KARST HYDROLOGY OF THE SIERRA DE EL ABRA, MEXICO3.2. Karst geomorphic features in the region 42 3.3 Summary 43 4. Hydrology and hydrogeology 45 4.1. Introduction 45 4.2. Temporal aspects

KARST HYDROLOGY OF THESIERRA DE EL ABRA, MEXICO

The Nacimiento del Río Choy in flood. Larger floodscompletely fill the entrance. Photo by Don Broussard.

KARST HYDROLOGY OF THESIERRA DE EL ABRA, MEXICO

John Fish

with a Foreword by Derek Fordand a Preface by Gerald Atkinson

ASSOCIATION FOR MEXICAN CAVE STUDIESBULLETIN 14

2004

AMCS Bulletin 14

This is a complete publication of a dissertation by JohnnieEdward Fish titled “Karst Hydrogelogy and Geomorphology ofthe Sierra de El Abra and the Valles–San Luis Potosí Region,Mexico” in 1977 for the degree of Doctor of Philosophy ingeology at McMaster University, Hamilton, Ontanio, under thesupervision of Derek Ford.

Cover photograph: The Nacimiento del Río Mante.The water rises from a shaft inside the cave that isat least 264 meters deep, the current limit of scubaexploration there. Photograph by Susie Lasko.

© 1977 Johnnie Edward Fish

Association for Mexican Cave StudiesPO Box 7672Austin, Texas 78713www.amcs-pubs.org

The AMCS is a Project of the National Speleological Society

Printed in the United States of America

4

AMCS Bulletin 14

Foreword by Derek Ford 7Preface by Gerald Atkinson 9Abstract 12Acknowledgements 13List of illustrations 14List of tables 16

1. Introduction 172. Geology and environment of the field area 21

2.1. Introduction and location 212.2. Geologic history 232.3. Stratigraphy and lithology 232.4. Structure and physiography 272.5. The environment 29

3. Geomorphic features of the surface 313.1. The Sierra de El Abra 313.2. Karst geomorphic features in the region 423.3 Summary 43

4. Hydrology and hydrogeology 454.1. Introduction 454.2. Temporal aspects of the hydrology 484.3. Spacial distribution of runoff, basin yield 614.4. Ground water in the Sierra de El Abra 674.5. Summary 72

5. Hydrogeochemistry 735.1. Introduction 735.2. Chemistry and environment of the karst waters 785.3. Flood pulse behavior of the springs 845.4. Summary 89

6. The caves 916.1. Introduction 916.2. Caves of the east face of the Sierra de El Abra 936.3. Caves on the crest of the Sierra de El Abra 1056.4. Caves on the western flank and western margin of the Sierra de El Abra 1146.5. Caves in other parts of the region 133

7. Analysis and synthesis 1357.1. Hydrochemical analysis 1357.2. The ground-water hydrology of the region 1507.3. Hydrology of the Sierra de El Abra 1617.4. Morphology and origin of the caves of the Sierra de El Abra 1697.5. Summary and conclusions 176

Appendix 1: Analytical methods and sampling procedures 179Appendix 2: Cave map symbols 179Bibliography 183

CONTENTS

5

AMCS Bulletin 14

The Nacimiento del Río Choy in 1891.Photographs taken by William H.Jackson on 18-by-22-inch glass plates,commissioned by the Mexican CentralRailway. A train can be seen on therailway bridge over the cave. Thisbridge was destroyed during the revo-lutions of the 1910–20 decade. Rubblefrom it still lies in the rapids outsidethe entrance. The bridge has beenreplaced, and a railroad still runs overthe cave. Low-resolution scans down-loaded from the Library of Congressweb site: above, call number LC-D43-1606; right LC-D43-1607.

6

AMCS Bulletin 14

FOREWORD

by Derek Ford

Johnnie Edward Fish was born in Kansas in thefall of 1942, attended local schools, and completeda BA degree with distinction in physics and math-ematics at Kansas State Teachers College in 1964.While there he became passionately interested incaving. This led him to shift his academic inter-ests towards earth sciences, and he enrolled in anMSc program in geophysics at the University ofTexas at Austin in the later ’60s. It was the time ofthe explosion of cave exploration in Mexico, ledby the AMCS. Johnnie came to the forefront of itsactivities. He was amongst the first to descendSótano de las Golondrinas and the deep caves inthe Huautla area, to explore the swallet caves onthe west flank of the Sierra de El Abra, and to chopthrough the dense thorn forest on the plateau abovethem to push down into the magnificent Hoya deZimapán and other great caverns there.

In those years I was building a graduate researchteam in cave science at McMaster University,Ontario. Our primary focus was on cave and karsthydrologic exploration in the Rockies and theSelkirk Mountains of western Canada, opening upalpine caves such as Nakimu, Castleguard,Gargantua, and Yorkshire Pot there. Weather prettymuch confines alpine scientific work to the briefsummer season. During the fall and spring aca-demic terms my gung ho cavers could get away toWest Virginia or Kentucky for long weekends. Forthe lengthier Christmas vacation they wantedsomething a bit meatier, however, and so headeddown to join the party at Huautla, exploring andmapping in Sótano del Río Iglesia, San Augustín,etc. Johnnie met them there over Xmas 1968–9,liked what he saw, and so applied to me for a PhDin cave studies, which he began in the fall of 1969.

We decided that his PhD research should attemptan integrated study of the surface landforms, caves,and dynamic hydrology of the karst around the

southern end of the El Abra mountain range, be-cause Johnnie already knew it quite well and itwas comparatively easy to get around in. Over thenext three summer field seasons he, with the re-doubtable Don Broussard and other assistants,roughly doubled the extent of known and mappedcaves in the region. Basic temperature and pre-cipitation data were available for some of theneighbouring towns, and so were discharge datafor El Choy and El Coy, the principal springs.Nothing was known of the swallet hydrology orwhat happened beneath the plateau, however. Withhard but meticulous fieldwork, John and Don de-termined that . This included such epic efforts asinstalling the large and clumsy water-level record-ers of that era over the terminal sump in Sótanode Jos and above the lake at the bottom of the 200m shaft in Soyate; the aquifer dynamics during thesucceeding hurricane season proved deleterious tothe health of the instruments, but they died in aworthy cause.

The first season’s fieldwork yielded sufficientinformation for the U.S. Karst Committee of theInternational Hydrological Decade to organize aspecial field visit to the region in April the fol-lowing year. At the close of the final season, in1973, John was honoured with an invitation to bea lead speaker at the prestigious O. E. MeinzerSymposia of the Geological Society of America.Such highly focused and demanding work canstrain personal relationships, however; John andhis first wife parted in 1974. He left McMaster in1975 with his second wife and an appointment atthe University of Indiana, and he was able tofinish the thesis late in 1977. Soon after, it wasacclaimed by all the external and internal examin-ers, and I hoped that he would then prepare a setof important papers from it. But John decided thatthe academic life was not for him and, in 1978,

7

AMCS Bulletin 14

took a responsible position in the Miami officesof the Water Resources Division of the U.S. Geo-logical Survey. The karst aquifer around Miami isarguably amongst the most fragile and stressed inthe United States. He worked on its many prob-lems until the close of the ’80s, when he electedto return to the faith of his childhood, resignedfrom geoscience, and entered the Christian minis-try.

Johnnie’s PhD thesis is large in scope, attempt-ing to describe, interpret, and integrate all perti-nent lithologic and tectonic, karst and cave, andhydrologic and climatic data for a large geographi-cal area in one major synthesis. Readers must judgefor themselves how well it succeeds.

Such overarching regional studies are out offashion these days, being replaced by theses com-posed of a few short papers on closely related top-ics. Parts of the speleological analysis are now alittle dated, chief among them our failure in the

Johnnie Fish prepares to lower a board for mounting a stagerecorder into Sótano del Soyate. Photo by Don Broussard.

1970s to appreciate the quantitative significanceof corrosion by condensation waters, which canape the effects of renewed phreatic conditions. Theunderground hydrology and spring hydrochemistrythat Johnnie did were ahead of their time and standfirm today. For me, the analysis and regional inte-gration that concludes the work in Chapter 7 con-tinues to be the best study and explanation of deepphreatic, meteoric cave genesis, as I sought toemphasize by placing a summary of it at the frontof the case studies of meteoric caves in Speleo-genesis: Evolution of Karst Aquifers (Klimchouk,et al., eds., 2000, Huntsville, Alabama; NationalSpeleological Society). The Sierra de El Abra is adistinctive and wonderful karst region. I hope thatthis publication by AMCS will inspire cave scien-tists to return to it. In every regard, there is a lotstill to be discovered and understood.

February 2004

8

AMCS Bulletin 14

PREFACE

by Gerald Atkinson

In the years since the completion of John Fish’sthesis in 1977, there has been a marked decline incave exploration in the Sierra de El Abra range. In1977, there were approximately 175 known cavesin the region, with Hoya de Zimapán at –320meters and Sótano del Arroyo at 7200 meters be-ing the deepest and longest caves respectively. In-deed, Arroyo was the longest known cave inMexico at the time. Few El Abra caves have beendiscovered in the intervening years, and the titleof longest Mexican cave has long since passed onto other contenders, but what little activity has tran-spired has been interesting and significant.

In the mid-1970s, cavers became increasinglyaware that the El Abra was not their exclusivedomain. Encroachment by farming and mininginterests was beginning to have a significant im-pact on the area. The completion of the OtatesMine road in 1974–1975 provided both cavers andminers alike access to the heart of the range. Agri-cultural ejidos were appearing at the base of therange, and fences were becoming more common.Access was increasingly restricted, and the long-term preservation of the karst and ecosystem ofthe range was in serious doubt. It was a land intransition, and all indications suggested that itcould very well go the way of other areas in Mexicowhere conservation had not faired well.

In the same year that Fish completed his thesis,Mitchell, et al. (1977) published a landmark pa-per on the blind cave fish genus Astyanax of theeastern ranges of the Sierra Madre Oriental, witha major emphasis on the cave localities within theSierra de El Abra. Included within the paper weremore than sixty pages of descriptions and historyof exploration of the area’s caves, with additionalmaterial on the geology and geomorphology of therange. Though eclipsed by Fish’s monumental tome,it was nonetheless an important speleological

contribution that remains a key reference for work-ers of the area.

In December 1977, Mark Minton, Neal Morrisand others successfully completed their four-yearcampaign to bottom Cueva de Diamante. It hadtaken five expeditions and over fifteen hundredman-hours to reach a depth of 621 meters througha veritable jungle gym of tight canyons, razor-linedpits, and horrendous top-outs (Minton, 1978). Atthe time, Diamante was the fourth deepest cave inthe western hemisphere. The extreme difficulty ofthe cave has since deterred anyone from return-ing, though several leads remain, including a ma-jor passage at –430 meters that was taking water.

In 1979, Sheck Exley and others began what wasto be a ten-year odyssey to determine the depth towhich they could explore the Nacimiento del RíoMante using open-circuit scuba gear. In March1979, Sheck Exley and Paul DeLoach reached adepth of 101 meters in a steeply descending, nar-row rift that continued downward. The subsequentdevelopment of mixed-gas diving technology, us-ing mixtures of helium, nitrogen, and oxygenrather than compressed air in order to avoid suchultra-deep diving problems as nitrogen narcosisand oxygen toxicity, made deeper cave dives pos-sible. In April 1987, Sheck returned and soloed toa depth of 159 meters, setting a new Americandepth record for diving. Two months later, Sheckbroke the world surface-to-surface diving recordby reaching a depth of 201 meters at the naci-miento. Determined to push the limits of divingtechnology, Sheck returned in April 1988, and af-ter twenty-four minutes of descent reached a depthof approximately 238 meters before tying off hisline and beginning what would be nearly elevenhours of decompression during his ascent. Finally,in March 1989, Sheck reached a depth of 264meters in the cave during a fourteen-hour solo dive.

9

AMCS Bulletin 14

Sheck never returned to the Nacimiento del RíoMante after 1989; though he went on to pursuediving opportunities at other caves in Mexico. Hedied attempting to break the world diving depthrecord at El Zacatón in April 1994 (Exley, 1979;1988; 1995; DeLoach 1988).

The Nacimiento del Río Choy was pushed to adepth of 43 meters and a length of 379 meters be-yond the entrance-chamber sump by Sheck Exley,Dan Lenihan, Terry More, Ken Fulghum, JamieStone, Carol Vilece, Frank Fogarty, and PaulDeLoach in early 1979 (Exley, 1979). The diversdiscovered a large submerged canyon passage thatled to a series of airbells and continued beyondtheir farthest penetration.

DETANAL released the first provisional topo-graphic maps of the Sierra de El Abra in 1981,which greatly aided reconnaissance efforts in thetrackless jungles of the high range. Prior to this

release, cavers had had no accurate maps for usein navigating the region. Mark Minton and othersimmediately began a series of jungle chops to theSuper Sink, a large dolina located several kilome-ters north of the Otates Mine road. Unfortunately,only a few small caves were found, and interest inthe high range subsequently began to wane. Thelast major El Abra chop of significance was madein March 1986, when Mark Minton, GeraldAtkinson, and others discovered Hoya de losGuacamayos, a large bird pit in the northern endof the range that was bottomed at –151 meters(Horowitz and Minton, 1987).

In 1988, Jim Bowden and Steve Gerrard ex-plored the Nacimiento del Río Santa Clara to adepth of 76 meters and a length of approximately427 meters using scuba gear. They were forced toturn back in going passage due to the great depthsencountered at the termination of their dive

(Bowden, 1988; Anonymous, 1998).The 1989 Mexpeleo convention

injected a short-lived, but intense pe-riod of activity into the region. Fromconvention headquarters at the HotelCovadonga south of Ciudad Valles,cavers fanned out in all directions toexplore, survey, or visit the area caves.Notable accomplishments during theweek-long affair included a resurveyby Bill Farr and others of Sótano delTigre (about 1500 meters surveyed),and the beginning of the resurvey ofSótano de Venadito by Don Broussardand crew (Farr, 1990; Broussard,1990). Don, later assisted by Joe Ivy,led a series of trips to Venadito from1989 to 1991 and from 1997 to1999that finally brought the cave to alength of 3341 meters and a depthof 204 meters, with going leads(Broussard, 1999; Ivy, 1999). Effortsat the cave were put on hold duringthe mid-1990s, when Africanized-bees were found to be inhabiting theentrance area of the cave.

During the 1980s, a growing con-servation movement within Mexico

10

AMCS Bulletin 14

began to systematically identify areas within thecountry that were in need of preservation. Amongthe first to receive recognition was the RanchoCielo region in the Sierra de Guatemala, whichwas set aside as a UNESCO-sponsored biospherereserve in 1986. Biosphere reserves are defined asareas of high biological diversity that are inhab-ited by species that are considered to be endemicand either threatened or in danger of extinction.To be considered for designation, a reserve mustbe at least 100 square kilometers in area and in-clude at least one ecosystem that has not beensignificantly altered by human activity. Unlike na-tional parks, biosphere reserves allow people tocontinue to live within the reserve. In addition tobiological surveys and ecological studies, researchis encouraged on sustainable resource use in or-der to encourage the local communities to partici-pate in the protection of the plants and wildlife(Anonymous, 2003).

In June 1994, the Mexican federal governmentestablished 215 square kilometers of the centralportion of the Sierra de El Abra as the Sierra de ElAbra–Tanchipa biosphere reserve. The reserveboundaries are defined as being within a box withcoordinates 22.08° to 22.40° N, and 98.89° to99.01° W (Barragán, 2003; Gurrola, n.d.). Infor-mation on the current status of the reserve and anyongoing research programs is fairly non-existentat present. There have been reports that access tothe reserve is difficult to obtain. Regardless of itsimpact on future caving, the designation of theSierra de El Abra as a reserve for preservation andstudy can only be considered a positive step, giventhat the natural resources of the region were in-creasingly being pursued by private interests.

Finally, it should be mentioned that in his the-sis, Fish refers to an anticipated book by NealMorris regarding the El Abra caves. The book inquestion was never actually published, though thedrafted and printed cave maps were subsequentlycompiled into a folio that was produced in 1989(Morris, 1989). Much of the original survey datafor the caves of the region has unfortunately beenlost, but a considerable amount of descriptive in-formation on the El Abra caves has been compiled

by the Association for Mexican Cave Studies andis available to interested individuals.

References

Anonymous. 1998. Expeditionary log El Proyecto de BuceoEspeleologico Mexico y America Central, April 15–25,1998. http://www.onr.com/user/zacaton/, Journals [ac-cessed 3/12/2004].

Anonymous. 2003. Eco travels in Mexico: Mexico parksexplained. http://www.planeta.com/ecotravel/mexico/mexparks2.html [accessed 3/12/2004].

Atkinson, Gerald. 1982. An updated list of the caves of theSierra de El Abra. AMCS Activities Newsletter 12, pp.87–92.

Barragán, Salatiel. 2003. Biosphere reserves: Sierra del Abra– Tanchipa: Reducto de fauna y vegetacion neotropical.http://www.mexicodesconocido.com.mx/espanol/n a t u r a l e z a / r e s e r v a s _ b i o s f e r a / d e t a l l e . c f m?idcat=2&idsec=8&idsub=0&idpag=3613 [accessed12/10/2003].

Bowden, Jim. 1988. Mexico news: Tamaulipas. AMCSActivities Newsletter 17, p. 20.

Broussard, Don. 1990. Trip reports: Venadito revisited.Texas Caver, 35(1):3–5.

Broussard, Don. 1999. Personal communication.DeLoach, Ned. 1988. The Deepest Dive: A Study in Con-

trolled Paranoia. Ocean Realm, summer, pp. 80–89.Exley, Sheck. 1979. Nacimientos! Diving the big springs

of the Sierra Madre. AMCS Activities Newsletter 10, pp.23–31.

Exley, Sheck. 1988. World depth record broken in Mexico.AMCS Activities Newsletter 17, pp. 96–99.

Exley, Sheck. 1995. Caverns measureless to man. CaveBooks: Cave Research Foundation, 176 pp.

Farr, Bill. 1990. Trip reports: Return to Sotano del Tigre.Texas Caver, 35(1):8–9.

Gurrola, Gerardo. n.d. Reserva especial de la Biósfera Si-erra del Abra-Tanchipa: Informacion general. http://maya.ucr.edu/pril/reservas/sierradelabratanchipa/sierradelabratanchipa1.html [accessed 3/12/2004].

Horowitz, Jeff, and Mark Minton. 1987. Mexico news:Tamaulipas. AMCS Activities Newsletter 16, pp. 14–17.

Ivy, Joe. 1999. Mexico news: Tamaulipas. AMCS Activi-ties Newsletter 23, p. 20.

Minton, Mark. 1978. The Diamante story. AMCS Activi-ties Newsletter 8, pp. 6–15.

Mitchell, Robert W., William H. Russell, and William R.Elliott. 1977. Mexican eyeless Characin fishes, genusAstyanax: Environment, distribution, and evolution. TheMuseum, Texas Tech University, Special Publications No.12, 89 pp.

Morris, Neal. 1989. Sierra de El Abra cave map folio.AMCS, 10 sheets.

March 2004

11

AMCS Bulletin 14

ABSTRACT

The general objective of this work was to develop abasic understanding of the karst hydrology, the natureand origin of the caves, the water chemistry, the sur-face geomorphology, and relationships among theseaspects for a high relief tropical karst region having athick section of limestone. The Valles–San Luis Potosíregion of northeastern Mexico and, in particular, theSierra de El Abra was selected for the study. A Creta-ceous Platform approximately 200 km wide and 300km long (N-S) delimits the region of interest. A thickLower Cretaceous deposit of gypsum and anhydrite,probably surrounded by Lower Cretaceous limestonefacies, is overlain by more than 1000 m of the thick-bedded middle Cretaceous El Abra limestone, whichhas a thick platform-margin reef. The Sierra de El Abrais a greatly elongated range along the eastern marginof the Platform. During the late Cretaceous, the re-gion was covered by thick deposits of impermeablerocks. During the early Tertiary, the area was folded,uplifted, and subjected to erosion. A high relief karsthaving a wide variety of geomorphic forms controlledby climate and structure has developed. Rainfall in theregion varies from 250 to 2500 mm and is stronglyconcentrated in the months June through October, whenvery large rainfalls often occur.

A number of specific investigations were made tomeet the general objective given above, with specialemphasis on those that provide information concern-ing the nature of ground-water flow systems in the re-gion. Most of the runoff from the region passes throughthe karstic subsurface. Large portions of the region haveno surface runoff whatsoever. The El Abra Formationis continuous over nearly the whole Platform, and itdefines a region of very active ground-water circu-lation. Discharge from the aquifer occurs at a numberof large and many small springs. Two of them, the Coy

and the Frío springs group, are among the largestsprings in the world, with average discharges of ap-proximately 24 m3/sec and 28 m3/sec, respectively.Most of the dry season regional discharge is from afew large springs at low elevations along the easternmargin of the Platform. The flow systems give ex-tremely dynamic responses to large precipitationevents; floods at springs usually crest roughly one dayafter the causal rainfall, and most springs have dis-charge variations (Qmax/Qmin) of 25 to 100 times. Thesefacts indicate well-developed conduit flow systems.

The hydrochemical and hydrologic evidence, in com-bination with the hydrogeologic setting, demonstratethe existence of regional groundwater flow to severalof the large eastern springs. Hydrochemical mixing-model calculations show that the amount of regionalflow is at least 12 m3/sec, that it has an approximatelyconstant flux, and that the local flow systems providethe extremely variable component of spring discharge.The chemical and physical properties of the springsare explained in terms of local and regional flow sys-tems.

Local studies carried out in the Sierra de El Abrashow that large conduits have developed and that largefluctuations of the water table occur. The large fossilcaves in the range were part of great deep phreatic flowsystems that circulated at least 300 m below ancientwater tables and discharged onto ancient coastal plainsmuch higher than the present one. The western marginswallet caves are of the floodwater type. The caves arestructurally controlled.

Knowledge gained in this study should provide abasis for planning future research and, in particular,for water resource development. The aquifer has greatpotential for water supply, but little of that potential ispresently used.

12

AMCS Bulletin 14

ACKNOWLEGEMENTS

A large number of persons and organizations have as-sisted this research. Their contributions are consideredinvaluable, and the author wishes to thank the follow-ing:

Prof. Derek C. Ford for acting as supervisor and forhis encouragement and patience.

Prof. G. V. Middleton and Prof. Brian McCann fortheir guidance and helpful criticisms as members ofthe thesis committee. Prof. James R. Kramer for dis-cussion of the hydrochemistry.

Sandra Patrick, my wife, for the great amount ofassistance she offered in the field, typing, drafting,editing, and discussing the thesis, and for her spiritualsupport.

Ing. Francisco Lavín Ortiz of the Tampico office ofthe Secretaría de Recursos Hidráulicos, who provideda large amount of unpublished hydrologic and climaticdata.

Ing. Diego A. Cordoba, director of the Instituto de

Geología, who provided assistance in many ways, andProf. Ernesto Lopez Ramos of the same organizationfor providing helpful geologic information.

The help and companionship in the field of manycavers, especially Don Broussard, who was my fieldassistant, Steve Bittinger, Neal Morris, who also hasprovided some maps, Craig Bittinger, Robert Hemperley,Logan McNatt, Skip and Kathy Roy, and William H.Russell.

Fellow graduate students Julian Coward, Peter Thomp-son, Melvin Gascoyne, Ralph Ewers, and John Drakefor assistance, discussion of thesis and other mattersof interest, and for friendship.

The foreman of Rancho Florida for granting accessto the Choy spring.

Señor Gloria, owner of the Restaurant Condesa inCd. Valles, for help with many matters.

Robert Bignell, who printed the photographs.Martha Nixon, for typing the final manuscript.

13

AMCS Bulletin 14

ILLUSTRATIONS

Figures (drawn)

1.1 Geography and physiography of the Valles–San LuisPotosí region

2.1 Middle Cretaceous reef trend around Gulf ofMexico and Valles–San Luis Potosí Platform

2.2 Geologic cross section of the Valles–San Luis PotosíPlatform during the Cretaceous

2.3 Gross carbonate facies distribution on the Valles–SLP Platform

2.4 Platform limits during the Cretaceous2.5 Structural map of the Valles–SLP Platform

3.1 Sierra de El Abra and surrounding area3.2 Geomorphic and hydrologic map of southern Si-

erra de El Abra (foldout)3.3 Geomorphology of North El Abra Pass area3.4 Map of Caldera, a large collapse sinkhole

4.1 Hydrogeologic map of Valles–SLP Platform4.2 Drainage basins and gauging stations of the region4.3 Annual rainfall in the region4.4 Annual flow of Choy spring compared with rain-

fall for 1960-19714.5 Monthly average, standard deviation, and coeffi-

cient of variation of discharge and precipitation atEl Choy station for 1960-1971

4.6 Monthly average and coefficient of variation of dis-charge of Río Coy (Estación Ballesmí) and RíoTampaón (Estación El Pujal), for 1960-1971

4.7 Choy and Coy hydrographs and rainfall for Jan. 1,1971–June 30, 1973 (foldout)

4.8 October 1966 flood events—Hurrican Inez4.9 Hydrographs of Río Mante and Frío springs in 19514.10 Río Santa Clara hydrograph, June–October for

1972, 19734.11 Magnitude of flooding of springs and rivers as

shown by flood exceedance4.12 Flow duration graph of Choy spring for 1960–19714.13 Soyate and Jos cave-lake hydrographs compared

with rainfall and Choy discharge, July and August1971

4.14 Soyate cave-lake hydrograph and correlation withrainfall and Choy discharge, June 1972

4.15 Flood events at Jos, Soyate, and Choy, September,1971

4.16 Relationship between Soyate cave lake stage andChoy discharge

5.1 Hydrochemical environments in the Sierra de ElAbra

5.2 Comparison of chemistry of water samples fromvarious environments in the Sierra de El Abra

5.3 Ca/Mg against SO4 for El Abra region karst water5.4 Temperature regime of springs in the region5.5 Trilinear plot of samples from six springs5.6 Effect of flooding on chemistry of Choy spring,

June–August, 19715.7 Effect of flooding on chemistry of Choy and Coy

springs, May–June l9725.8 Relationship between Mg and SO4 for Choy, Coy,

and Mante during flooding5.9 Relationship between T (°C) and SO4 for Choy, Coy,

and Mante during flooding5.10 Effect of flooding on degree of saturation with re-

spect to calcite, dolomite, and gypsum at Choy andCoy, May–July 1972

5.11 Effect of flooding on chemistry of Taninul SulfurPool, June–July, 1971

6.1 Map of Cueva del Nacimiento del Río Choy6.2 Maps of Cuevas de Taninul No. 1 and 2 and the

Taninul springs6.3 Profile of Cuevas de Taninul No. 1 and 26.4 Map of Grutas de Quintero (foldout)6.5 Map of Cueva de las Cuates6.6 Map of Cueva de la Ceiba6.7 Map of Ventana Jabalí6.8 Profile of Ventana Jabalí6.9 Map of Sótano de Escalera6.10 Map of Cueva Pinta6.11 Map of Sótano de la Cuesta (provided by Neal

Morris)6.12 Map of Sótano de los Loros (provided by Neal

Morris)6.13 Map of Cueva de Taninul No. 46.14 Map of Cueva de Tanchipa (provided by Neal Mor-

ris)6.15 Map of La Hoya de Zimapán (provided by Neal

Morris)6.16 Map of Cueva and Sótano de los Monos (foldout)6.17 Map of Sótano de Soyate6.18 Map of Cueva del Prieto6.19 Map of Sótano de Japonés (foldout)6.20 Map of Sótano de Jos6.21 Map of El Sistema de los Sabinos6.22 Map of Cueva de los Sabinos (foldout)6.23 Map of Sótano de Arroyo (foldout)6.24 Map of Sótano de la Tinaja (foldout)6.25 Profile of part of Sótano de la Tinaja6.26 Maps of Las Cuevas y El Nacimiento del Rio Coy

18

21

23

24

2727

3335

3738

46484952

54

55

57

565659

60

6169

69

70

71

73

79

81828384

85

86

86

87

88

9294

959998

101102102104106107

108

109110

111

100114115

117–118122123117120127126134

14

AMCS Bulletin 14

7.1 Plot of Mg against SO4 for best base flow analysisof thirteen springs

7.2 Two-source mixing model for the Choy and Coysprings

7.3 The hydrochemical model7.4 Evolutionary calcite solution diagram, HCO3

against pH7.5 Ca against SO4, and Ca + Mg against SO4 for springs

and the theoretical models of type B ground water7.6 Mixing curves of type A with type B ground water7.7 Plot of Ca++ against T (°C) for various geographi-

cal regions from Harmon et al. (1973), with a newMexico point from this study

7.8 Calcite saturation index (SIc) versus negative loga-rithm of partial pressure of carbon dioxide (PPCO2)for region springs

7.9 Regional E-W hydrogeologic cross section show-ing hypothesized local and regional ground-waterflow systems

7.10 Map summarizing spatial distribution of runoff inthe region, possible directions of ground-water flow,base flow chemistry, discharge and elevation of themajor springs

7.11 Hypothesized regional dry season potentiometricsurface

7.12 N-S cross section of front ranges of Sierra MadreOriental

7.13 Comparison of elevation of cave lakes and deepestpoint known in caves with radial distance to andelevation of El Choy rise pool

7.14 Hypothesized N-S dry season ground-water profilein the Sierra de El Abra

7.15 Hydrographs in February, 1957, of Frío springs,Mante, and Río Guayalejo showing an unusual, iso-lated dry season flood

7.16 Comparison of Mante, Santa Clara, and RíoComandante hydrographs and precipitation forJuly–August 1972

7.17 E-W cross sections showing relationships of geol-ogy, hydrology, topography, and caves in the Si-erra de El Abra

7.18 Models of the east face caves

Plates (photographs)

2.1 View of the thick-bedded El Abra limestone2.2 View of the thick-bedded El Abra limestone

3.1 Aerial view of the Gulf coastal plain and Sierra deEl Abra

3.2 Flat-topped hills: remnants of an older coastal plainsurface

3.3 Typical relief on the coastal plain near the El Abrarange

136

136

137138

139

144149

150

152

153

156

157

164

166

167

169

170

172

2424

32

34

34

34

34373738

40

42

93

93

95959697

103103103

104104

105105

108108109109

112112

112

113

123124125

125125130130

130131133

3.4 The eastern escarpment of the El Abra range in thehigh central zone

3.5 The Mante spring3.6 Fracture-controlled recesses in the east face3.7 The surface of the crest of the El Abra range3.8 The Caldera, a giant collapse doline on the crest of

the El Abra3.9 Stream capture along the western margin of the

range3.10 The spectacular Xilitla-Aquismón area karst

6.1 View from inside Choy cave looking out the loweror Resurgence Entrance

6.2 El Choy ceiling showing phreatic pocketing anddomes

6.3 The Choy spring at base flow6.4 The Choy in moderate flood6.5 The Entrance Passage of Grutas de Quintero6.6 The Main Passage of Quintero6.7 View out of the entrance of Ventana Jabalí6.8 The immense chamber of Ventana Jabalí6.9 The high dome and bedrock natural bridge in

Ventana Jabalí6.10 The rear portion of Ventana Jabalí6.11 A 5 m sediment section exposed in a test pit near

the entrance of Ventana Jabalí6.12 The entrance to Cueva Pinta6.13 View toward the entrance in Cueva Pinta from near

the test pit in sediments6.14 Phreatic fissure in Tanchipa6.15 The entrance to La Hoya de Zimapán6.16 The nearly circular entrance passage of Zimapán6.17 The stalagmite barrier across the end of the entrance

passage in Zimapán6.18 The Zimapán Room6.19 Large stalagmites in Zimapán that have undergone

“phreatic” re-solution6.20 Partially redissolved stalagmite and solution fea-

tures in the Zimapán Room6.21 A pocket dissolved by a “phreatic” re-solution phase

in Zimapán6.22 Entrance to Cueva de los Sabinos6.23 The entrance pit of Sótano del Arroyo6.24 View out Main Passage of Sótano del Arroyo into

the entrance sink6.25 Main Passage in Sótano del Arroyo6.26 A log in Main Passage in Arroyo6.27 Entrance to Sótano de la Tinaj6.28 Entrance Passage in Tinaja just above Cable Lad-

der Drop6.29 Part of the Sandy Floored Passage in Tinaja6.30 An old fill in the Sandy Floored Passage6.31 The slot-like entrance to Sótano de Yerbaniz

15

AMCS Bulletin 14

1.1 Translation of Spanish Words

2.1 Correlation of Stratigraphic Units2.2 Mean Monthly and Annual Air Temperatures (°C)2.3 Daily Maximum and Minimum Temperatures at

Estación Santa Rosa (Cd. Valles)

3.1 Drainage Area of the Swallet Caves

4.1 Gauging Stations in the Valles–SLP Region4.2 Selected Climatology Stations in the Region4.3 Longer Term Rainfall Compared with Base Period

Averages4.4 Monthly Distribution of Gulf Coastal Zone Hurri-

canes, 1921–614.5 1951 Rainfall (mm)4.6 Low Flows and Low Flow Budget4.7 Special Gaugings4.8 Indirect Calculations of Flow of Smaller Frio

Springs4.9 Runoff from Basins, 1961–19684.10 Annual Regional Runoff4.11 Runoff of Full Record Compared with 1961–1968

Average4.12 Unitary Discharges of Various Basins

5 .1 Chemical Analyses and derived Geochemical Mea-sures

5.2 Analyses for Na, K, and Cl5.3 CO2 of Soil and Cave Air

7.1 Equilibrium Constant Functions7.2 Possible Theoretical Compositions of Type B

Ground Water7.3 Mixing Model Calculations for the Choy and Coy

Springs7.4 Dry Season Regional Flow (QB)7.5 Flood Dilution Mixing Model Calculations7.6 Equilibrium Model with Greater Kd7.7 Some Published Data for Comparison with the

Geochemical Model7.8 Ratio of Minimum to Average Flows for Basins7.9 Data Used to Construct the Potentiometric Map7.10 Calculated Temperatures for Type B (Regional)

Water7.11 Drainage Area of the Sierra de El Abra “Basin”7.12 Cave Survey Data and Elevations for Water-Level

Profile7.13 Base-Level Points of the Sierra de El Abra

TABLES

19

222829

40

505253

53

53626363

646566

67

74

8081

137140

143

145146147148

154157160

162163

165

16

AMCS Bulletin 14 — Chapter 1

1

INTRODUCTION

There has been exploration and utilizationof caves and springs for at least severalmillennia. A classic example is the use ofcenotes in Yucatan by the Mayas (JohnLloyd Stephens, 1842; Rack and Hanshaw,1974). There were occasional astute obser-vations, but more commonly there weremyths or expressions of curiosity concern-ing these features as well as the exotic karstlandforms in some regions. The scientificstudy of karst did not begin in earnest untilabout one century ago (Roglic, 1972).Some of the work was prompted by water-supply or engineering activities such as rail-road building. Other studies of a broadernature were made by physical geogra-phers and geologists, essentially concurrentwith the growth of geomorphology as ascience.

As a result of the past century of explo-ration and study, the karsts of Europe andthe United States have become famous evenamong the general public. Many conflict-ing theories of origin or process have beengiven, and there has been much debateabout the relative importance of variousfactors, such as climate and rock structure,in the development of karst features andabout the nature of karst ground-water cir-culation. It appears that the study of karstrequires the integration of knowledge andtechniques from as many separate disci-plines as any subject in geology (to place itin a category). For example, hydrology,aqueous and isotope geochemistry, geomor-phology, climatology, petrology and sedi-mentary processes, structural geology, andfluid mechanics all have an important bear-ing on karst studies. Modern work has avery strong emphasis on karst hydrologyand related fields, not merely for academicenlightenment, but for development ofwater resources and control of environmen-tal problems such as pollution of carbon-ate aquifers. The great preponderance ofresearch has been conducted in the middlelatitudes, although carbonate rocks form asignificant portion of the sedimentary rockcolumn in most parts of the world. Al-though there have been many significant

contributions, work in tropical karsts hasbeen limited, most of it has been largelydescriptive, and much of it has dealt pri-marily with surface landforms. It was ageneral goal of this dissertation research tostudy in as great a depth as possible onetropical karst area, emphasizing thehydrogeology and hydrology, the develop-ment of caves, and the relationship of thesurface geomorphology to the first threesubjects.

It was especially desirable that suchstudies be undertaken in an area having athick, well karstified carbonate-rock sec-tion and high relief, because so much ofthe work done in the United States has beenconducted where the relief is low to mod-erate and the section of carbonate rocksseldom exceeds 200 m. Personal familiaritywith many karst areas in Mexico indicatedthat they were of scientific and practicalimportance. Excepting the Yucatan Penin-sula, most of the rest of the Mexican karstlies in extremely rugged mountainous ter-rain posing severe logistic problems, andis thus poorly known. The area selected,the Sierra de El Abra in northeasternMexico (Figure 1.1), offered caves in avariety of geologic and geomorphic settings(although not many were actually surveyedwhen the research was begun), some verylarge springs and many smaller ones, rela-tively high relief, a thick carbonate rocksection extending below base level, a longerosional history so that the paleohydrologycould be examined, and perhaps the mostreasonable access (although still difficult)of any of the Mexican karst areas. Somegeologic information was available becausethe karstic limestone that outcrops there isa prolific oil producer where buried in otherparts of Mexico. Also, there were exten-sive runoff and climatic data published forthe much larger surrounding region, al-though there was little information concern-ing individual springs. Aside from someearly work on cave fauna in the region, F.Bonet (1953a) conducted the first speleo-logic and karst morphologic investigationsin the Sierra de El Abra, where he examined

part or all of 15 to 20 horizontal caves, andin the nearby Xilitla karst area (1953b). Inthe 1960s, the Association for MexicanCave Studies (based in Austin, Texas) be-gan exploring and surveying caves inMexico; it has published two bulletins(Russell and Raines, 1967; and Raines, ed.,1968) which briefly describe a number ofcaves in the El Abra range and in other partsof the Sierra Madre Oriental in northeast-ern Mexico. A recently published study ofblind fish caves in the El Abra (Mitchell,Russell, and Elliott, 1977) utilizes a largeamount of their independent work plus in-formation from this research and othermembers of the Association for MexicanCave Studies. The observations and ideascontained in these reports as well as somecomparisons with other areas will be re-ferred to in appropriate discussions in laterchapters.

Initially, it was planned to confine thestudy to the Sierra de El Abra, with only aminor discussion of the regional hydrologyand geomophology. However, it becameclear from the first results of the waterchemistry, from observations of the hydrol-ogy, and from analysis of the geologic re-lationships that a much broader regionaltreatment was both desirable and in factnecessary to solve many of the fundamen-tal problems of the El Abra. Based on thegeology and the hydrology, it was ulti-mately judged that the area of interest wasa paleocarbonate platform measuring about200 by 300 kms (Carrillo Bravo, 1971) thatextends from Tamuín to San Luis Potosíand from Jalpan as far north as Cd. Victoria(Figure 1.1). This region includes parts ofthe states of San Luis Potosí, Tamaulipas,Querétaro, Guanajuato, Hidalgo, andNuevo León. It was decided that the re-search would comprise two levels of study:the Sierra de El Abra would be studied ex-tensively in the field to develop a detailedunderstanding of its hydrology, cave devel-opment, geomorphology, and water chem-istry; and the regional hydrology andhydrogeology would be analyzed fromgauging station data, the published geologic

17


Figure 1.1. Geography and physiography of the Valles–San Luis Potosí region.

literature, and from reconnaissance hydro-logic and geomorphic observations andwater sampling. Specifically, the thesisobjectives are: For the region, to determinethe hydrogeologic units and the regionalhydrogeologic framework including geo-morphic relationships, to analyse quantita-tively the hydrology, including regional andlocal runoff, magnitude and importance ofkarst springs to the total runoff, flow be-havior (temporal variations) especially ofthe springs, and character of the regionalground-water body, and to conduct a

hydrochemical survey of springs for flowsystem analysis and comparison with otherregions. For the Sierra de El Abra, to studyin detail the contemporary hydrology, in-cluding the water budget, ground-waterlevels, calculation of hydraulic gradients,distribution of ground water, and dynam-ics of the aquifer, to make detailed surveys,descriptions, and morphologic analyses ofcaves in all geomorphic positions across therange, to analyse the role of caves in con-temporary or paleohydrology, to place thecaves in the general theory of cavern genesis

and karst ground-water circulation, to de-scribe the geomorphic features and theirrelationship to the caves and the hydrol-ogy, and to establish the place of the Sierrade El Abra in the region.

To attain the above goals, support wasobtained from agencies of the Mexicangovernment. The Instituto de Geología(geological survey), directed by Ing. DiegoA. Córdoba, provided geologic maps andreports, air photos of the El Abra range, apickup truck for some of the more ruggedfield work, and other assistances as well.

18


The Sectretaría de Recursos Hidráulicosprovided two bulletins of published hydro-logic and climatologic data for the Panucobasin, which includes all of the region ofinterest, and their Tampico office, directedby Ing. Francisco Lavín Ortiz, made avail-able a large quantity of unpublished dis-charge and rainfall data. A major hindranceto this and other geologic field studies wasthe lack of good topographic maps. The old1:100,000 maps have only a 50 m contourinterval, cover only part of the region, andare often grossly in error. Presently, theMexican government is producing a goodquality new series at 1:50,000 having acontour interval of 10 or 20 m; unfortu-nately, they were begun after the field workfor this project was completed. Most of theSierra de El Abra is not yet available, butabout 80% of the region of interest is nowcovered. Air photos were used for study ofsurface features, for guidance in the field,and for preparation of some maps. Themajority of the field work was accom-plished during the 1971 and 1972 summerfield seasons; three other trips were made,including one during December 1971–January 1972 to obtain some early dry sea-son data and conduct some cave surveyswhen there was less danger of flooding. Oneach trip, a base of operations was estab-lished in Ciudad Valles to protect the volu-minous equipment, set up a laboratory, andprovide refuge.

The results of this study are reported inthe following pages. Each chapter from 2through 6 will concern a separate subject

area, presenting considerable informationand analysis. Chapter 2 gives the geologicbackground and some aspects of the envi-ronment. Chapter 3 focuses on the geomor-phic features of the El Abra range and sur-rounding area, and briefly describes karstdevelopment in other parts of the region.Chapter 4 establishes the regional hydro-geologic framework, provides a quantitativetreatment of spatial and temporal charac-teristics of the regional hydrology, empha-sizing karst springs, and discusses someaspects of ground water in the Sierra de ElAbra. In Chapter 5, the chemistry of vari-ous karst waters is examined, and variations

Table 1.1Translation of Spanish words

cañón canyon

ciudad city, town

cueva cave, usually with a horizontal or sloping entrance

estación station, such as Estación Ballesmí (gauging station)

gruta cavern

manantial spring, fountain

municipio municipality, township

nacimiento spring, source (of river)

pozo well

presa dam

río river

sierra mountain range

sótano pit

in concentrations that occur in some springsduring flood pulses are described. Chapter6 gives maps, descriptions, and morphologicinterpretations of the caves as individualentities. In Chapter 7, data and interpreta-tions from previous chapters are summa-rized, further analyzed, and synthesizedinto models and speculative discussions ofthe karst ground-water systems and thecaves. Also given are a summary and con-clusions.

Mexican geographic names are gener-ally used in this thesis; Table 1.1 providesa translation.

19


2

GEOLOGY AND ENVIRONMENT OF THE FIELD AREA

2.1. Introduction and Location

The purpose of this chapter is to pro-vide the geological framework for the dataand interpretations of subsequent chapters.This section will locate and identify theregion of interest. It will be followed by abrief geologic history, stratigraphic andlithologic descriptions, the regional struc-ture and physiography, and some environ-mental considerations. The stratigraphysection is lengthy because important recentknowledge is not covered in the Englishliterature. A general location map of theregion is given as Figure 1.1, and a moredetailed map of the Sierra de El Abra andenvirons is given in Chapter 3 as Figure3.1.

The area considered is the Valles–SanLuis Potosí Platform (Carrillo Bravo, 1971;Figure 2.1), and in particular the Sierra deEl Abra, which lies along its eastern mar-gin. It is part of a great Cretaceous carbon-ate complex that nearly encircles the Gulf ofMexico. In Mexico along the west-centralportion of the complex there has been ex-tensive uplift, deformation, and erosion that

has led to the development of high reliefkarsts which contrast with the low reliefkarst plains of the Yucatan and Florida pen-insulas (Fish, 1977). Certain aspects of thefield area, particularly the hydrology andwater chemistry, cannot be fully understoodexcept in the broader context of the wholePlatform. Thus, some parts of this study willdeal with a large portion of the Platform,while the remainder will focus on the Si-erra de El Abra.

Several studies in recent years, espe-cially by Pemex geologists, have consider-ably clarified the picture of the Cretaceoussystem in northeastern Mexico, althoughmuch remains to be done. Enos (1974) hasprovided a good general description of thecarbonate complex. He distinguishes threemajor facies and paleoenvironments: fine-grained cherty basinal limestones, platformlimestones and dolomites (El Abra Forma-tion in the middle Cretaceous), and basinmargin carbonate debris surrounding plat-forms. The three principal platforms in theregion were the large Valles–San LuisPotosí Platform, the Golden Lane Platformsouth of Tampico (which has been Mexico’s

most prolific oil producer), and the El Doc-tor Platform, south of the Valles–SLP Plat-form. (It should be remembered that thesewere paleoplatforms and the names are usedto denote the area of interest, even thoughtectonism has since altered their character.)Enos (1974, pp. 801–802) succinctly sum-marizes the character of the margins as fol-lows:

Several lines of evidence indicate thatthe middle Cretaceous platforms werebounded by steep edges, probably asthe result of rapid vertical accretion ofthe platform margin facies (reefs). Seis-mic reflection profiles (Rockwell andGarcia Rojas, 1953) and structural con-tour maps (Guzman, 1967, Fig. 7;Viniegra and Castillo-Tejero, 1970,Fig. 5) of the Golden Lane platformindicate relief of about 1000 m at theplatform edge by the close of the middleCretaceous. The east edges of the ElDoctor platform and of the Valles plat-form north of Valles form prominentescarpments which appear to reflectdepositional topography, although therelations in both are complicated byfaulting. Well records (Viniegra andCastillo-Tejero, 1970, Figs. 6,7) andisopach maps (courtesy of PetróleosMexicanos) of the middle Cretaceousrocks along the west side of the GoldenLane platform show that the platformfacies is about 1000 m thicker than thebasinal facies. Finally, the sedimentaryfacies show abrupt changes at the plat-form edges and indicate very little lat-eral migration with time.

The great difference in thickness, thestructure contour map, and the change offacies are critical evidence that the basinmargin carbonates (Tamabra Limestone ofthe Poza Rica trend near the Golden Lane)are forereef and turbidite deposits ratherthan downfaulted or downfolded true reefrocks. (Coogan et al., 1972, advocate thatthey are the “true reef” rocks of the El Abracomplex.)

Figure 2.1. Middle Cretaceous reef trend around Gulf of Mexico andValles–San Luis Potosí Platform.

21


Table 2.1Correlation of stratigraphic units*

22


The Sierra de El Abra is the classic areafor studying the middle Cretaceous reeftrends in outcrop, and several reports havebeen written about it as well as about theother “front ranges” of the Sierra MadreOriental. However, there was little infor-mation concerning the central and westernportions of the Platform until the regionalsynthesis of studies by Pemex and otherswas published by Carrillo Bravo (1971).Indeed, this was the report that first identi-fied the Platform, and the geologic infor-mation that follows relies heavily on it.

2.2. Geologic History

The geologic history of the region,largely adapted from Carrillo Bravo (1971),is summarized as follows:

During the late Precambrian, a regionaldiastrophism produced the majority of themetamorphic rocks found in the subsurface.

In the early Paleozoic, an intercontinen-tal geosyncline began to form in the regionof the Mexican high plains (or plateaus) andthe Sierra Madre Oriental (Figure 1.1).Considerable subsidence occurred duringthe Carboniferous and the Permian and alarge volume of orogenictype terrigenousrocks was deposited. In the closing stagesof the Paleozoic or the beginning of theTriassic the region emerged and wasstrongly folded.

All of the region remained emergent inthe Triassic and through the middle Juras-sic. Erosion was great in some places, andthick continental red beds were depositedin others. Along the western and southernmargins, marine basins developed. At theend of the early Jurassic, vertical move-ments and faulting occurred.

A regional marine transgression beganin the late Jurassic. Large areas remainedemergent; other areas may have received athin section of sediments. To the west, theMesozoic Basin of central Mexico wasformed. The “Platform” was a stable unitduring the Triassic and Jurrassic, receiv-ing almost exclusively continental red beddeposits, whereas numerous basins hadformed on the margins. By the close of theJurassic, the inundation was nearly com-plete, setting the stage for the Cretaceous.

The Cretaceous was the time of devel-opment of the great carbonate platforms inthe Gulf region, continuing to the presenton the Yucatan Platform. On the Valles–S.L.P. Platform, it began with a thin sec-tion of clay, sand, and calcareous sedimentsin the early Neocomian. From the geology,it is inferred (but not adequately confirmed

as yet by drilling) that large reef growthsdeveloped on the margins of the Platformfrom the middle Neocomian through theAptian. Within the Platform (lagoon?) athick evaporitic sequence, principally ofsulfate rocks, was deposited. A slight in-flux of argillaceous sediment deposited inthe deeper water surrounding the Platformmarks the beginning of more rapid subsid-ence at the close of the Aptian.

During the Albian-Cenomanian (middleCretaceous) maximum reef-platform de-velopment and platform-margin relief oc-curred. This gave rise to the Sierra de ElAbra reef, the Cerro Ladron bank, theJacala bank, and other reefs. In the centralzone, a thick section of backreef or plat-form carbonates was deposited.

In the Turonian, most of the Platformwas emergent. A transgression began on themargins, depositing pelagic limestones andcalcareous shales in the east and north, andclay, sand, and calcareous sediments in thesouth and west. Platform carbonate depo-sition occurred in the central region duringthe Senonian-Maestrichtian, while progres-sively siltier and shalier limestone and fi-nally a thick section of calcareous shaleswere deposited on the margins. Transgres-sive and regressive movements near the endof the Cretaceous covered the central andwestern portions of the Platform with athick section of clay and sand and signaleda new period of instability.

At the beginning of the Tertiary, the seawithdrew permanently and the area becameaffected by the Laramide Orogeny. Carbon-ate deposition ceased entirely, and early

Tertiary terrigenous deposits in the paleoGulf of Mexico (east of the Platform) recordthe accelerating diastrophism in the SierraMadre Oriental. Some granite bodies wereemplaced. The Tertiary and Quaternarywere times of uplift, erosion, and someterrestrial deposition. Extensive rhyoliticoutpourings occurred in the southwesternportion of the region during the middle Ter-tiary. There are some remnants of Pliocene-Pleistocene basalt flows in valleys of thecentral and eastern portion of the Platformand basalt-capped mesas on the coastalplain to the northeast.

2.3. Stratigraphy and Lithology

This section describes the stratigraphyand lithology of the sedimentary rocks thatoccur in the region, and briefly indicatesthe minor occurrences of igneous rocks.The age and correlation of the sedimentaryunits may be seen in Table 2.1.

Sedimentary Rocks

Triassic

Huizachal (La Boca) Formation. TheHuizachal Formation (redbeds) in theHuizachal-Peregrina anticline consists ofa sequence, locally more than 2000 m thick,of red, gray, and green shale, sandstone, andconglomerate. The formation containsUpper Triassic and Lower Jurassic flora,including silicified tree trunks. It is also ex-posed in the Miquihuana arch and has beenreached by wells on the Platform.

Figure 2.2. Geologic cross section of the Valles–San Luis Potosí Platformduring the Cretaceous.

23


Plates 2.1 (top) and 2.2 (bottom). Two viewsof the thick-bedded El Abra limestone. Thesesections occur in a quarry in the south ElAbra Pass, about 1.5 km west of the reeffacies. There are a few thin shale partingshere that become slightly thicker farther west.Note the numerous solution openings in 2.1and the thin to spotty soil cover in 2.2.

Figure 2.3. Gross carbonate facies distribution on theValles–San Luis Potosí Platform.

24


Jurassic

La Joya Formation. The La Joya For-mation west of Cd. Victoria generally con-sists of less than 65 m of red beds made upof a poorly sorted basal conglomerate, over-lain by thin limestones, mudstones, andsandstones of about middle Jurassic age(Mixon, 1963).

Olvido Formation. The Olvido Forma-tion in northern San Luis Potosí, Tamau-lipas, Nueva León, and Coahuila is a lateJurassic sequence of approximately 150 mof anhydrite, gypsum, limestone, and shale(Carrillo Bravo, 1963).

Tamán Formation (Mixed Facies). TheMixed Facies of the Tamán Formation isfound in wells along the east margin of thePlatform and designates the interdigitationof basin and platform rocks. This late Ju-rassic unit has a basal conglomerate ofquartz fragments and clay overlain by upto 200 m of fine-grained microcrystallinelimestone with frequent layers of ooliticlimestone and calcarenite. Some doleriteintrusions were encountered in the Tama-lihuale No. 1 well. (Locations of wells areshown in Figures 2.2 and 2.3.).

Pimienta Formation. Twenty-eightmeters of this dark compact argillaceouslimestone was found in Tamalihuale No. 1,concordantly above the Tamán and belowthe Lower Tamaulipas (Carrillo Bravo,1971). This late Jurassic formation cropsout south of Tamazunchale as a black chertylimestone and shale.

La Casita Formation. The La Casita For-mation is a series of shale, siltstone, sand-stone, conglomerate, and limestone of lateJurassic age that forms a thin transgressiveunit over the Platform.

Cretaceous

Lower Tamaulipas Formation. TheLower Cretaceous Lower Tamaulipas for-mation refers to a dense, dark, stylolitic,cherty lime mudstone considered to haveformed as an open-sea basin deposit in theTampico Embayment (that part of the an-cient Gulf of Mexico east of the Valles–SLP Platform). About 30 m has been foundin wells in the El Abra region, where it iscoarser grained, includes some quartzgrains, and is partially dolomitized. Out-crops are known on the western and north-ern margins of the Platform.

“Platform Limestone without FormalNomenclature.” This unnamed Lower Cre-taceous unit consists of 300 m of platformcalcareous sediments in medium to thick

beds that crop out in the axial portion ofthe Valle de Guadalupe anticline. A thin-ner section of somewhat similar rocks, withbentonites and of the same age, was pen-etrated by the Agua Nueva No. 1 well(Carrillo Bravo, 1971).

Guaxcamá Formation. A thick LowerCretaceous sequence of evaporites,mostly anhydrite and gypsum, but with asubstantial content of limestone and frac-tured dolomite beds, composes the Guax-camá Formation. Some red terrigenoussediments and silicified conglomerate havebeen found, but no salt has been recorded.The Guaxcamá crops out in an area approxi-mately 35 by 10 km in the central portionof the Santa Domingo and Guaxcamá anti-clines near the western margin of the Plat-form. It appears to concordantly underliethe El Abra Formation over much of theinterior of the Platform, but its true thick-ness is difficult to determine because ofplastic deformation. In the Agua Nueva No.1 well, 1987 m was penetrated, and 3009m (not including 500 m eroded from thetop) was cut by the Guaxcamá No. 1(Carrillo Bravo, 1971). These sections areprobably exaggerated by folding. The ge-ology of the Platform and small sections ofbackreef and reef facies cut in the Guax-camá No. 1 suggest that a massive LowerCretaceous reef developed as shown in Fig-ure 2.2, forming a shallow lagoon forevaporite deposition. The exact structure,distribution, and thickness of the LowerCretaceous reef trend is not yet known.Anhydrite with similar stratigraphic rela-tionships has been found in wells in thecentral portion of the middle CretaceousGolden Lane Platform (Boyd, 1963; seeespecially his Figure 2). The distributionof the evaporite facies is not well knownon either platform.

El Abra Formation. Investigation of themiddle and Upper Cretaceous limestone ofnortheastern Mexico has had a stormyhistory concerning almost every possibleaspect, including nomenclature. The termi-nology given here is that of Carrillo Bravo(1971) in his recent regional synthesis. Thename El Abra Formation is applied to thecalcareous complex principally of Albian-Cenomanian age that is found over theValles–SLP and Golden Lane Platforms.Previously, the name El Doctor Formationor Group was applied to the middle Creta-ceous limestones west of Valles, and thename El Abra Formation was restricted tothe reef and adjacent backreef facies of theSierra de El Abra and Golden Lane. The ElAbra formation is now subdivided into

three contemporaneous gross facies (seeFigures 2.2 and 2.3): forereef facies (pre-viously the Tamabra Formation or member),reef facies (Taninul facies or member), andbackreef and platform facies (including theEl Abra facies and El Doctor limestone ofolder classifications).

The forereef facies is a “mixed” rock unita few hundred meters thick found aroundthe margin of the Platform and is transi-tional to the basin facies (Cuesta del CuraFm.). It consists of porous dolomitizedcalcarenites and breccias derived from thereef and dark lime mudstone and wacke-stone with chert lenses typical of the basinlimestone. Relative proportions of thelithologies vary around the region and withdistance from the reef.

The “reef” facies (Taninul facies) is notstrictly a reef analogous to modern coralreefs. Bonet (1952, 1963) has extensivelydescribed this facies in the Sierra de El Abraand compared it with various modern car-bonate models. It is composed of muddyrudist banks (tabular bioherms), togetherwith the volumetrically much more abun-dant interreef bioclastic debris and mud.The rudist reefs are tabular in form, mea-suring a few hundred meters long and a fewtens of meters thick. They are composed ofabundant whole rudists, many in growthposition, weakly bound by stromatoporidsand coralline algae in a matrix of lime mudand skeletal sand. These reefs tended toform complexes of individual reefs and togrow along the margin of the Platform. Theinterreef and immediate backreef areas werefilled with reef debris, forming massive,coarse skeletal grainstones and packstonesto skeletal siltite. The skeletal sand and reeffacies display very little bedding and gradealmost imperceptibly into each other. Ac-cording to Griffith, et al. (1969), the com-plex Taninul facies can be divided into justthree subfacies: skeletal siltite, skeletalsand, and organic reef facies. They formwhat has been called a “calcareous bank.”Sparry calcite cement has eliminated allporosity except some cavities left by fos-sils and a system of massive fractures.Carrillo Bravo (1971) suggests the totalthickness of the reef may exceed 2000 m,but 800+ m is apparently the largest out-crop. In the Sierra de El Abra, the width ofthe reef zone is usually 600 m or less, butabout 2 or 3 times this width is known atGómez Farías in the Sierra de Guatemala.

Over the interior of the Platform, a widevariety of lithofacies make up the “back-reef” or platform facies of the El Abra For-mation. Carrillo Bravo (1971) outlines the

25


following gross distribution and types (Fig-ure 2.3): (1) backreef clastic unit border-ing the reef zone, (2) unit of calcilutites andcalcarenites with miliolids and toucasiabiostromes in a belt encircled by the clas-tic zone, (3) unit of fine to medium graineddolomite and partially dolomitized lime-stone, with some beds similar to lithofacies2, (4) unit of calcilutites and calcareniteswith miliolids and toucasias and of spo-radic dolomite horizons, which covers morethan 15,000 km2 in the central portion ofthe Platform, and (5) basal dolomite zone,in thick beds, apparently covering the en-tire Platform at the base of the El Abra For-mation. Types 1–4 occur in a bull’s-eyepattern overlying type 5. The total thick-ness of the backreef facies is 1000 to 1500 min the central region, including about 100 mof the basal dolomite, and up to 2000 mnearer the margins. The Tamalihuale No. 1well, located about 4 km west of the reefand southeast of Cd. Valles on the Sierrade El Abra, penetrated about 1450 m of litho-facies 2 above and 350 m of basal dolomite,which was described as having frequentsmall and irregularly distributed caverns.On the outcrop at the south El Abra pass,the backreef facies is predominately lightcream to gray lime mudstone and wacke-stone with some asphaltic stains. Exposuresurfaces marked by thin dolomite layers,flat-pebble conglomerate, and dessicationstructures are present. Some beds containabundant stylolites. Bedding varies from0.5 to 4 meters, with a few prominent shalepartings up to 4 cm thick (Plates 2.1 and2.2). Near the reef, the bedding is indis-tinct and massive, and the shale partingsand the thin dolomite layers are absent; allthese features become more abundant withincreasing distance from the reef. Numer-ous small solution openings are found inthe pass, but the rocks are very compact.Rose (1963) reports that only one of thethirty outcrop samples from the SouthernEl Abra pass had a measurable permeabil-ity (0.6 millidarcy), and the average poros-ity was 2.4% with extremes of 6.7% to 0.3%(measurements by Shell Oil).

Most of the upper surface of the El AbraFormation on the Platform is an uncon-formity. In the Valles area, the Agua Nueva,San Felipe and Méndez Formations directlyoverlie the El Abra at different places(Heim, 1940). The top of the El Abra lime-stone is now exposed and eroded in theSierra de El Abra, hence it is not directlyproven that it was ever covered. Since thishas important implications on developmentof permeability and karstification, evidence

bearing on this problem will be analyzedin a later section.

Agua Nueva Formation (Xilitla Flags).The Agua Nueva consists principally of fineto medium grained black argillaceousflaggy limestone in thin to medium bedsalternating with brown to black shale. Abasal conglomerate is often present. On thewestern flank of the Sierra de El Abra nearLos Sabinos, the Agua Nueva is not so“dirty” and appears as a thin band of whitelimestone circling hills (W. Russell, per-sonal communication; caution must be ex-ercised because some beds in the El Abralimestone also show up as “white bands”).Its thickness varies from 0 to 30 m in theSierra de El Abra and is about 50 m in theSierra la Colmena. It is also found on thesoutheastern and northern margins of thePlatform in thin units and in wells through-out the coastal plain.

Soyatal Formation. The Soyatal Forma-tion was deposited over the western mar-gin of the Platform and over a broad areain the deeper water basin to the south andwest. It is quite variable in thickness, rang-ing up to about 1000 m. It is transitionalfrom thin to medium bedded dark graylimestones with thin, often reddish shalepartings and chert at the base to a middleand upper section consisting predominantlyof gray shales with interbeds of sandstoneand shaley limestone. The Soyatal isthought to be equivalent to the Agua Nueva.

San Felipe Formation. The San FelipeFormation on the eastern margin of the Plat-form is a thin Upper Cretaceous body offine grained argillaceous limestone, withintercalations of olive gray shale and greenbentonite. This thin-bedded finely jointedunit concordantly overlies the Agua Nuevaand in some places discordantly overliesthe El Abra Formation. It is found in a nar-row belt along the base of the eastern scarpof the Sierra de El Abra, in the synclincalvalley and anticlincal Sierra Tamalave tothe west (4 to 50 m), in the Xilitla andGómez Farías areas, and under the coastalplain (several hundred m). At least threesmall patches of San Felipe occur on topof the El Abra southwest of Quintero, andnumerous tiny patches cap hills southeastof Los Sabinos on the dissected westernflank of the range. It weathers to shades ofyellow and brown.

Tamasopo Formation. The TamasopoFormation of Turonian-Senonian age des-ignates about 200 m of reef and backreeflimestones that cover (or covered) a largeportion of the Platform (see reef trend inFigure 2.4). These rocks are similar to the

El Abra Formation and represent the con-tinuation of shallow water conditions in thecentral region as the margins subsided andreceived terrigenous sediments.

Méndez Formation. The Méndez con-sists of monotonous gray, greenish gray,and brown calcareous shale and marl withsome sandy beds and frequent thin beds ofwhite bentonite. Because it is easily eroded,it is now found only in synclinal valleysalong the east front of the Sierra MadreOriental and on the coastal plain east of theEl Abra range. Up to 300 m remain in theValles valley and over 1000 m southeast ofAquismón. It is Campanian-Maestrichtianin age and conformably overlies the SanFelipe.

Cárdenas Formation. A suite of calcare-ous shale, siltstone, and sandstone, sometimesexceeding 1000 m in thickness, covers thecentral portion of the Platform. It correlateswith the Méndez Formation and generallyunconformably overlies the Tamasopo. Thetop of the Cárdenas is usually eroded andoften covered with alluvial material andsometimes by volcanic flows.

Tertiary-Quaternary

Santo Domingo Formation. The SantoDomingo Formation (Carrillo Bravo, 1971)designates a thin unit of fluvial andlaucustrine terrigenous sediments ofPliocene-Pleistocene (?) age. Fragments ofCretaceous limestone, chert, and basaltoccur in conglomeratic horizons. The for-mation is found in the San Bartolo–Cd.Maiz–El Huizache region. In many placesit is covered by basalt flows.

La Borreguita Formation. The LaBorreguita Formation (Carrillo Bravo,1971) is proposed for up to 30 m of finelybanded gypsum of lacustrine origin thatprobably correlates with the Santo DomingoFormation. It occurs in the broad Río Verdevalley and was likely washed down fromhigher exposures of the Guaxcamá to thewest. Another important occurrence of gyp-sum, not mentioned by Carrillo Bravo andof undetermined age and thickness, lies inthe valley from Matehuala to Huizache. Theupper 10 m is weathered, but bedding isbeautifully exposed in numerous caves de-veloped in the gypsum.

Alluvium. In the west and central areas,recent fluvial sediments are found in manyvalleys. Alluvium is relatively thin andpatchy in the eastern part of the Platform.Immediately east of the Sierra de El Abra,there are isolated occurrences of lacus-trine (?) limestone (capping a hill south of

26


Quintero) and calichefied conglomerates(about 2 m capping hills of Méndez shaleeast of the Choy spring).

Igneous Rocks

Intrusive igneous rocks are rare on thePlatform. Some were emplaced during theearly Tertiary orogeny.

Extrusive igneous rocks are of localimportance in the region. Tertiary rhyoliteis found in the southwestern portion of thePlatform and extensively in the centralMexican plateau. Numerous small basaltflows are found in the central, southern, andeastern zones. Many of the flows were “ba-salt rivers” in river valleys. In some places,they overlie lacustrine deposits containingPliocene-Pleistocene pollen (CarrilloBravo, personal communication). East ofthe Platform between Cd. Mante and Cd.Victoria, there are large basalt capped me-sas between 400 and 500 m elevation.

2.4. Structure and Physiography

During the late Cretaceous, tectonicmovements began to occur in the region,

as shown by the shifting sedimentary fa-cies. The Laramide movements culminatedduring Paleocene and Eocene times withuplift and intense folding. No marine Ter-tiary rocks are known on the Platform—the depocenter (ancient Gulf shoreline)moved eastward with time as the regionemerged.

Nearly all the large structural featuresof the Platform are folds having as muchas 2 km of structural relief. The structuraltrend varies from NW-SE to N-S and is dis-cordant with the preMesozoic fold trend(Carrillo Bravo, 1971). Four broad archesand two large anticlines, all with numer-ous associated folds, are shown in Figure2.5. In a belt of intense folding in the east-ern half of the Platform, the anticlines aregenerally narrower and asymmetric or over-turned to the east. Carrillo Bravo (1971)believes that rocks principally from UpperJurassic to Upper Cretaceous are involved,and that the structures were formed by asubsiding basement to the east causinggravitational gliding of rigid strata overLower Cretaceous evaporites. He believesthat many of the folds are injection folds.Along the eastern margins of the Platform

in the Valles area, the folds have less reliefand are separated by broad synclinal valleys.

Faults seem to be of lesser importancethan folds on the Platform. Many of theanticlines have longitudinal faults alongtheir eastern side and may have tear faultsat their terminations, as well as other trans-verse shearing within the structure. TheXilitla overthrust, which has many kilome-ters of displacement, is the largest faultstructure in the area. Several other over-thrusts have been described by Russell(1968) in the adjacent Aquismón area. Un-doubtedly, there are countless small faults,but mapping has not been carried out inthis detail.

Intense deformation of the rigid carbon-ate rock has produced massive fractures.Joints in the El Abra Formation normallypass through several beds; fractures thatextend vertically for over 100 m and some-times much more without observable off-set are known. Later, it will be shown thatthese fractures are very important to thecirculation of ground water.

The uplift and subsequent erosion of allof central Mexico has created three mainphysiographic provinces in the region: the

Figure 2.4. Valles–San Luis Potosí Platform limits duringthe Cretaceous.

Figure 2.5. Structural map of the Valles–San Luis PotosíPlatform.

27


coastal plain of terrigenous rocks extend-ing 110 km from Tampico to the Sierra deEl Abra and at less than 120 m elevation,the Sierra Madre Oriental foldbelt of car-bonate rocks, having high local relief andelevations mostly between 400 and 2500m, but reaching over 3000 m in places, andthe Central Mexican Plateau or highlandsof sedimentary and volcanic rocks, havingmore moderate local relief and elevationsranging from 1500 to 3000 m (Figure 1.1shows the general regional physiography).Another large regional feature is a broaddownwarped zone west of the main fold beltthat stretches from Río Verde to Tula (andbeyond), where there is now a broad valleyor system of wide intermontane valleys atabout 1000 m elevation. They contain al-luvial, lacustrine, and bolson deposits andhave been leveled by sedimentary pro-cesses. Erosional processes began to workon the Platform probably in the early Ter-tiary, but it must have taken a long time toestablish a regional fluvial system. Thesouthern half of the Platform is drained bythe Río Tampaón, which now extends tothe city of San Luis Potosí. Most of thenortheastern quadrant is drained by the RíoGuayalejo. The entire northwestern quad-rant lacks a fluvial network. The stratigraphy

of Tertiary continental rocks and late Ter-tiary-Quaternary erosion terraces suggestthat there has been intermittent uplift sincethe early Tertiary.

Although most of the Platform is now akarst area, the largest scale physiographicfeatures within it were formed by the flu-vial erosion of less resistant Upper Creta-ceous shales and thin bedded limestonesfrom the tops and sides of structural highsof the El Abra and Tamasopo formations.Hence, these features are the result of struc-tural relief and erosional processes. An ex-cellent example is the Sierra de Guatemala,which has over 2 kilometers of structuraland topographic relief. Another example isa large anticline southwest of Ayutla thathas about 1 km of structural relief and awidth of several kilometers. Previouslyoverlying shales have been stripped off thetop and about three fourths of the way downthe flanks; shale remains on the lowerflanks and in the valley. The crestal lime-stones having had the longest exposure arethe most intensively karstified. Most of thepositive structures are anticlines; thereforeridges of varying complexity are the domi-nant positive form, except in the Xilitla area,where a thrust block has formed a karst pla-teau. West of the Sierra la Colmena, most

of the shale has been removed, exposing abroad belt of folded limestone to karst pro-cesses. The largest purely erosional featuresare the transverse valleys or water gapscrossing the structures, which were formedby the major drainage channels. Some ofthe water-gap canyons are 1000 m deep.

The Sierra de El Abra is a long ridge ofresistant El Abra limestone at the easterncentral margin of the Platform. The rangeis much lower in elevation than the nearbyhigh sierras. Its geomorphology and struc-ture will be examined in Chapter 3. A broadsynclinal valley floored with the Méndezand San Felipe formations on the westernside of the range is referred to as the Vallesvalley, or sometimes in its northern end asthe Antiguo Morelos valley. This valleyconnects directly with the coastal plainaround the southern end of the range. Twosmall domes which expose the El Abralimestone occur along the slightly curvingstrike of the El Abra range structure beyondits southern limits. One is called theTantobal dome, and the other is called theSalsipuedes or Coy dome, the latter hav-ing a very large spring. There are manysmaller folds in the El Abra area that oftensignificantly affect local geomorphic fea-tures and drainage.

Table 2.2Mean monthly and annual air termperatures (°C)

ElevationStation4 (m) Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Dec. Avg. Range

Cd. Valles 95 20.0 21.5 23.8 26.8 29.1 28.8 28.5 29.5 28.0 25.1 22.3 20.2 25.3 9.5

Antiguo Morelos 220 18.8 21.3 24.4 27.5 29.2 29.3 283 28.8 27.5 25.0 22.4 20.1 25.2 10.5

Ahualulco4 (80) 19.5 21.2 24.2 26.7 28.7 29.2 28.4 29.1 27.5 25.0 21.2 19.1 25.0 10.1

Joya de Salas2,4 1550 9.2 11.4 16.6 17.4 19.0 18.4 17.9 18.7 17.4 14.1 13.1 11.7 15.4 (9.8)

Tamapatz3 (750) 16.1 16.1 19.5 22.1 23.3 23.5 22.4 23.3 22.9 21.5 18.3 17.9 20.6 7.4

Agua Buena4 372 18.1 19.7 23.0 25.9 27.9 28.0 27.1 27.5 26.2 23.9 20.5 18.4 23.9 9.9

El Salto (400) 16.8 19.0 21.7 24.5 26.4 26.2 25.0 25.7 24.7 23.0 19.1 18.0 22.5 9.6

Jaumave 750 15.3 17.3 19.9 22.4 24.8 25.3 24.6 24.8 23.8 21.5 17.8 16.1 21.1 10.0

Lagunillas4 950 17.6 19.4 22.4 24.2 26.1 25.9 24.7 25.5 23.8 21.7 19.9 17.4 22.4 8.7

Cd. del Maíz 1239 15.6 17.5 22.0 22.3 24.8 24.3 23.1 23.2 22.5 20.6 18.0 16.2 20.8 9.2

Río Verde 987 16.0 18.0 20.5 23.1 24.9 24.7 23.7 23.9 22.8 20.7 17.9 16.0 21.0 8.9

San Luis Potosí 1877 13.5 15.4 17.8 20.2 21.4 21.0 19.8 20.0 18.8 17.3 15.2 13.9 17.9 7.9

1. lowest and highest monthly average are underlined; data are from S.R.H. Boletín 19, except Tamapatz (unpublished)

2. only one year of data

3. five years of data

4. see Figure 1.1 for locations except for Ahualulco, 11.5 km S of Gómez Farías, Joya de Salas, 20 km NW of Gómez Farías,Agua Buena, 3 km N of Tamasopo, and Lagunillas, 30 km S of Rayón

28


2.5. The Environment

2.5.1. Climate and vegetation

The climate of the region varies greatly,both temporally and spatially. In terms ofprecipitation, the year is normally dividedinto wet and dry seasons. During themonths November through May, there isvery little precipitation in any part of theregion; most of the rainfall occurs duringthe months June through October. Mois-ture is brought into the region by warmsummer winds from the Gulf of Mexico.Occasionally there are hurricanes. Themean annual rainfall may exceed 2500 mmin the wettest localities along the easternflanks of the high sierras, such as the Xilitlaarea, and is as low as 300 mm in the semi-arid western half of the region, which is inthe rain shadow of the Sierra Madre Ori-ental (Figure 1.1). The Sierra de El Abrareceives about 1200 to 1400 mm per year.A more detailed discussion of precipitationin the region is given in Chapter 4.

Seasonal temperatures are related toboth solar cycles and precipitation. Thecoldest months are December and January,early in the dry season (Table 2.2). Gradualwarming occurs through the spring to thehot, dry days of April and May, when thedaily maximum occasionally exceeds 40°Cin the Valles area (Table 2.3); the vegeta-tion on the Sierra de El Abra usually ap-pears to be a parched brown at this time.Usually the rains begin modestly in Mayand come heavily sometime in June. Dur-ing the rainy season, the relative humidityincreases, and the daily maxima and meanmonthly temperatures are suppressed. In thelast half of the dry season, the potential forevapotranspiration is great, but there is verylittle moisture. However, once the rainsbegin, the actual evapotranspiration mustbecome quite high. Table 2.3 shows thedaily maximum and minimum temperaturesand their average differences for four se-lected months in 1971; there is a large dailyvariation in the dry season, and somewhatless in the wet season. The temperature datamay be compared with daily rainfall for1971 shown in Figure 4.7; very high dailytemperature maxima continued until therains came in mid-June.

Table 2.2 gives an indication of the re-gional and altitudinal variation in tempera-ture, as well as seasonal changes. The dataare arranged from easternmost at the top towesternmost at the bottom. Nearly all thestations are located in valleys; hence theydo not accurately reflect the full range of

Table 2.3Daily maximum and minimum temperatures

at Estación Santa Rosa (Cd. Valles) in degrees C

1971 January May June October

Day Max Min Max Min Max Min Max Min

1 28.0 19.0 30.5 21.5 38.0 27.0 33.0 21.5

2 30.0 16.0 35.0 18.5 38.0 27.0 32.5 22.5

3 34.5 14.5 36.0 17.5 38.0 25.0 32.5 23.5

4 23.5 15.0 37.0 22.5 37.5 25.5 32.0 23.0

5 15.0 11.5 40.0 24.0 37.0 22.5 32.5 23.0

6 13.0 11.5 46.0 22.0 37.0 25.0 29.0 23.0

7 17.0 9.0 43.5 22.5 37.0 24.0 31.0 23.0

8 18.0 8.0 38.5 24.5 38.0 26.0 33.5 22.0

9 25.0 8.0 36.5 26.0 39.5 25.5 34.0 22.0

10 29.0 9.5 37.0 23.0 38.0 26.5 27.5 22.0

11 30.0 14.0 39.0 20.0 39.0 26.0 29.0 20.5

12 29.0 16.0 32.0 23.5 30.5 23.0 29.0 18.5

13 30.0 18.0 29.0 22.5 35.0 24.0 29.0 20.5

14 28.0 19.5 31.0 17.0 36.0 24.0 30.0 21.5

15 30.0 20.0 37.5 19.0 35.5 23.0 32.5 22.5

16 23.0 20.0 38.5 19.0 34.5 23.0 32.5 22.0

17 32.0 15.0 37.5 18.5 30.0 24.0 33.5 24.0

18 34.0 15.0 37.0 19.0 32.0 22.5 33.0 23.5

19 23.0 19.0 37.0 23.0 31.0 23.5 33.5 24.0

20 22.5 14.0 37.5 25.0 31.0 23.0 32.5 22.5

21 29.5 12.5 37.5 22.0 23.0 22.0 30.5 20.0

22 29.5 12.5 38.0 23.0 30.0 22.0 32.0 19.5

23 30.0 12.0 36.5 26.5 31.0 20.0 31.5 18.0

24 30.5 13.5 39.5 25.0 30.5 22.0 30.5 21.5

25 29.5 13.5 31.0 22.0 28.0 23.0 32.0 20.5

26 28.5 17.0 34.0 23.5 31.5 22.5 32.5 20.0

27 26.0 18.0 36.0 23.0 32.0 22.5 32.5 20.0

28 30.0 20.0 36.0 23.0 34.0 23.5 33.0 21.5

29 30.0 18.5 36.0 24.0 32.5 22.0 32.0 19.0

30 32.5 16.5 38.0 24.5 32.0 21.5 33.0 20.0

31 35.0 14.0 37.5 26.5 — — 32.5 19.5

Avg. 27.3 14.9 36.6 22.3 33.9 23.7 31.7 21.4

Difference 12.4 14.3 10.2 10.3

Month mean 21.1 29.3 28.8 26.6

29


temperatures to be expected in the region.The temperature difference betweenAhualulco and Joya de Salas (at the east-ern base and high on the western flank ofthe Sierra de Guatemala, respectively) in-dicate the much cooler conditions that un-doubtedly exist in the high sierras. Snowoccurs only rarely in the Valles area, butlight snows are likely more common in thehigh ranges. Winter frost at high altitudesand dew at all altitudes are common events.

The variety of vegetation in the regionis of course very great in order to accom-modate the extremes of climate and groundconditions. The Sierra de El Abra is cov-ered with a dense thorn forest in which allthe plants must be adapted to a long dryseason with sparse water renewal and verylittle soil. There are many types of cactus,stinging vines and nettles, epiphytes, andthorny bushes and trees. Except in rela-tively rare places where there is a signifi-cant soil accumulation, trees usually are notmore than about 10 m high. On the coastalplain and in the low Valles valley, the natu-ral vegetation is a scrubby bush with lowtrees, except where there are stream chan-nels, especially those that are spring fed.In the much higher and wetter Sierra de

Guatemala, P. S. Martin (1958) has distin-guished several vegetation zones; amongthem are tropical deciduous forest, tropi-cal semi-evergreen forest, cloud forest, andpine and oak forest. Going farther west intothe semiarid region, there is chaparral andthorn desert.

2.5.2. Culture

Despite the ruggedness of the terrain,the region as a whole is moderately wellpopulated. Probably the majority of thepopulation lives in towns in valleys wherethere are water supplies, road constructionis more tractable, and ranching and farm-ing are more fruitful. Most of the largertowns and more productive ranches orfarms are located along the eastern marginof the region and on the adjacent coastalplain. The principal products are cattle,oranges, sugar cane, and sisal, but a greatvariety of other fruits, vegetables, andgrains are also grown; there is a smallamount of irrigation. On the lower eleva-tions in the front ranges of the Xilitla area,coffee, oranges, and corn are importantcrops. Much of the mountainous countryis still poorly serviced by roads. Industrious

natives have carved out small plots of landon slopes as great as 45° by slash-and-burntechniques; they grow corn, coffee, andbeans, which are carried on their backs tomarket. The high parts of the ranges aresparsely inhabited, there often being onlya few woodcutters and isolated farmers.Farther west in the semiarid zones, the oc-currence of water and arable land are themain factors in human occupation.

Access to the Sierra de El Abra is gen-erally poor, and very little was known aboutthe top of the range previous to this studyand other recent investigations by the As-sociation for Mexican Cave Studies. Threeroads cross the range transversely, two ofthem in wind gaps. There are no roadslengthwise along the crest, only a highwayparallel to the range several kilometers westof it and a dirt track parallel to the east facethat is impassable during much of the wetseason. There are numerous trails on thecrest that are used by woodcutters, a fewfarmers, and occasionally by jaguar, oce-lot, or deer hunters. In general, the crest isalmost totally uninhabited, and navigationoften could only be made by air photo andmachete.

30


3

GEOMORPHIC FEATURES OF THE SURFACE

The surface geomorphic features of the Si-erra de El Abra and environs are describedin this chapter. This is the area in whichmost of the field work was conducted, andwhich was studied by aerial photographs.Karst landforms in other parts of the re-gion will be briefly considered at the endof the chapter.

The region is very large, access is poor,and the geomorphic studies only just be-gun.

3.1. The Sierra de El Abra

The Sierra de El Abra is an extremelyelongate range at the eastern edge of theValles–San Luis Potosi Platform. It beginsin the south at the Río Tampaón and ex-tends approximately 140 km northward toPuerto Chamalito (a wind gap), which istaken as the northern limit of the range (Fig-ure 3.1). The Sierra de El Abra structureactually can be traced for 6 to 7 km northof Puerto Chamalito in a zone where itgradually merges with the much larger Si-erra de Guatemala. The main range variesin width from a minimum of 2 km to amaximum of 5 or 6 km. The crest zone ofthe range (Plate 3.1) has very low relief.The crest rapidly descends southward fromthe Sierra de Guatemala to a low of 190 to200 m just north of the Río Comandante,then gradually rises to the high central por-tion of the range, and finally descendsslowly to where the El Abra structure passesunder the Río Tampaón at 28 m. The higherpart of the range (that part 450 m andabove) extends from about the latitude ofAntiguo Morelos to a few kilometers northof Cd. Valles. Recent explorations in Cuevade Diamante (Peter Sprouse, personal com-munication) show that the crest probablyreaches at least 700 m in the highest areas,much higher than the 450 m shown on theold 1:100,000 topographic maps.

The well defined eastern margin is asteep scarp that descends to the coastalplain varying from 33 to 120 (?) m eleva-tion. The western flank of the range is muchmore complex. In many places, the El Abralimestone is structurally higher than the

Méndez and San Felipe Formations in theValles and Antiguo Morelos valleys. Thus,the western flank is quite variable in width,and the El Abra limestone is exposed bothin anticlinal ridges and synclinal valleys atelevations intermediate between the crestof the main range and the lower portionsof the Valles valley. The width of outcropof the El Abra limestone reaches about 10km in places. As pointed out by Mitchell,et al. (1977), the name Sierra de El Abra isoften applied to the complete exposure ofthe El Abra Formation, though technicallyit should not be. Also, it should be notedthat many other names are locally used forthe El Abra range, including Sierra Cu-charas, Sierra de Tanchipa, and Sierra delas Palmas.

Based on air photos and some field ob-servations, a geomorphic map of the south-ern part of the Sierra de El Abra is given asFigure 3.2. Most of the field work on theEl Abra range was conducted in this area.It will be helpful to refer to it in the fol-lowing sections and in the chapter on caves.Aside from the drainage pattern and othergeomorphic features, the map includes oneboundary of particular importance: the con-tact between the El Abra limestone and theoverlying formations. On the west side ofthe range, the contact can usually be accu-rately determined, but along the east faceit is drawn approximately at the base of thescarp. On the air photos, there is consider-able scale distortion where there is localrelief. The map was made simply from anair photo mosaic, using sightings along thehighways as the control. The distortion hasbeen corrected somewhat by the overlap-ping process, but there is undoubtedly stillsome error. The scale was determined byfield measurements and by comparison with1:50,000 topographic maps.

3.1.1. The east face and the coastal plain

The 110 km wide coastal plain is a lowrelief erosion surface developed on weaklyresistant Tertiary and Upper Cretaceousterrigenous rocks and sediments. In thewestern part of the plain, the surface becomes

more irregular, and near the Sierra de ElAbra, it is being rapidly eroded downwardby streams that head on or at the base ofthe scarp. West of Tamuín near the Choyspring, there are several flat-topped hillsthat are remnants of an older erosion sur-face at an elevation of about 77 m (Plates3.2 and 3.3). These hills have a cover of2 m of caliche, containing abundantrounded gravel, that rests on Méndez shale.The gravel cap has been removed from mostof the area, and the dissection by base levelstreams has reached down to elevations of33 m at the origin of the Río Choy, and 28 mwhere the Río Tampaón passes the south-ern end of the range. East of the high cen-tral portion of the range, the air photos showthat the surface has considerable relief,perhaps more than 50 m in places. The areabetween the Santa Clara and Tampaónsprings appears to be the most rugged.Some of the interfluves, e.g., on RanchoTampacuala northwest of Tamuín, also sug-gest an older erosion surface, although notnecessarily the one cited at El Choy. In thevicinity of Cd. Mante, the surface has lowrelief and rises to meet the scarp at 100to130 m elevation. There is only one re-sidual hill at 220 m elevation to suggestolder erosion levels in that area (Plate 3.1).

The topographic character of the eastface may be seen in Plate 3.4. The upperhalf of the scarp is very steep, locally nearlyvertical, and composed of the El Abra reeffacies. The surface of the coastal plain, de-veloped on the Méndez, becomes steeperand more dissected adjacent to the scarp.From a somewhat arbitrary “foot” of thescarp, the slope steepens and may be com-posed variously of the Méndez, San Felipe,or El Abra Formations, or of talus. In fourplaces where limited investigations weremade, the bedrock of the lower portion ofthe scarp was almost completely coveredwith boulders of El Abra limestone that hadfallen from above, or by soil and thickvegetation. At the Nacimiento del ArroyoSeco on Rancho Peñon near the middle(north-south) of the range, the El Abra lime-stone is exposed by the spring at the baseof the scarp, but Méndez shale crops out a

31

AMCS Bulletin 14 — Chapter3

few tens of meters downstream and inpatches a hundred meters to the south onthe lower slope of the scarp, where it is wellabove the spring level. The eastward dipof flaggy beds in the Méndez indicates thatfolding may be of considerable importance.Study of the air photographs reveals eastdipping beds (either San Felipe or flaggyMéndez beds) and probable small anticlinesin several localities along the east face. Itappears that impermeable rocks normallyextend some distance up the scarp, perhapsas much as 100 m or more in the centralportion of the range. South of the SantaClara spring, there are topographic featuresnear the base of the scarp that appear (onair photos) to be small east-dipping hog-backs, formed where streams have cutacross and behind resistant beds in theMéndez or San Felipe formations. Thebreak in slope that often occurs about half-way up the scarp might be caused by a fa-cies change within the El Abra limestoneor by the change from Méndez, San Felipe,or talus to barren limestone. Based on ob-servation and a few measurements, thescarp has an average slope of around 30°to 35°, a minimum of about 25°, and slopes

as great as 45° in the vicinity of springs.Many springs discharge from the base

of the range (Plate 3.5). Most of them aresmall. The two largest springs are the Choyand the Mante, at the opposite ends of therange. At most of the spring sites observedin this study, the discharged water risesvertically from some undetermined depthinto a “rise pool.” One exception is theArroyo Seco spring (on Rancho Peñon),which has a horizontal “water table” pas-sage for at least the first few tens of meters.The flow from the springs cuts more deeplyand laterally into the coastal plain at thescarp than the nearby surface streams, strip-ping away the impermeable rocks back tothe El Abra limestone. Thus, a sort of re-entrant is formed in the Méndez, but nosignificant steepheads or pocket valleysin the El Abra limestone are known. Will-iam H. Russell (personal communication,unpublished manuscript) and Mitchell,Russell, and Elliot (1977) have suggestedthat the scarp has (sometime in the past)been etched backward in a number of placesby swamps to form a “corrosion plain.”They cite as an example the Nacimientodel Río Santa Clara, where the “solutional

encroachment has progressed about one-half kilometer” (Mitchell et al., 1977, p.13), and Russell (personal communication)states that the Santa Clara spring now liesabout 500 m east of the base of the scarp.These statements must be incorrect becausethe erosional morphology and other char-acteristics shown on the air photos of thearea near the spring indicate the bedrock isthe Méndez shale (or San Felipe) ratherthan the El Abra limestone, and the springoccurs essentially at the base of the scarp(within 50 m). From the field observationsand air photo analysis made during thisstudy, it is concluded that solutional en-croachment has not steepened the scarp,and probably has never had any significanteffect on the El Abra range. This conclu-sion is supported by the lack of “swampslots” or “solution notches” commonlyfound where swamp waters impinge later-ally on a limestone surface (Wilford, 1964;Jennings, 1971). Lateral solution encroach-ment as suggested by Russell (personalcommunication) may be important nearMicos in the valley west of the Sierra LaColmena.

Common features of the upper part of

Plate 3.1. Gulf coastal plain and Sierra de El Abra. This aerial view looks southward along the strike of the rangefrom above the town of Quintero. The coastal plain of shale rises gently to the eastern reef-front escarpmentof the range. The crest typically has low relief, and the western margin dips westward to the Valles–AntiguoMorelos valley. Note the residual hill of Méndez shale which is capped by a lacustrine limestone.

32


the scarp are fracture controlled recesses,as shown in Plate 3.6. Usually they are ori-ented vertically, but a few run diagonallydown the slope. Some of the larger recessesare found where caves developed alongfractures are collapsing. It is believed thatmost of the recesses are formed by en-hanced solution and rockfall along frac-tures. Perhaps the majority of the talus onthe lower slope is derived from retreat ofthe fractured rock. However, in some placesgullies appear to reach high up the scarp,thus sometimes confusing the issue. Also,there are several clear examples of what areherein called “spillways” or “chutes” wherestreams have flowed off the top of the range,spilling down the face. Two prominent onesnorth of the Choy spring are indicated onthe geomorphic map (Figure 3.2). A few ofthese chutes appear to be supplied withwater from collapse dolines at the edge ofthe crest, and thus are still active.

3.1.2. The crest of the range

The top of the El Abra range has a rela-tively flat surface (Plates 3.1 and 3.4),rather than a rounded or sharply crestedridge. As indicated on the maps (Figures3.1 and 3.2), it maintains a remarkably simi-lar width of 2.5 to 3 km throughout mostof its length of 140 km. The crest strata havelow dips, and an anticline or west-dippingmonocline marks the western boundary ofthe crest. The present surface is developedentirely on El Abra limestone, except for afew small patches of San Felipe just northof the North El Abra Pass (Figure 3.1).Observations in the southern part of the ElAbra range show that the surface locallyfollows closely the structural attitude of thebedded limestone. For example, the anti-cline at Cueva Pinta and west of Monos(see Figure 3.2) has a moderately dipping(15° to 20°) western limb and a gently dip-ping eastern limb; the ground surface quiteaccurately reflects the structure. In otherplaces, one may walk for some distance onthe same bed of limestone. Thus, the top ofthe El Abra range is a structural surface, toa first approximation. However, over thewhole length of the structure, much morelimestone must have been removed fromthe high central zone than from the ends.

In the past, the Sierra de El Abra musthave been covered by a small thickness ofthe San Felipe formation and at least hun-dreds of meters of the Méndez formation(and possibly by lowest Tertiary beds aswell). The relevant facts are: (1) smallpatches of the San Felipe occur on the top Figure 3.1. Sierra de El Abra and surrounding area.

33


Plate 3.2. Flat-topped hills are remnants of an older coastalplain surface. At the right in the background is the presentlow relief erosion surface on Méndez shale.

Plate 3.3. Typical relief on the coastal plain near the ElAbra range. The gravel cap of the flat-topped hills hasbeen removed, and the hills are being dissected.

Plate 3.4 (left). Theeastern escarpment ofthe El Abra range in thehigh central area. Thesteep upper part isdeveloped on El Abrareef facies. Talus, soil,and occasional outcropsof San Felipe orMéndez are found onthe lower slope.

Plate 3.5 (right).Springs rise from thebase of the east face andflow across the coastalplain. Here the Mantespring flows from ashort cave into a smallartificial reservoir.

of the range, near the North El Abra Pass;(2) at its southern end, the El Abra lime-stone passes into the subsurface, beneath agreat thickness of Méndez; (3) J. Carrillo(1971) reports that the Méndez is over1000 m thick southeast of Aquismón andseveral hundred meters have been pen-etrated in coastal plain wells near the ElAbra range; (4) geomorphic evidence con-cerning the paleo-hydrology of El Abrarange caves (to be developed in later sec-tions) requires the coastal plain imperme-able rocks to have formerly reached at leastthe highest present level of the Sierra de ElAbra; (5) the Río Comandante (La Servi-lleta Canyon) and the stream that createdthe now dry North El Abra Pass (between

Mante and Antiguo Morelos) are believedto have been superimposed on the El Abrastructure; and (6) the structural and strati-graphic relations of the formations showthat the folding (and faulting?) took placeafter, or possibly began very late in, thedeposition of the Méndez; Muir (1936) saysthat the tectonism occurred during and afterdeposition of the lower Tertiary Chicon-tepec beds. Hence, at one time, the Sierrade El Abra had a thick impermeable cover.

As the less resistant impermeable coverwas eroded from the top of the range, thelimestone was laid bare to form approxi-mately a stripped structural surface. Thisprocess of course first exposed the lime-stone in the high central part of the range.

As the regional base level has lowered (rela-tive to the structure), the stripping pro-gressed to the north and south along theaxis of the structure and down its flanks aswell, creating the present condition. Vestigesof the paleo-fluvial system that loweredonto the El Abra limestone are indicatedby a network of dry valleys, “spillways” onthe east face, and notches in anticlinalridges along the western edge (Figure 3.2).The dry valleys are generally dendritic, butare structurally controlled in places. Somecan be traced to “spillways” on the east faceor to wind gaps on the western edge withassociated dry valleys on the flanks. Evi-dently, the channels continued to functionfor some time after reaching the limestone,

34


but gradually dried up as the impermeableinterfluves were removed and secondarypermeability developed in the El Abra for-mation. The valleys are most easily tracedin the lower parts of the crest, such as nearthe South El Abra Pass. Only the largestpaleo-fluvial features survive in the highpart of the range. Karst solution has beenactive in the higher areas much longer thanin the lower areas and has undoubtedly re-moved more of the El Abra limestone there;thus, most of the paleofluvial pattern hasbeen obliterated. Figure 3.3 shows themorphology and drainage pattern of theNorth El Abra Pass area. Particularly im-portant here are the patches of the SanFelipe formation, which occur in synclinaldepressions at 240 to 270 m elevation, andthe erosional features caused by runoff fromthese impermeable beds. A much greaterthickness of the San Felipe and Méndez willhave covered the whole area in the past,thus enabling streams to easily erodeheadward across the low anticlinal ridgealong the western margin of the crest, andthen to lower onto the El Abra limestone,where they cut narrow gorges. This pat-tern—an anticlinal ridge on the westernmargin of the crest, breached by a windgap that joins an adjacent synclinal (andsometimes faulted) dry valley on the eastwith a dip-slope valley on the western flankof the range—is common in the Sierra deEl Abra, but in most instances all the im-permeable rocks have been removed.

The present surface of the crest has avariety of small scale solution features. Insome places, the beds seem nearly unbro-ken, and only small solution hollows andchannels have formed. In other places,

Plate 3.6 (left). Fracture-controlled recesses inthe east face. In somecases caves are devel-oped along the fracturesand have partiallycollapsed. This photoshows Cueva de lasCuates in the highcentral part of the range.

Plate 3.7 (right). Thesurface of the crest ofthe El Abra range. Thereare many small vadoseshafts (lapiez wells).Note the vegetal debris,loose blocks, and cactus.

Figure 3.3. Geomorphology of North El Abra Pass area showing effect of SanFelipe beds on the range.

37


joints have been dissolved downward 1 to2 m to form lapies. Loose blocks of lime-stone are common. Much bare rock is ex-posed, and where there is soil, it commonlyconsists of no more than a few centimetersof vegetal matter (Plates 3.7 and 2.1). Somefine sediments and organic material arewashed into open joints and broad depres-sions. A few of these depressions have suffi-cient soil to grow crops, but nearly all of theSierra de El Abra is covered by vegetation(scrubby thorn forest) specifically adaptedto life with little soil, very low moisture,and a strongly seasonal climate.

Large scale karst features on the crestare limited to two types of sinkholes. Noneof the “typical” tropical karst forms suchas haystack hills, mogotes, or cockpits arepresent; nor are there poljes as found inmany karst regions. Study of air photo-graphs shows a considerable number of

Plate 3.8. The Caldera, a giant collapse doline on the crest of the El Abra. It is425 m long, 245 m wide, and 90 m deep. Another, having sloping sides, maybe seen in the background. Soil collects in these dolines and supports adifferent vegetation than the surface.

Figure 3.4. Map of Caldera, a large collapse sinkhole.

38


dolines that are believed to have beenformed by surface solution processes (in-dicated by an S). However, their densityappears to be much less than the dolinefields of Kentucky and Indiana, and theirmorphology is somewhat different becauseof the lack of soil. Their width variesgreatly, ranging up to several hundredmeters, but their depth below closure isusually not more than about 10 m. An at-tempt was made to trace the closure line ofthese dolines on the geomorphic map (Fig-ure 3.2). Most of the larger solution dolinesoccur along the supposed paleo-drainagelines. In some places along these fossildrainage channels dolines deeper than 10 mhave formed, apparently by surface solu-tion. Large lineations several kilometers longappear on the photographs. They indicatemajor transverse or longitudinal fractures,probably faults, in the El Abra limestone.Some of these fractures have been espe-cially favorable sites for surface solutionprocesses, thus forming greatly elongatedsolution dolines and possibly bogaz, andsolution depressions and trenches that arenot closed. In many places, the east face ofthe El Abra gives the impression of havinga double crest. At these places a parallelvalley, typically 30 to 50 m deep, lies a fewhundred meters behind the first ridge, andanother ridge sometimes higher than thefirst, lies along the west side of the valley.This morphologic feature (the valley) prob-ably has been created by some combinationof folding, faulting, paleo-fluvial erosionand enhanced solution along fractures. Oneor two fracture lineations commonly maybe traced (on air photos) in the valley overdistances of several kilometers.

The second type of sinkhole found inthe El Abra is formed by stoping of ceilingrock in underground chambers until theground surface is finally reached. These arecollapse dolines (marked with c on Figure3.2). They range in size from 10 m acrossto spectacular ones more than 300 m acrossand 100 m deep (Plate 3.8 and Figure 3.4).That they are indisputably created by cavecollapse will be shown in Chapter 6. Manyof them are about three times as long asthey are wide, and the elongated directionis usually parallel or perpendicular to theaxis of the range. Others are roughly equi-dimensional in plan view. Many of themhave some nearly vertical or overhangingwalls as in the Caldera, while others havesloping walls that have apparently beenpartially degraded. Virtually all of the largecollapse dolines occur between the NorthEl Abra Pass and the latitude of Los

Sabinos, i.e. the high central portion of therange (above 450 m).

3.1.3. The western flank and the Valles orAntiguo Morelos valley

Just as on the crest, the impermeablecover has been removed from a large partof the western flank of the Sierra de El Abrato expose the El Abra limestone. Betweenthe North and South passes, the belt of lime-stone on the flank usually has a width ofseveral kilometers, reaching a maximum ofabout 6 km. At the ends of the range wherethe structure subsides, the flank is verynarrow. Near the anticlinal western crestdips reach 20° or more to the west, whilethe lower part of the flank has lesser dips,circa 5°. Many small folds occur on theflank, causing greater variation in dip, lo-cal relief, and geomorphology than occurson the crest of the range.

The present and past drainage systemsof the western flank and the Valles–AntiguoMorelos valley are easily traced on air pho-tos (see Figure 3.2). It is clear that thepresent condition is merely an instant in thecontinuing process of erosion of the im-permeable cover from the El Abra lime-stone, followed by the development of akarst. The presently active fluvial systemsare developed on the impermeable rocks.Their valleys usually continue headwardonto the limestone surface, where they areequally well formed, but they are believedto be dry valleys. The fossil streams weregenerally oriented down the dip and theydissected the flank to form valleys often50 m deep (estimate). The local erosionalrelief appears to be greater on the flanksthan on the crest. It must have taken sometime to remove the impermeable rocks fromthe interfluves, but presently the only coverremaining is found on knobs and ridges lowon the flanks where the valleys leave thelimestone. A few of the “dry” valleys havethe appearance (on air photos) of being stillactive. For example, many of them do notseem to have a significant profile disconti-nuity between the parts developed on theEl Abra limestone and on the impermeablecover. More field observations of the char-acter of the dry valleys should be made,particularly during heavy storms, althoughaccess is difficult at those times. The wholelimestone surface is presently being low-ered by karst solution processes, and nu-merous closed valleys have formed. It isbelieved that the fossil surface drainagenetworks on the El Abra limestone havebeen almost totally destroyed, and that

water now enters the subsurface via a greatmultitude of small infiltration points(mostly opened joints, but also beddingplanes).

The small streams on the west flank ofthe Sierra de El Abra are tributary to larger,strike-oriented streams. A drainage divideoccurs 5.5 km north of the state line at anelevation of 290 m (the lowest point). TheArroyo El Lagarto (also called ArroyoTerroncillos) is the principal northernstream, and it flows to the Río Comandante.The Río Puerco (or Arroyo Grande) is theprincipal southern stream, and it flowssouthward along the western side of thevalley to join the Río Valles. From about3 km northeast of Cd. Valles, Highway 85follows for about 13 km the crest of anorth-south trending anticlinal ridge thathas had an important effect on the localdrainage pattern (see Figure 3.2). The ridge,here named the Sabinos Anticline, is a re-sistant structural high presently reachingjust over 300 m elevation that allowed an-other strike-oriented stream to develop onthe Méndez shale in the syncline betweenthe Sabinos Anticline and the main range.This stream is here named the Río Sabinos(distinct from the Río Sabinas in the GómezFarías area), after the village of LosSabinos. The name is applied to the fossilriver in this valley, and loosely to the presentdisrupted drainage as described below.

The Río Sabinos eroded downward un-til it impinged on the El Abra limestone ina number of places. Eight major and anunknown number of lesser stream piracieshave occurred over a period of time, so thatmost of the drainage is now diverted un-derground. The piracies occur at large jointsthat often have a vertical range of more than50 m. Once the diversion is effective, thestream begins to erode downward rapidly,because the piracy causes a sudden lower-ing of local base level. An arroyo, initiallydeveloped on San Felipe beds, is shown inPlate 3.9 where it has passed against thedip onto the El Abra limestone and has cuta deep canyon that leads to a degradedcapture point. The amount of downcuttingdepends on the size of the stream and theantiquity of the capture. Mitchell et al.(1977) have illustrated the process ofstream capture and degradation of the sinkfor this area. Most of the pirated waters areprovided by runoff from San Felipe bedswest of the Río Sabinos, on the east limbof the anticlinal ridge. Mitchell et al. (1977)and Russell (personal communication) havegiven estimates of the drainage area for eachof the pirated streams. Improved estimates

39


based on air photos are listed in Table 3.1.The areas should be accurate to ± 5%, ex-cept for the smallest ones. Only the drain-age area on impermeable rocks is given(basin boundaries are shown in Figure 3.2).The drainage basin of some of the swalletstreams may be significantly different (usu-ally smaller) than in the past, because up-stream piracies may divert some of the flowunderground or to another surface channelout of the basin. Also, the area of imper-meable cover may have been slightly re-duced during the time the swallet caveshave existed. The air photos show that for-merly there were many tributaries to the RíoSabinos from the east side of the channel,i.e., the west flank of the Sierra de El Abra.These sources have largely dried up as mostof the impermeable cover has been removedand the channels have become karstified.

A morphologic pattern similar to thatdescribed above occurs higher up the westflank of the range. Two and one-half kilo-meters east of the Río Sabinos, there is a verylarge (300 to 1000 m wide) strike-orienteddry valley (Figure 3.2) that developedwhere the dip reduces or possibly reversesdirection (anticlinal (?) ridge on the westside of the valley). It formerly received run-off from many dip slope tributaries on itseastern side. At the upper end of the valleyin the vicinity of Sótano de Soyate, thepresent valley floor has an elevation ofabout 290 m. The drainage was to the south,where it eventually turned southwest to join

limestone was reached.

3.1.4. Passes in the El Abra range

There are three large wind gaps and onlyone water gap in the Sierra de El Abra (Fig-ure 3.1). Two of the wind gaps are calledPaso de El Abra or sometimes simply ElAbra (the opening); hence, it is proposedhere to refer to them more specifically asPaso de El Abra del Sur (South Pass) anddel Norte (North Pass). South Pass occursin the southern El Abra east of Cd. Valles (seeFigure 3.2). Its fossil river first impingedonto the El Abra limestone probably atabout 300 to 350 m and cut downward toabout 160 m at the west end (high end) ofthe pass. Initially, the drainage may haveincluded all of Río Valles and the Río Sabi-nos (Section 3.1.3). If so, it was probablyaltered relatively early by stream piraciesto look nearly like the present drainagepattern (Figure 3.1), with only the ancientRío Sabinos continuing to flow through it.Finally, the pass was abandoned as the westflank of the El Abra became karstified,some streams were pirated underground,and still others captured by the Río Valles.

For some time there was also a significanttributary network developed on the crestin the vicinity of the pass, particularly onthe north side, that created the most dis-sected part of the crest.

North Pass crosses the range betweenAntiguo Morelos and Cd. Mante (Highway85 uses the pass). The river that created itfirst reached the El Abra limestone prob-ably around 450 m at the point where therange has a NW-SE kink and where it hassubsided from the high central zone. Theriver cut downward, exposing more lime-stone, down the SW dip until it reached250 m (see location map inset in Figure6.4). Then the stream was captured by themodern Arroyo El Lagarto, which is tribu-tary to the Río Comandante.

The third wind gap is called PuertoChamalito on the new 1:50,000 topographicmaps (Loma Alta, no. F-14 A-59), and it istaken as an arbitrary northern boundary ofthe Sierra de El Abra. On air photos, it islarger and more spectacular than all othergaps in the El Abra. It crosses the rangewhere the crest is presently about 400 m. Ariver that formerly drained the southernflank of the Sierra de Guatemala passed

Table 3.1Drainage area of the swallet caves

Cave area(km2)

Sótano de Venadito 7.7

Sótano de Yerbaniz 21.8

Sótano de Matapalma 1.9*

Sótano de Japonés 8.6

Sótano del Tigre 4.4

Sótano del Arroyo 12.4

Sótano de la Tinaja 5.1

Sótano de Montecillos 8.4

Sótano de Jos 4.3

Sótano de las Piedras 0.3*

Sótano de Palma Seca 0.7*

Total area of swallet cave basins 75.6

* was previously much larger, but is nowreduced by upstream piracy (see Figure3.2)

the Río Sabinos. Now the area iscompletely karstic and includesdevelopment of closed solutionvalleys, but the largest geomorphicfeatures are the paleo-fluvial val-ley and intervening ridges. An-other 1.5 km up the west flankfrom this dry valley there appearsto be yet another, although smaller,strike-oriented valley that likewiseeventually joined the Río Sabinos.

North of the Los Sabinos area,the Sabinos anticline apparentlyplunges, which allows local streamsto flow westward to join the RíoPuerco. However, 20 km north ofCd. Valles, this or another anti-cline rises sufficiently high thattwo streams have eroded down tothe El Abra limestone and havebeen pirated. Three large swalletcaves, sótanos Japonés, Matapal-ma, and Yerbaniz, have developedabout 3 km west of the El Abra–San Felipe contact along the westmargin of the range. One other im-portant swallet cave near the stateline, Sótano de Venadito, is formedwhere a stream flowing from asynclincal valley having an imper-meable cover crossed an anticlinalridge to the west and was capturedunderground when the El Abra

Plate 3.9. Stream capture along the westernmargin of the range. An arroyo developed onMéndez and San Felipe beds was capturedwhen the stream exposed a 70 m deep joint inthe El Abra limestone. The swallet is Sótano delArroyo (lower right). The Cueva de los Sabinosentrance sink is visible in the upper center.

40


through the canyon. Ultimately, the tribu-tary streams were either pirated under-ground or captured by the Río Comandante;the area is now completely karstic.

The Río Comandante (also called RíoOcampo and Río Boquilla) forms the onlycontemporary water gap in the El Abrarange at Cañón Servilleta. This river col-lects runoff from the northern AntiguoMorelos valley and the Ocampo valley westof the Sierra Tamalave. It has captured allthe surface runoff in this area because ithad the advantage of crossing the Sierra deEl Abra where the El Abra limestone isstructurally lower than at any other placein the north—about 220 m elevation. Theriver is believed to be superposed on theEl Abra. At present, it has cut downward to120 to130 m at the west side, where thereare rapids and waterfalls, and to 73 m onthe east side. A much deeper canyon oc-curs where the Comandante crosses theSierra Tamalave.

3.1.5. Origin of the escarpment of theSierra de El Abra

Although it was not an objective of thethesis research to determine the origin ofthe escarpment on the east face of the Si-erra de El Abra, some pertinent facts con-cerning this problem have been observed.The origin of this most striking feature hasbeen a controversial subject for many de-cades. A fault has been advocated by Bonet(1952), Muir (1936), Nigra (1951), andCarrillo (1971); a fold is strongly-promotedby Heim (1940) and Mitchell, et al. (1977);and a steep platform edge is suggested byGriffith et al. (1969), Enos (1974), and tosome extent by Rose (1963). A preliminaryreport on the results of this study (Fish andFord, 1973) gave a cross section having afaulted scarp, in accordance with mostMexican geological reports, but no directevidence of a fault has been observed dur-ing this research. Most of the authors rec-ognize that in the vicinity of the scarp thereis a transition zone from platform to basinfacies limestones, and a few seem to sug-gest that there may be a combination ofsteep platform edge and faulting. The fieldobservations and arguments of previousauthors will not be reiterated here, althoughthey are taken into some account, especiallywhere confirmed in this study. Instead, whatis offered are the observations made dur-ing the field work and an analysis of newand old drilling information. The lack ofdeep borehole data very near the scarp hashindered resolution of the problem.

Two important wells have been drilledin recent years (Carrillo Bravo, 1971). ThePozo Tamlihuale No. 1, located 4 km westof the reef and near El Pujal, penetrated1808 m of the El Abra formation, followedby 287 m of Lower Cretaceous and Juras-sic beds before reaching metamorphic rocks(basement?) at 2011 m below sea level.Pozo Trangulo No. 1, located 3 km east ofthe Sierra de El Abra, cut 394 m of UpperCretaceous beds, only 373 m of the El Abra,followed by 452 m of Lower Cretaceousand Jurassic beds, ending at a depth of1128 m below sea level in the Jurassic (wellinvaded by water, probably fresh). In thelatter well, the El Abra formation is the fore-reef facies and consists of 250 m of porousand permeable dolomite overlain by 115 mof dolomites that alternate with beds ofcalcarenites and oolites. Other wells 6 to10 km east of the Sierra de El Abraencounted either a “mixed facies” (Tama-bra) which is composed of interfingeringbeds of reef debris and dense, dark limemudstones with chert (basin facies), or justthe basin facies. Carrillo (1971) calls thebasin equivalent of the El Abra the Cuestadel Cura, rather than the Tamaulipas, whichhe correlates with the Lower Cretaceousplatform facies. These wells clearly showthat the middle Cretaceous beds thin by afactor of 5 in a very short distance and con-tinue to thin until diminished by a factor of10 to 20. Also, they show that the Jurassicbeds are about 800 m structurally higherunder the west edge of the coastal plain thanunder the east edge of the Platform. Thus,the reef zone cannot be more than perhaps2 km wide. In fact, it may scarcely be widerthan the outcrop, but may have migratedslightly with time. Exactly the same rela-tionships as described here are reported forthe Golden Lane Platform by Viniegra andCastillo-Tejero (1970), where the sameproblem has been argued.

Mitchell, et al. (1977) have argued thatthe El Abra reef is possibly 8 km wide,nearly all of which is buried below thecoastal plain; that the east face of the rangeis the upturned west edge of the reef (andthus is essentially a dip slope); and that thefolding occurred approximately at the junc-tion of the weaker, bedded backreef lime-stones with the massive reef. As shownabove, a “massive” reef does not exist inthe area. The most likely locus of deforma-tion would be the transition zone from themassive platform limestone to the thin ba-sin facies limestone and weak overlyingshales. The structural relief on the LowerCretaceous and Jurassic beds described

above suggests that a deep seated structuresuch as a fault block resisted the eastwardtectonic movements along the platformedge. The section of basin Cretaceous andJurassic rocks presently known hardlyseems capable of providing sufficient re-sistance to cause a major fold in the Plat-form limestones.

Other significant points concerning thescarp as noted in this study are:

1. The Méndez shale occurs directly atthe base of the scarp and probably for ashort distance up the face. Dips in thesebeds seem to be proof that some foldingdid occur.

2. There appears to be little structuralrelief on the crest of the range, althoughsmall folds and some longitudinal fault (?)traces may be seen on air photos. Strataexposed in the North and South El Abrapasses and in caves remain horizontal tothe reef zone. Thus, nearly all of the struc-tural relief of any alleged deformation mustbe located between the top and the base ofthe escarpment.

3. The dips of facies (bedding) planesin the reef (which may not have been hori-zontal originally) as exposed in caves, andstratal dips in the Méndez or San Felipe atNacimiento del Arroyo Seco on RanchoPeñon are much less than the steepness ofthe escarpment. An extreme case occurs atGrutas de Quintero (Chapter 6), where thedip of a major facies plane is about 5°, muchless than the slope of the east face there.

4. None of the caves in the east face oron top of the range that have been investi-gated in this thesis bear any trace of theintense crumpling, shearing, and faultingthat would be expected in the interior of amajor fold. A cave that is presently beingexplored (Neal Morris, 1976) is partly de-veloped in one of the supposed faults justwest of the top of the east face; however, itdoes not seem likely that there is a largevertical displacement along the fault, be-cause it has no effect on the surface relief.

5. The slope of the scarp must be slightlyreduced from times past by erosional pro-cesses, particularly solution and rock fall.

6. The width of the reef zone remainsabout the same throughout the length of theSierra de El Abra. Just south of the NorthEl Abra Pass, where the range turns to thenorthwest, N-S trending folds in the bed-ded facies on the crest are truncated at thenorthwest trending reef. The implication isthat the present escarpment represents theupper part of the Platform edge, and thatsome deep-seated basement structure or

41


topographic high controlled the locus oforiginal reef development.

The data indicate that some folding hasoccurred, but only part of the relief of theescarpment can be attributed to it. Muchof the relief must be caused by exhumationof a steep platform edge, or by faulting, orboth. Thus, some sort of combinationmodel, possibly involving all three sug-gested origins, must be developed. It issuggested that the Platform edge wasshoved eastward up a steep basement slopeor against a basement fault block to createthe present Sierra de El Abra structure. Theplatform edge (east face of the El Abra)would thus be uplifted relative to the basinrocks, causing some folding and possiblefaulting.

3.2. Karst Geomorphic Featuresin the Region

A brief description of some of the karstlandforms in the rest of the Valles–San LuisPotosí Platform region will be given in thissection. The discussion is unfortunately farfrom comprehensive, because so little is

really known about the region. Reconnais-sance trips have been made into some ar-eas, but a large portion of the karst areahas never been observed.

At its north end, the Sierra de El Abramerges into the Sierra de Guatemala, amuch larger structure. It is a huge anticli-norium, having more than 2 km of struc-tural relief and reaching an elevation of2320 m. The upper Cretaceous rocks wereremoved long ago, but the paleofluvial val-leys can be traced up to at least the 1000 mlevel. A very rugged surface has developedthat includes intense karren formation, karstpinnacles, hoyas (irregular closed karstvalleys, commonly 200 to 1000 m wide and25 to 100 m deep), areas of cone karst, frac-ture controlled solution corridors or val-leys, collapse structures, caves and pits, anda polje at the village of Joya de Salas. Onthe east and south flanks of the range, thewettest areas, the precipitation probablyexceeds 2000 mm, and there is a thick tropi-cal jungle. Except on the west side of therange, where there may be some surfacerunoff, all rainfall sinks rapidly into thekarst. Most of the ground water flows tothe great springs at the head of the Río Frío;

another important spring lies on the RíoSabinas.

The most spectacular karst landformsand scenery in the region occur in theXilitla and Aquismón areas, south of Cd.Valles (Plate 3.10). There the elevationsrange from 100 m to 3000 m, and the rain-fall may exceed 2500 mm per year. Intensekarstification has produced landforms simi-lar to the classic tropical karst found inCuba, Jamaica, Puerto Rico, and the EastIndies. Russell (1968) has briefly describedthe geology and geomorphology of the areawest of Aquismón; Bonet (1953) andRussell and Raines (1967) have describedsome of the caves. Included in the Xilitla-Aquismón area are large karren forms andpinnacles throughout, abundant cone andcockpit karst up to about 1300 m, cones onsteep slopes up to about 2000 m, hoyas andshallow dolines, uvalas, at least one polje,located west of Tamapatz, that is about 1 kmwide and 10 km long, and caves and pits.The area west of Aquismón contains someof the world’s most spectacular collapsestructures, such as the 333 m deep Sótanode las Golondrinas (Fish, 1968). West ofXilitla, there is a massif that has a large

Plate 3.10. The spectacular Xilitla-Aquismón area karst. The Aquismón area at about 600 to 1000 melevation is on the left, and the high plateau above Xilitla is on the right, at 2000 m or more. Karst devel-opment here is the most intensive in the region.

42


summit area between 2000 and 3000 m.Solution effects there are intense, butkegelkarst and cockpit karst are generallyabsent. The Huichihuayán and San Juanito(also called Mujer de Agua) are large per-manent springs located along the eastmargin of the area at the base of an escarp-ment; other seasonal springs may be foundin the interior, particularly along the Ar-royo Seco, a major valley in the area.

West of Cd. Valles, the Sierra MadreOriental comprises a series of N-S trend-ing folded ranges (see Figure 1.1 and Fig-ure 2.5). Most of the ranges reach muchhigher elevations than the Sierra de El Abra.Nearly all of the impermeable rocks havebeen removed except in some valleys, andkarst solution processes have been activefor a very long time. Karren, dolines, largeclosed karst valleys, and other karst land-forms are abundant; but the classic formsoften considered omnipresent in tropicalkarsts (namely cones and cockpits) have notbeen observed yet, or are not sufficientlydeveloped to be recognizable. Most of thearea has not been visited, but air photos ofLa Sierra la Colmena and the new 1:50,000topographic maps show many vestiges ofan older fluvial network, and there are someactive channels. Part of the area may beconsidered holokarstic, while the remain-der is fluviokarstic. Massive travertinedams, constructional karst forms, are foundin Micos canyon, El Salto falls, in the low-est Río Tampaón canyon, and on the RíoGallinas just west of Tamasopo.

Very little is known about the extent ofkarstification of the El Abra limestone ex-posures in the western half of the Platform.From a few observations, it is clear that thelimestone has undergone significantsolutional modification, even though theprecipitation is only 300 to 500 mm per year(rain shadow of the Sierra Madre Orien-tal). Very few closed valleys are shown onthe topographic maps, which have a con-tour interval of either 10 or 20 m. One largespring, Manantiales de Media Luna, isknown. It flows from a 75 m deep lake afew hundred meters across, at the southernend of the large, flat Río Verde valley. Threeorifices at the bottom yield the dischargedwater. Two high mountain karsts have de-veloped that are of particular interest anddeserve more study. One is the area west ofJalpan, Querétaro, where the mountainsare high enough to cause increased pre-cipitation. From Jalpan to Pinal deAmoles (about 3000 m elevation), karstforms are developed in a variety of climaticzones, structural and lithologic settings, and

topographic positions. Some very largecaves and collapse pits have been found,and a few springs are known. The other sig-nificant high mountain karst is located atValle de los Fantasmos, approximately 30km east of San Luis Potosi at about 3000m elevation. The area is named Valley ofthe Phantoms because heavy ground fogoften occurs, so that people, animals, andobjects sometimes seem like apparitions.The area is replete with karst pinnacles,sinkholes, and pits.

A large portion of the western half ofthe Platform is covered by valley flats.Many of the valleys are or were bolsons,although the sediment thicknesses reportedby Carrillo (1971) are surprisingly thin.Gypsum karst has been observed in someplaces and may be fairly common. A widevariety of other landforms occur, but willnot be discussed.

3.3. Summary

The erosional history of the region hasreceived very little attention in the litera-ture and in this study. The early history willbe difficult to deduce because the knowncontinental sediments are regarded as lateTertiary or Quaternary age. Segerstrom(1961, 1962) has briefly discussed the ero-sional history of two adjacent areas to thesouth and southwest. It appears that ero-sion on the Platform very likely started inthe early Tertiary, or possibly very late Cre-taceous. It is suggested that the major riv-ers were established approximately in theirpresent positions relatively early (middleto early Tertiary) and were gradually low-ered onto the El Abra limestone. The RíoTampaón is located in the structurally low-est part of the Sierra Madre Oriental foldbelt, and passes onto the coastal plain justat the point where the Sierra de El Abrastructure dives beneath the Méndez shale.It could more quickly erode downward andcapture a large drainage area than any ofits competitors. Thus, the Río Tampaón isnow the master drain for the southern halfof the region, and the lowest point in theregion occurs where it passes onto thecoastal plain. In the north, the Río Coman-dante and the Río Guayalejo were also es-tablished long ago, but their developmenthas been hampered by very high folds andsmaller areas of structurally low imperme-able rocks or sediments to form catchments.Hence, the Río Guayalejo has not providedas low a base level in the north as the RíoTampaón in the south because its dischargeis less and because it has been “held up”

by exposures of the El Abra limestone andby basalt flows on the coastal plain. All themajor rivers cut across large folds, form-ing spectacular water gaps. For example,the canyon of the Río Guayalejo at the northend of the Sierra de Guatemala is as muchas 1000 m deep.

As erosion exposed the El Abra lime-stone on structural highs, karstificationbegan. The limestone in the high centralportion of the Sierra de El Abra probablywas first exposed during Pliocene time.Remnants of the paleo-fluvial system maybe seen on the crest towards the ends ofthe range where it is lower (and more re-cently exposed), and all along the westernflank. Solution features are most highlydeveloped in the higher part of the range,but virtually all of the range has karsticdrainage. The eastern flanks of the higherranges nearby receive much more rainfallthan the Sierra de El Abra, have been ex-posed much longer, and have tropical karstlandforms. The karst landforms found inthe region show that a variety of karst land-forms occur in tropical karst regions—conekarst and cockpits do not predominate.

Flat-topped hills on the coastal plain infront of the El Abra range and waterfalls inthe Micos and Servilleta canyons suggestintermittent uplifts of the area during laterPleistocene time. A deposit of continental(?) limestone (Heim, 1940) on the hill southof Quintero (Plate 3.1) may have beenformed very near sea level during a stableperiod and then uplifted.

Among the many geomorphic problemsthat need further examination are:

1. The erosional history of the region.2. The processes and rate of stripping of

impermeable rocks from the crest andflanks of ranges and of subsequentkarstification; stripping of the SierraTamalave has lagged well behind thatof the Sierra de El Abra.

3. The occurrence of large fossil swalletcaves higher up the west flank of the ElAbra range; fossil swallet caves withcanyons similar to those in the LosSabinos area should exist, particularlyalong the previously discussed dry val-leys parallel to the Río Sabinos; how-ever, none have yet been discovered andonly a few potential sites are knownfrom the air photos.

4. Karst landforms as a function of climate,structural attitude, carbonate rock typeand impurities, slope, time and inher-ited paleogeomorphic forms.

43


4

HYDROLOGY AND HYDROGEOLOGY

4.1. Introduction

In this chapter the basic elements of thehydrology of the region are examined. Thisincludes aspects of the hydrogeology, thehydrology information base, the temporaland spatial distribution of runoff, and theground water. The treatment will be on twolevels. Most of the chapter concerns theregional hydrology, but ground water isdiscussed at a local level with reference tofield studies conducted in the Sierra de ElAbra. The objective is to develop an un-derstanding of the hydrology of this majorkarst area, and to lay the groundwork forthe analysis and synthesis given in Chap-ter 7.

4.1.1. Hydrogeology

An important part of the study of thehydrology of a region is to identify thehydrogeologic units and their hydrologicproperties. In this section, the rock forma-tions will be grouped into four broad cat-egories of hydrogeologic behaviour anddescribed.

Using state geologic maps provided bythe Instituto de Geología, geologic mapsby J. Carrillo (1968) and Segerstrom(1961), and the new 1:50,000 Cetenal topo-graphic sheets, a preliminary hydrogeologicmap (Figure 4.1) has been prepared for theValles–San Luis Potosi Platform and fringe.Unfortunately, the source maps differ con-siderably in age, detail of mapping, nomen-clature, and type of stratigraphic unitmapped. Recent work of Pemex geologistssynthesized by J. Carrillo (1971) has re-solved some of the geologic problems inthe area and has provided a much betterunderstanding of the regional geology.However, the scale and paucity of culturalfeatures on his 1968 map make it difficultto use, e.g., to determine exactly the traceof the middle Cretaceous platform edge. Itis desirable that more detailed geologicmaps be made that employ a uniform no-menclature for rock stratigraphic units,so that, among other things, a systematic

study may be made of the hydrogeologicunits and an improved map assembled.

The El Abra and Tamasopo Formationsplus the suspected lower Cretaceous plat-form facies limestones form the principalaquifer of the region, and its permeabilityappears to be karstic, i.e., through solutionconduits. The El Abra Formation occurs atthe surface or in the subsurface over nearlythe whole Platform. Around the margin ofthe Platform, exposures of the deeper wa-ter equivalents of these formations mayexhibit some secondary permeability. Also,the lower Cretaceous shelf facies lime-stones may extend north of the middle Cre-taceous platform (Enos, 1974). Where thedeeper water or lower Cretaceous lime-stones are mapped adjacent to the platformfacies, a question mark (?) indicates uncer-tain hydrologic conditions. It is thought thatmost of the total runoff from the region haspassed for at least some distance throughthis karstic aquifer. Where there are steepslopes, thick soil, or insufficiently devel-oped secondary permeability such as on“recent” exposures, surface runoff mayoccur.

The second unit comprises all the im-permeable rocks. It includes the upper Cre-taceous and Tertiary formations of marineorigin except the Tamasopo Formation (ma-rine Tertiary formations only occur east ofthe Platform), all the igneous rocks, and afew small exposures of preCretaceousrocks. The San Felipe, part of the Cardenas,and some of the volcanic rocks may haveminor permeability. This unit yields directsurface runoff.

The third unit comprises all the Tertiaryand Quaternary continental deposits ofsedimentary origin. These sediments havea variety of origins and are poorly known.Most maps simply show Quaternary allu-vium, but several Tertiary units are givenon the San Luis Potosí map, and J. Carrillo(1971) briefly describes but does not maptwo Pliocene age formations. Because ero-sion of the region commenced in early Ter-tiary time, there must be deposits datingback to that time in the bolsons of the

northwestern part of the Platform and, per-haps, beneath some of the volcanic rocks.The sediments include fragments of volca-nic rocks and Cretaceous limestones, finermaterial derived from upper Cretaceousterrigenous rocks and volcanics, and chemi-cal precipitates such as gypsum, tufa, andcaliche. The thickness of these sedimentsis reported by J. Carrillo (1971) to be 20 to30 m in the Río Verde valley and more than50 m in the large valley northwest of Cd.del Maiz. They are also generally thoughtto overlie the El Abra Formation, thoughin places the impermeable Cárdenas For-mation might intervene. Thus, precipitationmay be able to pass through the sedimentsinto the El Abra limestone. Their perme-ability is expected to be highly variable,from insignificant to moderate. J. Lesser(unpublished S.R.H. report, c. 1973) hasshown that the potentiometric surface of thelarge Río Verde valley sediments indicatesflow toward the Río Verde. From dischargemeasurements, he estimates the averageannual seepage to be 16×106 m3, or an av-erage of 0.5 m3/sec. Gypsum karst has beenobserved at a few localities in the long val-ley south of Matehuala (personal observa-tions, and personal communication withWilliam H. Russell).

The fourth unit is the lower CretaceousGuaxcama Formation, which has about 300km2 of outcrop in the interior of theGuaxcama anticline, 40 km northwest ofthe town of Río Verde. It is dominantlygypsum and anhydrite (at depth), and itwould not be conducive to deep circulationof water were it not for interbeds of lime-stone and dolomite. At least some shallowflow seems possible, and a gypsum karstmay have formed. It is distinguished fromthe other units because its hydrologic prop-erties are unknown, and because it isthought to play an important role in theground water chemistry. In the subsurface,it likely represents an impermeable barrierbelow the El Abra Formation.

In summary, the El Abra and theTamasopo Formations plus the platformfacies of the lower Cretaceous limestone

45


Figure 4.1. Hydrogeologic map of Valles–San Luis Potosí Platform.

46


are thought to form the only significantaquifer in the region, and its permeabilityis karstic. The areal extent of this unit definesa hydrogeologic province that is expectedto have only a small flux of ground wateracross its boundary. Other rocks in the re-gion are impermeable and yield surfacerunoff, or are slightly permeable. Most ofthe rest of this dissertation will be con-cerned, either directly or indirectly, with thehydrologic character and function of thekarst aquifer, particularly the El Abra For-mation.

4.1.2. The drainage basins and gaugingstations

The principal streams of the fluvial net-work that drain the region are shown inFigure 4.2. About 37% of the Platform areais not integrated into the surface drainagesystem. Part of the remaining portion yieldsdirect surface runoff from impermeablerocks (refer to the hydrogeologic map, Fig-ure 4.1), but much of the Platform is afluvio-karst. The two principal rivers arethe Río Tampaón in the southern part ofthe area and the Río Guayalejo in the north-ern part. Most other streams are tributariesto these two either on the Platform or onthe coastal plain a few kilometers from theeastern margin of the Platform.

Mexican water resource planners haveestablished a good system of gauging sta-tions, controlled mainly by the Secretaríade Recursos Hidráulicos (hereinafter re-ferred to as the S.R.H.). Figure 4.2 is takenfrom a map given in S.R.H. BoletínHidrológico No. 32 (publication date notgiven, but after 1968). It shows the loca-tion of the gauging stations with referenceto the rivers. When more detailed mappingis accomplished, it will be seen that theamount of karst between the channels ismuch greater than shown on the S.R.H.map.

Some information about the gauging sta-tions, brief basin descriptions, and othercomments are given in Table 4.1. Detailedinformation concerning the type of instal-lation, method of measurement, and thecomplete discharge record of each stationthrough 1968 in the whole Río Pánuco ba-sin is contained in S.R.H. Boletín Hidro-lógico No. 32. Additional discharge recordswere obtained from the Tampico office ofthe S.R.H. for some stations through 1971,the latest year for which some data had beenprocessed. The period of record shown inTable 4.1 is the period available for thisstudy. The oldest stations were begun in

1927 for water availability studies for irri-gation development of the Mante and Fríosprings. Much later, El Salto (1948) andLa Encantada (1950) were added, again forwater use projects. In 1954, three stationsbegan operations on major rivers in thesouthern part of the region, and in the pe-riod 1959–61 several more important sta-tions were added that nearly completed thenetwork of gauges as it now stands. Forbasin yield calculations given in Section4.3, the base period selected is 1961–68.

4.1.3. Distribution of precipitation

The geographic distribution of rainfallin the region is shown in Figure 4.3. Datasources used were the S.R.H. Boletín No.19 (December 1962), which has informa-tion through 1961, and unpublished datathrough 1973 provided by the S.R.H. for alarge number of the stations in the easternpart of the region. Because a base periodof 1961–68 is used in Section 4.3 for basinyield calculations, the average precipitationused for the stations in the principal areaof interest (the eastern part of the region)is for the same period. For the rest of theregion, the data points are drawn mainlyfrom stations having long records listed inBoletín 19. Many of these stations have 25to 40 years of data. A comparison of theaverage precipitation measured at sevenstations that have long records against theiraverages for the 1961–68 interval is givenin Table 4.3. The base period average an-nual precipitation appears to be not too dif-ferent, possibly slightly drier, than thelonger term average annual precipitation inthe region.

There are two factors that influence theuse of the data as a measure of the actualprecipitation in the region. The first factoris that nearly all of the climatology stationsare located in valleys in or near towns. Aspreviously described, a large part of theregion has a very large local relief. Some-what greater precipitation must occur on theridges or massifs adjacent to the valleys.In a few places where gauges are appropri-ately situated, this effect can be clearly seen.The second factor is the uncertain capabil-ity of the stations to measure the oftenheavy dew, the frost, and the light snow-falls that occur. These forms of precipita-tion in the area could not make an importantdirect contribution to the runoff, but mayprovide some moisture to vegetation.

The rainfall distribution is presented asan isohyetal map having a contour intervalof 350 mm, beginning at 650 mm. Greater

detail is not justified because of the firstfactor discussed above and the mixed yearsof data. Considerable judgement has beenused in drawing the isohyets, based on to-pography, exposure, and to a limited ex-tent on field observations.

Warm, moisture-laden air moves fromthe Gulf of Mexico westward across the110 km wide coastal plain (650 to 1100 mmprecipitation) to bring water to the regionof interest here. The zones of greatest pre-cipitation are coincident with the first groupof high ranges (higher than 1000 m) of theSierra Madre Oriental. There, the summerrains are torrential. The Xilitla area isknown to have an average of more than2000 mm, and it is likely that the easternflank and crest of the Sierra de Guatemalaand small similar areas of the fold beltridges northwest of Micos exceed thatvalue. In many years the total rainfall atXilitla and at Tamapatz has exceeded 3000mm. The Sierra de El Abra only reaches anelevation of about 750 m in the central partof the ridge. It was observed to have someeffect on clouds, but unfortunately there areno climatology stations on the range. Theeffect, however, is thought to be small com-pared with the effect of the much higherranges. The central portion of the Sierra deEl Abra might receive 1300 to 1400 mm,whereas the nearby valley stations receivebetween 1100 to 1200 annually.

On the westward (leeward) side of thefirst high ranges, a very strong rain shadowexists. Most of the western two-thirds ofthe region receives 350 to 550 mm of rain-fall. Broad valleys between 1000 and1300 m comprise a large portion of the area(Figure 1.1). A second series of high rangesis present along the western margin of thePlatform. West of Jalpan, ridges up to 3000m elevation occur. The sequence of greaterrainfall on eastern slopes and crests fol-lowed by a leeside rain shadow is repeated,and a high mountain karst has formed. Asimilar pattern has been observed about 40km east of San Luis Potosí, at the Valle delos Fantasmos.

4.1.4. Springs

The development of karst ground-waterflow systems is the single most importantaspect of the regional hydrology. There area great many springs, principally in the east-ern third of the Platform, and some of themare among the largest in the world. Thephysical and chemical characteristics of thesprings and the ground-water systems willbe extensively described and investigated

47


Figure 4.2. Drainage basins and gauging stations of the region. The numbers are keyed to Table 4.1.Springs: (a) Choy, (b) Santa Clara, (c) Mante, (d) Frío springs (includes Poza Azul and others), (e) Sabinas,(f) Coy, (g) Pimienta, (h) San Juanito, (i) Huichihuayán, (j) Agua Clara, (k) Tanchachín, (l) Puente de Díos,(m) Media Luna, (n) name unknown.

in later sections of this and other chapters.The more important springs that this thesiswill be concerned with are the Coy, Choy,Santa Clara, Mante, and Frío, all of whichlie along the eastern edge of the Platform.Locations of the large springs known to theauthor are shown in Figure 4.2.

4.2. Temporal Aspects of the Hydrology

Some of the temporal aspects of the hy-drology of the region will be discussed inthis section. It is not the purpose of thisthesis to quantitatively determine the longterm periodicities of annual rainfall and

discharge in the region, or to make statisti-cal or predictive models of their relation-ships. Rather, the objectives are to analyzeshort term events and the seasonal patternto establish a general understanding of therainfall-runoff relationship and the dynam-ics of the karst aquifers.

48


Figure 4.3. Annual rainfall in the region. Numbered stations refer to Table 4.2.

4.2.1. Long-term trends at El Choy spring

The relatively short record (1960–1971)at El Choy is given in Figure 4.4 to showwhat had happened in the years leading upto this study (1971–1973) and to suggestthe controls on the annual runoff. The ra-tio of the yearly runoff to the average an-nual runoff for the 12 years of El Choy datais compared with similar ratios for thesame time base of the rainfall at the El Choyand the Santa Rosa (Cd. Valles) stations.

Considerable variation does exist, con-trolled by the very large scale phenomenathat create tropical storms and hurricanesin the Atlantic and Gulf of Mexico regionand guide them onto the mainland. Theperiod 1960–1965, excepting 1961, was asequence of dry years and below normaldischarge, followed by wet or about aver-age years from 1966 up to the thesis fieldstudy period.

For most of the years the runoff ratiohas the same sign as and is more extreme

than the rainfall ratios. Thus, as might beexpected, the annual precipitation is usuallythe most important factor in determiningthe annual runoff at El Choy. Three otherfactors are the number, intensity, and dura-tion of precipitation events during eachyear; ground water stored during antecedentyears; and the degree to which the pre-cipitation measured at the climatology sta-tions around the periphery of the Sierra deEl Abra reflect the true rainfall on the range.

Despite only slightly deficient rainfall,

49


Table 4.1Gauging stations in the Valles–SLP area

Number on Elevation Surface drainageGauging station Figure 4.2 River (m) basin area (km2) Period of record

Tancuilín 1 Tancuilín 7 321 1961–68+

Requetemú 2 Axtla 71.1 661 1954–68+

Ballesmí 3 Coy 29.1 194 1954–71+

Ojo Caliente 4 Santa María 1768 2574 1966–68+

Tansabaca 5 Tampaón 174 17532 1959–72+

El Pujal 6 Tampaón 28.2 23373 1954–71+

Nogal Obscuro 7 Verde 1027 2444 1965–68+

(Media Luna) 8 Media Luna 1000 — sporadic measurements

Vigas 9 Verde 897 3571 1958, 1961–78+

Tanlacut 10 Verde 257 6039 1961–71+

Galllinas 11 Galllinas 280 789 1959–71+

El Salto 12 El Salto 399 900 1948–64 suspended

Micos 13 los Naranjos 210 1978 1961–71+

Santa Rosa 14 Valles 72.8 3521 1959–71+

El Choy 15 Choy 32.2 12(4.9) 1960–71+

Santa Clara 16 Santa Clara 70 16.5 ? 1972–73+

Mante 17 Mante ~80 42(69) 1927–30, 32–35, 38–61, 63, 67–69, 71, 72+

Canal Oeste (West) 17 West Canal — — 1932–35, 38–61, 63, 67, 68+

Canal Este (East) 17 East Canal — — 1932–35, 38–61, 63, 67, 68+

La Servilleta 18 Comandante 75 2532 1928, 1960–72+

Río Frío 19 Frío 53.7 2785 1961–68+

Ahualulco 20 Poza Azul 78 (?) 17 1946, 1948–68+

Salida Conducto Cerrado 20 Canal Alto 87–88.5 — 1949–68+

Ahualulco 21 Nacimiento 78 25 1946–68+

Ahualulco 22 Frío 75.5 45 1927–35 suspended

Canal Bajo 22 Canal Bajo 77–77.5 — 1960–68+

Puente “T” 22 Canal Bajo — — 1951–54, 56–59 suspended

Sabinas 23 Sabinas 88.8 497 1961–68+

La Encantada 24 Guayalejo ~310 3725 1950, 52–54, 56–68+

San Gabriel 25 Guayalejo 129 4937 1943–68+

50


Table 4.1continued

Gauging station Description of basin; comments

Tancuilín short, high-gradient river in karst area; probably springs

Requetemú becomes R. Axtla below junction of R. Huichihuayán and R. Tancuilín; drains high-relief karst area;Huichihuayán, San Juanito, and Pimienta known large springs

Ballesmí small, low-relief surface basin on shale; contains the great Coy spring and other smaller ones in theCoy dome; possible small springs near Aquismón

Ojo Caliente headwaters of R. Santa María; off platform; very small flow; used for irrigation since 1968

Tansabaca R. Santa María becomes R. Tampaón below confluence with R. Verde; drains high-relief region of variedrock types; many important tributaries, especially in eastern portion of the basin; springs gauge R. Tampaón

El Pujal the largest and base-level river of the region; varied rock types, high relief, cuts across tectonic structures

Nogal Obscuro measures flow from headwaters of R. Verde to R. Verde valley; high relief; varied rock type

Media Luna actually, there are six stations on the five irrigation canals and the channel of this large spring

Vigas near boundary of flat R. Verde valley and Sierra Madre fold belt to the east; some flow from Media Luna

Tanlacut located 10 km upstream of confluence with R. Santa María; varied geology and relief

Gallinas high-relief basin; karst and springs; some sediments and impermeable rocks

El Salto high-relief karst basin; high-gradient stream; used for hydroelectric power just below station

Micos below confluence of R. El Salto and R. Gallinas, the river becomes R. de los Naranjos; high-reliefkarst basin, but relatively low gradient in broad valley from El Salto to Micos

Santa Rosa at Micos the R. de los Naranjos crosseds an anticline and descends to the broad, low-relief Valles valleywhere tributaries are developed on impermeable rocks; below R. Mesillas, the river becomes R. Valles

El Choy very small surface basin; nearly all the runoff comes from the large Choy spring; joins R. Tampaón

Santa Clara small surface basin; contains a spring group (3) that has third largest discharge from Sierra de El Abra

Mante small surface basin of shale; Mante spring, largest in S. de El Abra, occurs along western margin; asmall control dam creates a small reservoir and divides the base flow into two irrigation canals;elevation of water surface in reservoir is about 80 m; before it was dammed, natural elevation wasperhaps 1 to 3 m lower; excess flow not taken by canals is meansured by wading near dam;flood flow measured at bridge 4 km downstream

Canal Oeste uses about one half of base flow of the Mante spring for irrigation

Canal Este uses about one half of base flow of the Mante spring for irrigation

La Servilleta about one half of basin is high-relief karst; other half is composed of broad, low-elevation valleyson impermeable rocks

Río Frío gauges combines flow of R. Comandante and R. Frío; contains many springs closely spaced and designatedthe “Frío spring group”; they drain the massive S. de Guatemala karst area; some irrigation losses

Ahualulco gauges flow from the large Poza Azul spring; very little surface runoff; irrigation losses

Salida Conducto Cerrado takes base flow of Poza Azul under R. Nacimiento by tunnel for irrigation; undesired water goes toR. Nacimiento

Ahualulco gauges flow from several springs that form the R. Nacimiento; no irrigation losses; occasional small gain

Ahualulco station is 600 m below junction of R. Nacimiento with R. Poza Azul where it becomes R. Frío;measures most if not all of discharge of “Frí spring group”; no man-made losses or gains

Canal Bajo located 800 m downstream of old Ahualulco (on R. Frío) station and 400 m below the controlreservoir on R. Frío

Puente “T” old gauging station on Canal Bajo; located 7.4 km downstream of the control reservoir

Sabinas high-relief, mostly karstic basin; some non-karst rocks along eastern margin; dry-season flow comes fromNacimiento del R. Sabinas, but there must be several (many?) seasonal springs and some surface runoff

La Encantada gauges flow of R. Guayalejo as it leaves mountains; complex, high-relief basin, but includes the broadJaumave valley; various rock types; possibly springs

San Gabriel gauges R. Guayelejo downstream of La Encantada; one of oldest stations in the region

51


partially maintained by water stored dur-ing 1967. The importance of stored watermay be seen by comparing the 1965 and1971 years, which had nearly equal precipi-tation and reasonably similar numbers ofconcentrated rainfall events.

At the bottom of Figure 4.4, the mini-mum discharge that occurred in March ofeach year is plotted. This is an excellentindicator of the quantity of stored ground-water available, and it follows perfectly instep with the previous season’s rainfall-runoff trend.

4.2.2. Seasonal pattern

One of the two dominant features of thetemporal distribution of rainfall and runoffin the region is its strongly seasonal pat-tern. The year is normally discussed interms of wet and dry seasons instead of ourfour temperature-oriented divisions. Thesecond dominant feature is the rapid anddistinct response of the karst ground-waterflow systems to large precipitation events.

Figure 4.5 compares the average month-ly rainfall and discharge for the years

Table 4.2Selected climatology stations

in the region

# in Avg. annualStation name Fig. 4.3 rainfall (mm)

El Choy 1 1124

Santa Rosa 2 1172(Cd. Valles)

Tanuín 3 888

El Pujal 4 1210

Ballesmí 5 1383

Penitas 6 1064

El Tigre 7 1048

Micos 8 1468

Antiguo Morelos 9 1223

El Refugio 10 —

San Felipe 11 939

Cd. Mante 12 921

Hac. Santa Elena 13 949

Bellavista 14 943

La Servilleta 15 1099

La Boquilla 16 960

Ocampo 17 1383

Ahualulco 18 1585

Joya de Salas 19 801

Jaumave 20 519

El Salto del Agua 21 1503

Cd. del Maíz 22 608

Agua Buena 23 1846

Presa Palomas 24 504

Río Verde 25 488

Requetemú 26 2025

Aquismón 27 2057

Tamapatz 28 (2124) Figure 4.4. Annual flow of Choy spring compared with rainfall for 1960–1971. Also shown is the minimum instantaneous flow in March of each year.

the year 1964 had the lowest runoff and byfar the smallest annual flood (only 12 m3/sec)on record at El Choy, because no largesustained storms occurred and the ground-water reservoir was very low from the pre-ceding years of below average rainfall. Thehighest annual runoff and the greatest flood,143 m3/sec, occurred in 1967 when, after awet 1966 season, about 800 mm of rain fellin three storms during a five week periodin August and September. Though 1968precipitation was above normal, it was dis-persed in many small storms, and flow was

52


Table 4.3Longer-term rainfall compared with base-period averages

Period Number Average annual 1961–68 % greater or lessClimatology station of record of years rainfall (mm) average (mm) than 1961–1968

Ahualulco 1949–73 25 1581.9 1585.0 –0.2

Hacienda Santa Elena 1923–70 48 1023.0 949.1 +7.8

C. E. I. Mante 1952–73 22 1020.3 920.9 +10.8

Antiguo Morelos 1927–30 31 1127.8 1222.5 –7.71942–68

Requetamú 1954–73 20 2128.2 2024.8 +5.1

Ballesmí 1957–73 17 1483.6 1324.3 +12.0

Agua Buena 1942–73 32 1788.3 1846.5 –3.2

Table 4.4Monthly distribution of gulf coastal

zone hurricanes, 1921–1961

January–April 0

May 7

June 16

July 16

August 27

September 53

October 16

November 4

December 3

Table 4.51951 rainfall (mm)

Station May June July Aug. Sep. Oct. Nov. Year total

Ahualulco, Tamps. 181 171 60 514 266 37 15 1311

Hac. Santa Elena, Tamps. 63 140 127 402 280 66 5 1103

Antiguo Morelos, Tamps. 148 139 270 530 270 67 4 1448

Peñitas, SLP 87 245 124 261 180 173 11 1126

Presa Palomas, SLP 19 90 141 360 172 30 0.1 853

1960–1971 at the Choy station. The rain-fall pattern at Cd. Valles on the other sideof the range is similar to that at El Choy.Nominal showers may occur during themonths December to April, but often oneor more months pass with no precipitation—hence the high coefficient of variation. Maybrings an increase in shower activity. Thefirst intense storms of the rainy season usu-ally come in June (average 200 mm or 8inches), but are sometimes delayed untilJuly or August. September nearly alwaysreceives heavy storms, and occasionallyearly October does also. On the average,the wet season begins in early June and thedry season begins about mid-October toearly November.

The discharge pattern resembles thegross shape of the rainfall curve. The Choymaintains a strong, steady base flow aswater is supplied from karst storage at agradually diminishing rate until June. Junethrough October are the high dischargemonths. These months also have the high-est coefficient of variation of discharge,because the occurrence of heavy storms and

the arrival of the wet season varies con-siderably from one season to the next.Though July and October are relativelydeficient in rainfall, discharge is maintainedby ground water stored during the previ-ous wet month, and probably by increasedinfiltration efficiency and responsivenessof the aquifer.

The mean monthly discharge and coef-ficient of variation are shown for the RíoCoy and Río Tampaón (El Pujal station)for the same 12 year period in Figure 4.6.The Coy exhibits an almost identical pat-tern to that of the Choy spring, the onlydifferences being lower coefficients ofvariation of the Río Coy because of the verystrong base flow from the spring, and a ten-dency for rainfall-runoff events to beginearlier in the Coy basin. The graphs of theRío Tampaón represent the average runoffpattern for a very large portion of the re-gion. Floods are clearly a much more im-portant part of the annual runoff of the RíoTampaón than than the Choy or the Coy.

4.2.3. Short-duration events

The second time varying phenomena ofinterest in this thesis is the effect of majorstorms and hurricanes on the hydrology.These storms may occur as a single large

daily rainfall, but normally precipitation isdistributed over several days to two weeks,with several large downpours during thatperiod. More than 300 mm of rain may fallon the El Abra range in a few days, andover 600 mm on the adjacent higher ranges.The monthly distribution of hurricanes thathave affected some part of the Gulf coastalzone of Mexico in the period 1921–1961is given in Table 4.4 (data from S.R.H.Boletín 19). It bears a strong resemblanceto the regional runoff pattern as shown bythe Río Tampaón (Figure 4.6).

Examples of the response of variousbasins to storms are shown in the next sev-eral figures. The Choy and Coy hydro-graphs along with rainfall data are givenfor January 1971–June 1973 (S.R.H. un-published data), the period of field work inthis study, though the largest floods in theirrecords did not occur during that time in-terval. Daily averages have been calculatedby the S.R.H. for the Río Coy for 1972 and1973, but the data have not yet reached theyearly summary stage. The instantaneousflow hydrograph for El Choy has been de-rived from S.R.H. stage and discharge mea-surements and rating curves. As a wholethe year 1971 had slightly below averageprecipitation in the southern Sierra de ElAbra and slightly above average in the Coy

53


area, the northern Sierra de El Abra, andthe Sierra de Guatemala. In 1972, the rain-fall was 15 to 30% greater than average atmost stations, but was about normal for ElChoy. The daily precipitation at the clima-tology stations is collected at 8:00 A.M., sothe measurements shown in the graphs areplotted at 8:00 P.M. of the 24 hour periodthey represent.

As may be seen in Figure 4.7, there areonly a few large precipitation-runoff eventseach year. At the springs, a response oc-curs within 24 hours (thought to be muchless), and the flood peaks within 48 hoursafter heavy precipitation. Because the timeinterval of rainfall measurement is aboutthe same as or larger than the aquifer re-sponse time, it is difficult to tell what thetime lag really is. Furthermore, when theChoy hydrograph and accompanying rain-fall data are examined closely, it is evidentthat the daily precipitation values of thestations are often a poor indication of theamount that must have fallen on the Sierrade El Abra. This demonstrates the consid-erable spatial variation in rainfall that isalso apparent in the climatology stationdata. It is also noted that, quite often, earlywet season showers or heavy rains (nearthe climatology stations, anyway) do notproduce floods comparable to similar rain-falls later in the year. Probably, a large per-centage of the early rains is captured byvegetation or dissipated as replenishmentof ground-water storage. Nevertheless, itis clear that large precipitation events causefloods at the karst springs, and that the aqui-fer response time is very rapid. For stormsthat are reasonably widespread over therecharge area, it seems doubtless that a one-to-one correspondence of rainfall todischarge events could be established. Thisis in marked contrast with big springs ofFlorida (G. Ferguson, et al., 1947), whichhave a broader response and have ratios ofmaximum to minimum flow of less than 2.

There are many flood events which de-serve attention, but only a few will be con-sidered here to illustrate the nature of thehydrographs of the other basins andsprings. The effect of Hurricane Inez inOctober 1966 on several basins is shownby the hydrographs published in BoletínHidrológico No. 32 of the S.R.H. and re-produced in Figure 4.8. The Río Guayalejoabove La Encantada, the Río Comandante,and the Río Valles receive large portionsof their runoff from shale valley catch-ments, whereas the Río El Salto and RíoFrío are predominantly spring fed. The flowat Micos has been drawn with the flow at

Santa Rosa (Río Valles) and the flow at LaServilleta with that at the Río Frío station(where the Río Comandante and the flowfrom the Frío springs join) so that the com-ponents may be compared. Relatively littleprecipitation had occurred in the previousmonth; nearly all of the base flow of theRío Frío and at least two thirds of the RíoValles was supplied from springs. On Oc-tober 10, rainfall began in the late morning

and became very intense in the afternoonand evening over the northern and the cen-tral portion of the region (from S.R.H. de-tailed observations in Boletín 32). Nearly200 mm fell in about 15 hours at many sta-tions. Micos recorded 300 mm, the largestamount. Runoff from the shale catchmentsbegan to reach the gauging stations withinhours of the storm’s initiation. The verysharp pulses of the Río Guayalejo and from

Figure 4.5. Monthly average, standard deviation, and coefficient of variationof discharge and precipitation at El Choy station for 1960–1971.

54


the Valles valley on top of the flow fromthe Río El Salto are especially notable.Additional smaller rains fell in the morn-ings of the 11th and 12th on the Valles val-ley and the Sierra de El Abra and aroundmidnight on the 14th over most of the east-ern part of the region. These later stormsexceeded 100 mm in some places, and theireffects may be seen either as distinct peaksor as bumps on the hydrographs. The dailyaverage flow hydrograph of the Frío springs(sum of Río Nacimiento + Río Poza Azul+ Canal Alto) shows that it too has a veryrapid response, and it peaks 18 to 24 hours

after intense rainfalls; however, it recedesfrom the peak much more slowly than thehydrographs of the shale catchments. Com-parison of the Río Valles and Río Guayalejowith the Frío springs and El Choy springsshow that the flood pulses from the sur-face basins have a narrower width and thepeak is reached more quickly than at thesprings. The Santa Rosa and La Encantadagauging stations measured the highestfloods in their years of record during thisOctober 1966 event. Most other basinsyielded a moderate response, but producedgreater floods in June of that year, and in

many other years.Another event of particular interest oc-

curred in August 1951 (flow data fromBoletín 32). Daily precipitation data are notavailable, but the discharge and monthlyrainfall values (Table 4.5) indicate an ordi-nary year until near the end of August. Arelatively large flood of 178 m3/sec oc-curred in late June at the Frío springs, andboth the Frío and Mante springs had a mod-est flood in mid-July (see Figure 4.9 top).Judging from the daily average flow, show-ers began in the third week of August, andthen a large part of the monthly precipita-tion (by no means are they extraordinarymonthly totals) fell on August 22 and 23.The phenomenal response of the springs isthe greatest in their records (Figure 4.9bottom). Some surface catchment is in-cluded in the measurements of the RíoMante gauging station. Based on old topo-graphic maps, the S.R.H. gave 42 km2 asthe area, but using the new 1:50,000 mapsand air photos, a new estimate of 69 km2 isgiven, assuming the canals do not interferewith the runoff. Thus for the Mante, as withthe Coy, runoff from a surface basin mustprovide a significant part of the floodhydrograph. Most of the surface runoffwould have occurred rapidly, as discussedfor the Río Valles. Thus, the average flowof the Mante spring on August 23 is sug-gested to be in the range 120 to 200 m3/sec,and the spring peak flow between 150 and250 m3/sec. Most of the discharge on thesucceeding days is interpreted as springflow recession. The combined flow of theRío Frío springs was even more re-markable, probably the greatest known inthe world. The peak discharge at the stationswas 554 m3/sec on August 24—304 m3/ secfor the Río Nacimiento and 250 m3/ sec forthe Poza Azul. However, because there ap-pears to be considerable additional inputfrom springs just downstream of the gaug-ing stations, the maximum may have beennearer 700 m3/ sec. Since there were no hy-drologic stations in operation in the south-ern part of the Platform in 1951, the effectof the event in this area is not known. How-ever, the Río El Salto, flowing from a 900square kilometer karstic basin above the ElSalto station, produced a 1260 m3/sec creston August 23, its largest during 16 years ofmeasurements.

One of the biggest surprises of the re-gional hydrology is the large floods thatcome from springs that have a small baseflow or are seasonal. Access to the smallsprings along the east face of the Sierra deEl Abra is almost impossible during the wet

Figure 4.6. Monthly average and coefficient of variation of discharge of RíoCoy (Estación Ballesmí) and Río Tampaón (Estación El Pujal) for 1960–1971.

55


season, so that little is known of their flowbehaviour at this time of year. However, theS.R.H. established a hydrologic station onthe Río Santa Clara in February 1972.When the data were obtained, they had notyet processed their measurements into dailyflows. Hence a simple rating curve withmore than 150 points was constructed fromS.R.H. measurements. A computer programwas then used to convert the stage read-ings at two-hourly intervals into dischargefor 1972–1973 (Figure 4.10). The dry sea-son flow of the spring is about 0.3 m3/sec,one-tenth that of the Nacimiento del RíoChoy, but the flood flows appear to benearly the equal of El Choy. Rememberingthe sharp pulses that may come from sur-face basins in the area, some of the extremepeakedness may be due to surface runoff

included at the gauging station. However,most of the discharge must come from thespring (actually three springs); i.e., springflow is the primary factor in the hydrographshape.

Another example observed during thisstudy is the Nacimiento del Río Pimienta,which rises from holes in a limestonestream bed about 10 km south of Aquismón.In the summers of 1971 and 1972, the lowflow of a few observations was establishedto be several m3/sec. The high flow mayhave been on the order of 100 m3/sec, in-cluding many additional inputs fromsources at the base of the adjacent scarp.Yet when the spring was observed in Janu-ary of 1972, it had dried up completely!

Thus, the karst springs exhibit extremeresponses to storms. The flood events aresuperimposed on a strong base flow fromstorage at some springs, and at others on asmall or intermittent flow. As would beexpected, the response on the surface basinsis even more peaked. During the recession

period, many of the springs maintain a sig-nificantly elevated flow for some time,whereas the flow from the shale basins andsome springs recedes very rapidly. The RíoTampaón, the principal river in the region,carries the integrated output of a large num-ber of karst springs and surface basins, andthis is reflected in its hydrograph (Figure4.6).

4.2.4. Flood exceedance, major events

To lend some perspective to the climatic-runoff events in the region and to summarizethe magnitudes of floods that various springsor basins experience, a flood-frequencyanalysis was made according to proceduresdescribed by Dalrymple (1960). For thepurposes given above, it was not necessaryto extend the records for the stations hav-ing the shorter records—only the real dataare used. Flood data for a few stations weremade available by the S.R.H. for 1972–1973 in addition to the period of record

Figure 4.8. October 1966 floodevents—Hurricane Inez.

Figure 4.9. Hydrographs of Río Mante and Frío springs in 1951: top monthly,bottom daily August–September.

56

AMC

S Bulletin 14 — C

hapter 4

Figure 4.10. Río Santa Clara hydrograph, June–October for 1972 and 1973.

59


listed in Table 4.1. The highest annual peakdischarges are plotted with their calculatedrecurrence interval in Figure 4.11 (note theten times difference in the discharge scalefor the two groups). In this method, the re-currence interval is defined as the averagetime interval in which a flood of a givenmagnitude will be equalled or exceeded asan annual maximum. Unfortunately, theMante data cannot be used because for mostyears only the maximum annual averagedaily flow is reported, rather than the in-stantaneous maximum. The Frío springs arerepresented here by the Río Nacimiento.Inspection of their records shows that theRío Nacimiento and the Río Poza Azul peakalmost simultaneously, but in six of twenty-three years the calculated maximum annualdischarge of each occurred during differ-ent flood events. The peak for the Fríosprings (the combined flow) may be ap-proximated as 1.75 times the value shownfor the Río Nacimiento.

The pertinent points may be summarizedas follows:

1. Spectacular floods from the manysprings in the region are commonplace andcompare with the largest reported in the lit-erature. This is particularly remarkable be-cause the outflows are mostly derived froma myriad of small infiltration points ratherthan from the sinks of major rivers.

2. Many of the stations display erraticpatterns. In particular, there are greatdiscontinuities in the exceedance flow thatoccur in the interval 1.1–2.3 years. Thisappears to be caused by long term cyclesof wet and dry years that may be partiallyself-dependent, by inhomogenous drainagebasins comprising both surface and complexunderground watersheds, and by possibleinhomogeneity of the causative climaticevents.

3. According to Dalrymple (1960),streams in large regions which are homog-enous with respect to flood-producingproperties will have frequency curves ofabout equal slope independent of the sizeof drainage area. The slopes of the curvesin Figure 4.11 are clearly widely disparate.The Frío and Choy springs have the lowestslopes, while the basins with mixed karstand surface drainage have higher slopes.This suggests that the apparent inhomoge-neity may be caused by differences betweensurface channel and cave conduit hydrau-lics.

4. The 1955 flood was an exceptionalregional event in the relatively short his-tory of gauging in the area. It appears to

have produced record floods over most ofthe region, except at the Mante and Fríostations. The Río Tampaón yielded a “l in300 year” flood at El Pujal, and part of theLa Encantada installation was destroyed.The 1951 event also may have been a ma-jor regional event, but the documentationis insufficient. It may have been more simi-lar to other events, such as the October 1966floods that were (are) felt at many placesin the region, but that produced a verystrong response at only a few localities.

5. No abnormally large flood occurredat the Choy spring during the period 1960–

1973; however, the rainfall measured at Cd.Valles suggests that the 1955 event mayhave been substantially larger than any inits record.

4.2.5. Flow duration

The data presented so far have indicatedthe stability of spring discharge during thedry season and the variability during thewet season. To quantify the time distribu-tion of discharge, a flow duration graph ofthe daily average flow of the Choy springfor the years 1960–1971 was prepared

Figure 4.11. Magnitude of flooding of springs and rivers as shown by floodexceedance. The basins and possibly the causal precipitation events areinhomogenous.

60


(Figure 4.12). The curve is composed oftwo segments—the gently sloping baseflow regime and the steeply sloping floodregime. Curiously, the 12 year average of5.47 m3/sec falls at the junction of the twosegments. It is equalled or exceeded only13.5% of the time.

The graph confirms that the base flowregime is strongly dominant, that flood flowis short lived, and that about 50% of thetotal volume of water is discharged infloods. The median value (50th percentile)was 3.3 m3/sec, and the flow either equalledor exceeded 56 m3/sec or fell to less than1.63 m3/ sec only 1% of the time. Alsoshown are the curves for 1967, the year withthe highest annual discharge, and 1964,with the lowest annual average. All of theflows less than 1.65 m3/ sec occurred in1965 following several years of drought.

4.3. Spatial Distribution of Runoff,Basin Yield

In order to gain a quantitative knowl-edge of the magnitude of the springs andtheir contribution to the regional runoff,an analysis will be made here of the lowand average flows and the unitary yield ofthe drainage basins. The analysis will alsoestablish some areas that have excess or de-ficient runoff. This will be used in Chapter7, where the regional surface and ground-water systems are discussed.

4.3.1. Low flows

The dry season runoff from the regionis provided almost entirely by springs.Table 4.6 gives the lowest recorded dis-charge at various gauging stations and com-pares the average daily discharge from thebasins on several days during seasons ofunusually low flow. Inspection of the

S.R.H. daily discharge data indicates thatthe May 1, 1963, runoff budget is aboutthe lowest total runoff thus far measured.Thus, the aggregate discharge from thesprings very rarely falls below 60 m3/sec.The lowest flows for the Frío springs andthe Mante springs are not given becausethe water is split between three channels ineach case and the Mante has a small con-trol reservoir as well. Virtually all the baseflow of the Río Valles comes from aboveMicos. It may be subject to control—thevalue given by the S.R.H. as the minimumrecorded certainly appears affected in someway.

Some springs maintain a strong baseflow, while others are strictly seasonal orintermittent. This seems to be controlledpartly by the elevation of the spring. TheCoy spring, 32 m elevation, is probably atthe lowest point in the whole region whereground-water discharge from the limestonecould occur (the only possible exception is

Figure 4.12. Flow duration graph of Choy spring for 1960–1971. Despite theoccurrence of severe floods, base flow predominates.

61


Table 4.6Low flows and low-flow budget

Elevation lowest recorded m3/secStation or basin (m) m3/sec flow date 3/25/63 5/1/63 3/25/65 1/6/72

E. Requetemú 71.1 1.28 Mar. 25, 1964 2.317 1.591 1.928 —

E. Tancuilin — 0.290 July 3, 1960 0.900 0.674 0.850 —

Huichihuayán valley = 71.1 — — 1.417 0.917 1.078 —Requetemú – Tancuilin

E. Ballesmí (Coy) 29.1 13.00 Apr. 15, 1963 15.07 14.08 14.33 19.92

E. El Pujal (Río Tampaón) 28.2 19.60 May 9, 1964, 26.300 21.20 24.100 87.5June 14, 1965

E. Tansabaca 174 4.51 May 25, 1966 7.820 7.020 6.092 30.21

E. Tanlacut 257 0.00 many days in 0.741 0.00 0.814 —1963–64

E. Vigas 897 1.03 June 9, 1964 3.004 2.075 1.804 —

E. Gallinas 280 0.458(?) June 24, 1960 2.077 1.553 2.270 7.2

E. Santa Rosa 72.8 0.34(?) Apr. 24, 1965 4.672 3.276 4.431 21.6

E. Micos 210 0 ? Apr. 23, 1965 4.442 3.495 3.190 13.7

E. El Salto 399.3 5.51 July 1961 6.68 6.000 8.10 —

E. El Choy 32.3 1.53 June 11, 1965 2.023 1.906 1.690 (3.75)

Micos – El Salto 210 — — –2.24 –2.50 –4.91 —

Santa Rosa – Micos 72.8 — — 0.230 –0.219 1.241 7.90

Tansabaca – Tanlacut 174 — — 7.079 7.020 5.278 —

Tampaón upper — — — 14.569 11.849 13.243 59.0

Tampaón lower 28.2 — — 11.731 9.351 10.857 28.5

Tanlacut – Vigas 257 — — –2.263 –2.075 –0.990 —

E. La Servilleta 75 0.00 many months 0.002 0.002 0.022 0.067

Mante spring (3 stations) (80) ? ? 6.685 6.296 ? —

E. Río Frío 53.7 0.605 June 11, 1965 1.779 1.058 1.586 —

Poza Azul (Canal Alto) (90–95) ? — 3.559 2.939 3.735 —

Río Nacimiento (90–95) ? — 1.792 1.330 1.960 —

Canal Bajo — — — 2.588 2.307 2.491 —

Frío springs (76) — — 5.351 4.269 5.695 —

E. Sabinas 88.8 0.120 May 14, 1965 0.391 0.272 0.242 —

La Encantada (310) 0.537 May 3, 1954 3.548 3.145 3.012 —

ungauged springs — — — (2.00) (2.00) (2.00) —

Laguna Media Luna (1000) ? — (4.00) (4.00) (4.00) —

Regional total 67.687 58.761 (63–64)

where the Río Tampaón passes the south-ern end of the Sierra de El Abra). Its lowestinstantaneous flow in 20 years, 13.0 m3/sec,is equal to two thirds the low flow of theentire Río Tampaón at El Pujal, one quar-ter of the low flow of the entire Platform,and at least twice the low flow of any otherspring in the region! The big springs between

Huichihuayán and La Pimienta are at about100 to 120 m elevation and do not yieldsuch a large base flow (measured atEstación Requetemú). In the northern partof the region, the Mante and Frío springsare the principal points of base flow dis-charge. They are at nearly the lowest el-evation of the El Abra limestone available

for discharge in that area, approximately80 and 90 to 95 meters, respectively. Theflow from each probably does not fall be-low 6 m3/sec. The surprisingly large baseflow of the Río Guayalejo at La Encantadamay be derived from springs in the canyonupstream. The relatively small dry seasonflow of the Río Sabinas comes from one

62


spring, the Nacimiento del Río Sabinas, at160 m elevation. The channel characteris-tics above and below the Sabinas spring andthe topographic maps indicate that there isextensive wet season runoff produced far-ther upstream.

The selected daily budgets given inTable 4.6 allow the location of zones ofwater loss or excess discharge (springs).Reference to the basin map (Figure 4.2) willassist the reader. The basin of the RíoTampaón above El Pujal may be dividedinto contributions from the Tampaón up-per basin (Tansabaca + Gallinas + SantaRosa) and the Tampaón lower basin (ElPujal – Tampaón upper). For the three daysof very low flow given, 45% of the baseflow of the Río Tampaón was derived fromthe Tampaón lower basin, which containsonly 1,531 (6.6%) of the 23,373 square kmsof the Tampaón basin. The eastern half ofthe Tampaón lower basin is formed in shaleand is an extension of the Valles valley.Thus, karst springs in the canyons of theRío Tampaón below Tansabaca, along theRío Gallinas below its gauging station, andthe Nacimiento del Río Tanchachín form amajor zone of ground-water discharge in theregion. On January 6, 1972, a day of higherbase flow early in the dry season, the Tam-paón lower basin contributed 28.5 m3/ sec(33%) of the total flow of 87.5 m3/ sec atEl Pujal. On the same day, the discharge ofthe Nacimiento del Río Tanchachín wasmeasured to be 2.207 m3/ sec (special gaug-ings, Table 4.7), and the Nacimiento de AguaClara was estimated to be 1 to 2 m3/sec.Hence, most of the water comes fromsprings still unknown.

Just downstream of the Ahualulco RíoNacimiento station lies the Nacimiento LaFlorida, and other small springs occur inthe Sierra de El Abra north of the RíoComandante. Their discharge can only beestimated by a complicated budget, shownin Table 4.8. The sources of the Río Frioare the Nacimiento Poza Azul (where thebase flow is abstracted into the Canal Altofor irrigation), the Río Nacimiento, and theRío Comandante. The difference betweentheir total and the sum of Estación Frío andCanal Bajo (which is derived from the RíoNacimiento below the gauging station)gives the discharge of the ungauged springs.The low flows are calculated for two dryseason months in 1963 that follow a pe-riod of regional drought. There would beno contribution from soil moisture or sur-face runoff at these times. Because it is notknown how much water is returned by theCanal Bajo to the Río Frío, the 2.5 to 2.7 m3/sec

Table 4.7Special gaugings

DischargeSpring Location Date m3/sec

Santa Clara Sierra de El Abra Jan. 7, 1972 0.379

Tantoán thermal Sierra de El Abra Jan. 7, 1972 0.0142

Arroyo Seco Sierra de El Abra, Jan. 7, 1972 0.0071Rancho Peñon

Huichihuayán near Huichihuayán Jan. 6, 1972 3.70

Tanchachín Sierra de la Colmena Jan. 6, 1972 2.21

San Juanito near Huichihuayán Jan. 6, 1972 2.53(Mujer de Agua)

Río Pimienta bridge to Rancho Huiche June 2, 1972 22.1(San Juanito and Pimienta springs)

Table 4.8Indirect calculation of flow of smaller Frío springs

Discharge (m3/sec)Station Mar. 1963 Apr. 1963 1966 year 1961–1968

Río Frío 1.905 1.430 48.863 27.68

Canal Bajo 2.645 2.365 2.70 2.54

4.550 3.795 51.563 30.22

La Servilleta 0.002 0.002 11.49 4.95

Río Nacimiento 1.844 1.256 18.62 12.25

Poza Azul 0.0 0.0 9.31 6.10

1.846 1.258 39.42 23.30

Q of other springs* 2.704 2.537 12.14† 6.92†

* Difference of the two sums above† Includes flow of other springs and surface basin

figures are the maximum possible duringthose months. The dry season minimumvalue is 1.1 m3/sec, estimated by subtract-ing the Río Nacimiento flow from the Ca-nal Bajo flow and assuming that the RíoFrío is supplied entirely by return flow fromthe Canal Bajo. The true value probably iscloser to the maximum possible.

Two areas of considerable water lossesare easily determined from Table 4.6. Theseare herein termed the Rayón basin (Tanlacut– Vigas) and the El Naranjo basin (Micos– El Salto). The El Naranjo basin and theeastern portion of the Rayón basin are in-tensively karstified. Water losses occurthroughout most of the dry season. How-ever, it is not known whether all the lossesare due to piracy underground, or in part to

anthropogenic causes such as irrigation orthe El Salto hydroelectric plant. The Rayóndeficit is most likely due to karst piraciesalong the Río Verde.

If there are losses along the Río SantaMaría, they must occur above the EstaciónTansabaca (above the Tampaón lower ba-sin). Unfortunately, there are no helpfulgauging stations. The strong base flow sug-gests that there is no loss. In fact, most ofthe flow is probably derived from karstsprings in the thirty kilometer section ofthe river immediately above EstaciónTansabaca and in the karst area west ofJalpan. Other areas where losses or gainsmight occur are the canyons at Micos, ElPuente near Ocampo, La Servilleta, andLa Encantada. Field observations of the

63


Table 4.9Runoff from basins, 1961–1968

Gauge or basin Elevation Area Average discharge Q for each year, m3/sec Qavg unit Ubasin/(E. = Estación) m km2 1961 1962 1963 1964 1965 1966 1967 1968 m3/sec σ σ/Qavg runoff* Uregion %†

E. Requetemú 71.1 661 42.03 28.96 24.45 20.29 31.89 43.57 44.01 47.71 35.36 10.27 0.29 0.0535 6.32 12.50

E. Tancuilín ? 321 12.13 7.12 6.92 5.58 10.04 14.05 19.66 17.39 11.61 5.15 0.44 0.0362 4.28 4.10

Huichihuayán = 71.1 340 29.9 21.84 17.53 14.71 21.85 29.52 24.35 30.32 23.75 5.88 0.25 0.0699 8.26 8.39Requetemú – Tancuilín

E. Ballesmí (Coy) 29.1 194 27.76 22.05 21.18 21.44 25.36 32.16 32.41 30.24 26.575 4.74 0.18 0.1370 16.19 9.39

E. Choy 32.2 4.9 6.07 3.36 3.62 2.21 3.98 7.36 8.53 5.43 5.07 2.16 0.43 1.035 122.30 1.79

E. Micos 210 1978 24.64 14.45 18.34 16.64 19.62 33.16 34.40 27.29 23.57 7.54 0.32 0.01192 1.41 8.33

E. Santa Rosa 72.8 3521 36.12 15.81 23.73 18.89 25.71 59.37 46.72 36.72 32.88 14.84 0.45 0.00934 1.10 11.62

Valles valley = 72.8 1543 11.48 1.36 5.39 2.25 6.09 26.21 12.32 9.43 9.32 7.91 0.85 0.00604 0.714 3.29Santa Rosa – Micos

E. Gallinas 280 789 38.08 13.87 14.74 22.63 30.82 45.49 42.02 32.50 30.02 11.98 0.40 0.0380 4.49 10.61

E. Tansabaca 174 17532 34.98 19.64 19.21 13.34 16.71 31.07 66.33 46.80 31.01 18.12 0.58 0.00177 0.209 10.96

E. Tanlacut 257 6039 7.21 3.06 2.23 2.51 3.36 9.31 22.11 10.80 7.57 6.73 0.89 0.00125 0.148 2.68

Santa María = 174 11493 27.77 16.58 16.98 10.83 13.35 21.76 44.22 36.00 23.44 11.72 0.50 0.00204 0.241 8.28Tansabaca – Tanlacut

E. Vigas 897 3571 4.72 3.36 3.08 2.34 2.18 5.68 12.56 6.73 5.08 3.42 0.67 0.00142 0.168 1.80

Rayon basin = 257 2468 2.49 0 0 0.17 1.18 3.63 9.55 4.07 2.64 3.23 1.22 0.00107 0.126 0.95Tanlacut – Vigas (–0.30) (–0.85) (2.49) — — (0.00101) (0.119) (0.88)

E. El Pujal 28.2 23373 162.04 69.54 79.19 76.02 105.39 186.97 227.64 169.34 134.52 59.70 0.44 0.0058 0.685 47.54

Tampaón upper = — 21842 109.18 49.32 57.68 54.86 73.24 135.93 155.07 116.02 93.91 40.50 0.43 0.00430 0.508 33.19E. Tansabaca + E. Santa Rosa + E. Gallinas

Tampaón lower = 28.2 1531 52.86 20.22 21.51 21.16 32.15 51.04 72.57 53.32 40.60 19.54 0.48 0.0265 3.13 14.35E. El Pujal – Tampaón upper

E. La Servilleta 75 2532 4.85 2.77 1.85 1.36 5.17 11.49 7.30 4.81 4.95 3.29 0.66 0.00195 0.230 1.75

Mante ? 69 12.39 (8.4) 10.57 (10.1) (12.0) (14.0) 13.23 13.47 (11.77) (1.93) (0.16) (0.171) 20.21 4.16

Frío springs — 42 19.77 17.00 19.45 21.57 22.99 32.08 21.84 21.76 22.06 4.46 0.20 0.5252 62.06 7.80

E. Sabinas 89.8 497 16.27 11.33 8.27 10.38 12.35 25.69 15.96 11.31 13.945 5.46 0.39 0.0281 3.32 4.93

E. La Encantada est. 310 3725 9.90 9.90 6.20 6.13 5.79 17.80 20.95 10.36 10.88 5.63 0.52 0.00292 0.345 3.85

E. Río Frío 53.7 2785 25.66 18.10 19.87 21.59 29.24 48.86 30.87 27.22 27.68 9.676 0.35 0.00994 1.17 9.78

E Canal Bajo — — 1.76 2.38 2.81 2.71 2.66 2.70 2.63 2.69 2.54 0.34 0.13 — — 0.90

E. Canal Alto — — 3.02 3.31 3.72 4.19 3.67 4.15 3.79 3.82 3.71 0.39 0.11 — — 1.31

E. El Salto 399.3 900 20.77 14.22 16.81 14.53 — — — — — — — — —

El Naranjo basin = 210 1078 3.87 0.23 1.53 2.11 — — — — — — — — —E. Micos – E. El Salto

* Q average / Area

† (Q average / Q regional) × 100

canyons need to be made during the fullrange of discharge.

4.3.2. Average flows

To assess the spatial distribution of to-tal runoff from the Platform, the eight yearperiod 1961–1968 was selected for analysis.

The average flow for each year and thedrainage area are given in Table 4.9 for allthe major gauging stations on the Platform.Unfortunately, the Mante record is incom-plete. The record was “completed” by plot-ting the mean annual discharge for nineyears against the average of four nearbyrainfall stations, estimating the best fit line,

and then reading the discharge for the miss-ing years. A small subjective adjustmentwas made if the previous seasons’ rainfallwas abnormally high or low. Also listed arevarious subbasins obtained by subtractionor addition of appropriate station values.For example, the runoff from the Vallesvalley is obtained by subtracting the Micos

64


values from the Santa Rosa values. Exceptfor the Río Santa María above Tansabaca,which needs more stations, a reasonablydetailed analysis is possible. It is immedi-ately evident (as would be obvious fromthe rainfall map, Figure 4.3, or a casualperusal of the area) that the wet easternbasins provide much more runoff than thedry western basins.

As was done in the section on low flows,Table 4.9 may be checked for unusualflows. Amazingly, the Rayón basin showeda net loss, or negative average discharge,during the years 1962 and 1963, and onlya marginally positive flow in 1964. Theeastern half of this basin and the bed of theRío Verde are thus important zones of karstinfiltration. The El Naranjo basin, whichhas a steady loss during base flow, shows asmall plus in the average flow figures (onlyfour years overlap of data). This basin of1078 square km has a high average pre-cipitation and should have a runoff com-parable to that above El Salto. The deficitis believed to be greater than can be ac-counted for by domestic, irrigation, andhydroelectric plant use. Much of the defi-cit is thought to occur by seepage in theriver bed into alluvium where it is lost byevapotranspiration or by continued seep-age down into the limestone aquifer, andby karst subsurface drainage out of the

basin (to the Río Tampaón?) from rechargeinto high relief karst over a large part ofthe basin.

The lower Frío basin, shown previouslyto have a spring or springs providing a sig-nificant base flow, has an eight year aver-age of 6.92 m3/sec. Most of this must beassigned to flow from springs, because thesurface catchment (208 km2) is too smallto account for the excess. Furthermore, re-ferring back to Figure 4.8, the daily budgeton October 14 and October 15, 1966, yieldsan average runoff of 69 and 89 m3/sec, re-spectively, coming from the area below theAhualulco and La Servilleta gauging sta-tions and above the Río Frío station. Nocorrection for lag time has been allowed,but it will not seriously affect the result onthose dates. Also, on each of those days,surface runoff will have diminished to asmall fraction of the water discharged fromthe springs in the lower Frío basin. Thus,these springs and the Santa Clara spring(Figure 4.10), despite their comparativelymodest base flow, produce an importantpart of the karst discharge from the Sierrasde Guatemala and El Abra. The averageflow at Santa Clara station in 1972 was 2.53m3/sec and in 1973 was 5.09 m3/sec. Sev-eral ungauged El Abra springs that haveeither a very small or nil dry season flowwill probably be found to have a significant

mean discharge. The lower Tampaón basinis again identified as a major zone ofground-water discharge. It provided an av-erage of 40.6 m3/sec.

The determination of the total regionalrunoff is shown in Table 4.10. There is nosurface runoff from the semiarid northernquadrant of the Platform. What happens tothe meager quantity of ground-water re-charge is not known. Also, the S.R.H. dis-charge data (Boletín No. 32) and elevationconsiderations indicate that the loss alongthe southern edge of the Platform into theRío Moctezuma could not be more than afew percent of the total regional runoff, andis probably negligible. Significant lossesare not likely along the northern margineither. Runoff from the arid western half ofthe Río Santa María basin could not bemore than a few cubic meters per second,and much of it is used for irrigation. Thus,the runoff from the Platform is closely ap-proximated by adding the flow of the Axtla,the Coy, the Tampaón, and the Choy riv-ers, and the subbasin of the Río Guayalejowithin the Platform. The flow from theungauged springs along the east face of theSierra de El Abra, the Santa Clara springs,and the springs in the lower Frío basin hasbeen set equal to the flow from the Mantespring. Due to the extremely flashy or evenintermittent flow from these springs, their

Table 4.10Annual regional runoff

drainage Year % ofStation area km2 1961 1962 1963 1964 1965 1966 1967 1968 Qavg total*

E. Requetemú 661 42.03 28.96 24.45 20.29 31.89 43.57 44.01 47.71 35.36 12.50

E. Ballesmí 194 27.76 22.05 21.18 21.44 25.36 32.16 32.41 30.24 26.575 9.39

El Pujal 23373 162.04 69.54 79.19 76.02 105.39 186.97 227.64 169.34 134.52 47.54

El Choy 4.9 6.07 3.36 3.62 2.21 3.98 7.36 8.53 5.43 5.07 1.79

Mante 69 12.39 (8.4) 10.57 (10.1) (12.0) (14.0) 13.23 13.47 11.77 4.16

La Servilleta 2532 4.85 2.77 1.85 1.36 5.17 11.49 7.30 4.81 4.95 1.75

Frío springs 42 19.77 17.00 19.85 21.57 22.99 32.08 21.84 21.76 22.06 7.80

Sabinas 497 16.27 11.33 8.27 10.38 12.35 25.69 15.96 11.31 13.945 4.93

La Encantada 3725 9.90 9.90 6.20 6.13 5.79 17.80 20.95 10.36 10.88 3.85

Ungauged El Abra and — (12.39) (8.4) (10.57) (10.10) (12.0) (14.0) (13.23) (13.47) (11.77) 4.16S. de Guatemala springs

Irrigation losses — (6.0) (6.0) (6.0) (6.0) (6.0) (6.0) =(6.0) (6.0) (6.0) 2.12

Subterrean drainage 2335area not included above

Region totals 33433 319.47 187.71 191.75 185.60 242.92 391.12 411.10 333.90 282.93 99.99

* (Q station average / Q regional average) ×100

65


total discharge is likely to be considerablyless than the Mante spring during dry yearsand somewhat more than the Mante duringvery wet years. The average flow might beslightly less than the Mante.

One other factor in the budget is lossesdue to irrigation. The water taken from theMante and Frío rivers has been accountedfor at the source. So far, other irrigationusage is estimated to total 5 to 10 m3/secor only 1.5 to 3% of the runoff. Areas ofirrigation on the Platform include numer-ous small sites along many streams andvalleys, the Río Verde valley, supplied bythe nearly constant flow of the Manantiales(springs) de Media Luna, and the UpperRío Puerco basin in the Valles valley. ThePresa (dam) de Las Lajillas on the RíoPuerco was built sometime during the pe-riod 1961 to 1968 and is capable of storingabout one half of the runoff from its catch-ment of approximately 275 square km.Hence, the flow of the Río Valles will havebecome affected sometime during this pe-riod. The regional average discharge dur-ing 1961–1968 is thus about 283 m3/ sec.

Table 4.11 lists the average discharge ofvarious basins during their complete record.The averages range from 2.0% to 40.0%greater than for the eight year study period.The eastern, karst-spring dominated basinsshow only a small increase, whereas basinsthat have a significant percent of imperme-able cover, such as Tanlacut and Valles val-ley, have a much greater increase. Also, the

variations are due in part to differences inthe period of record. The large increase atEl Pujal reflects the enhanced runoff of theimpermeable basins during wetter years andthe very large 1955 event (Section 4.2). Afew years of the very long record at ElMante could not be used because of uncer-tain data reporting by one of the many agen-cies that have controlled the spring. As withthe rainfall data, these data suggest that theeight year base period may have beenslightly drier than the long term mean.

4.3.3. Basin yield

The average flows given in the previ-ous section are, in reality, the basin yields.However, the data are not directly compa-rable. Table 4.9 includes columns of figureswhich make comparison more tractable.One column gives the yield of each basinexpressed as a percentage of the total re-gional runoff. About one half of the area,but only 11% of the discharge, comes fromthe basin above Estación Tansabaca. Mostof the region of low annual precipitationthat is included within the surface drain-age network is included within this basin.The Coy, the Frío, and the Mante springs(actually includes some surface drainage)produced a remarkable 9.4, 7.8, and 4.2percent, respectively, of the total averagedischarge. The Choy, with an average dis-charge of 5.07 m3/sec, is a first magnitudespring, but this was only 1.8% of the regional

total. Nearly all of the flow from theHuichihuayan and Lower Tampaón basinscomes from springs, and this appears to betrue also for the Gallinas and Micos basins.Thus, probably between 70 and 80 percentof the total regional runoff passes throughthe subsurface karstic systems and is dis-charged at springs.

Further comparison may be made bycalculating the unitary discharge (i.e., theaverage discharge divided by the area) ofeach basin (Table 4.9). During the eightyear period, the calculated values rangedfrom a low of 0.0010 m3/sec/km2 for theRayon basin to a high of 1.035 m3/sec/km2

for the Choy spring. The Rayón basin haspreviously been identified as an area ofsubstantial karst losses. The very low yieldfrom the La Servilleta basin was caused bytwo factors: about 50% of the area is highrelief karst which yields little runoff tothis basin, and in the shale valleys there isinsufficient rainfall in the drier years toproduce an excess of water (available forrunoff) above the requirements of soil in-filtration, surface detention, and evapo-transpiration. The other basins with lowunitary discharges are affected by a dry cli-mate. The northwestern portion of the Plat-form produces no surface runoff becauseno integrated surface drainage has devel-oped, though there is likely a small amountof infiltration into the subsurface. The ba-sins with very high values are influencedby springs which bring in water from out-side the topographic basin. Except for theChoy and the Mante basins, which weremeasured in this study from aerial photo-graphs, the areas listed are those given bythe S.R.H. in Boletín No. 32.

From Table 4.10 the regional unitarydischarge, Uregional, is calculated to be282.93/33433 or 0.00846 m3/sec/km2. Thearea used in this calculation includes theareas on the S.R.H. map which are markedas having total internal drainage plus acrudely estimated 100 km2 of the crest ofthe Sierra de El Abra, which does not ap-pear to be included within the region. An-other column in Table 4.9 shows the ratioof the unitary discharge of each basin tothe unitary discharge of the region. Thisratio makes comparisons of basin yieldeasy. Values greater than about 3 stand outclearly as influenced by karst ground watercaptured from outside the surface basin. Forareas having only moderate rainfall, a lowerratio might still indicate captured groundwater.

Because of the wide range of climaticand geologic conditions which have been

Table 4.11Runoff of full record compared with 1961–68 average

Spring or Qavg σ/ % > 61–68station m3/sec years σ Qavg average

El Pujal 175.66 18 83.99 0.48 30.6

Ballesmí (Coy) 29.28 18 5.70 0.19 10.2

Choy 5.45 12 2.00 0.37 7.5

Mante 12.41 31 3.62 0.29 (5.4)

Frío springs 22.5 20 5.02 0.223 2.0

Est. Ahualulco 30.89 10 6.06 0.196 —(Frío springs)

Valles valley 11.86 11 7.95 0.67 27.3

Micos 27.26 11 9.53 0.35 15.7

El Salto 20.38 16 8.75 0.43 —

Tansabaca 41.52 14 23.17 0.56 33.9

Gallinas 33.17 13 13.36 0.40 10.5

Tanlacut 10.62 11 7.79 0.73 40.2

La Servilleta 6.48 13 4.11 0.64 30.9

66


included in the regional mean unitary dis-charge, attempts were made to determinethe actual yield of some basins having anarrower range of conditions. This is im-possible for most of the basins in the area,because the “underground drainage di-vides” or source areas for the springs arenot known. The best candidate is the Vallesvalley, which is underlain by shale. Therunoff from the Valles valley may be de-rived by subtracting the flow at Micos fromthe flow at Santa Rosa, provided that thereare no gains or losses in the Micos canyon.The 1543 km2 area of the Valles valleygiven by the S.R.H. includes some of theLos Sabinos and Japonés area drainageswhich are pirated underground, the areawhich supplies the Presa de las Lajillas, andprobably too much of the Sierra de laColmena. Thus, 0.00604 m3/sec/km2 is theminimum value for this valley. From thenew 1:50,000 topographic map series ofMexico, the minimum effective area wasmeasured to be 975 km2, assuming that thePresa de las Lajillas retained all the runofffrom its drainage basin throughout the eightyear period. This gives a maximum valueof 0.00955 m3/sec/km2 for the Valles val-ley. As previously stated, it is not knownexactly when the Las Lajillas reservoir be-came effective. Its storage capacity is41.5×106 m3 (S.R.H., Bol. No. 32), whichis one half to two thirds of the expectedaverage annual runoff from its catchmentof 275 km2. Also, there are a few smallranch ponds which impound runoff. Thevalue of the unitary discharge is taken tobe about 0.008 m3/sec/km2 in this valley.For a mean annual precipitation of ap-proximately 1150 mm during the eight yearperiod, and using the estimated unitary dis-charge, the runoff was approximately 22%of the precipitation.

From the data presented, it is obviousthat most of the runoff from the region isproduced in the high rainfall belt in theeastern half of the Platform. A “reducedregion” unitary discharge may be calculatedby subtracting the runoff and area of the LaEncantada and Tansabaca basins and the irri-gation losses from the total regional values:Ureduced region = (282.93–10.88–31.01–6.0) /(33433–3725–17532) = 235.04/12176 =0.0193 m3/sec/km2. The resulting area, the“reduced region,” is 70 to 75% karst terrainwith predominantly subsurface drainage,versus 25 to 30% shale or sediment filledvalleys which produce surface runoff. Itcontains nearly all of the moderate to highrainfall area—the range of measurementswas 801 mm to 2125 mm. Thus, the average

yield of the “effective” runoff producingregion was 0.0193 m3/sec/km2. An upperlimit of 0.03 m3/sec/km2 is suggested forthe average yield of the wettest areas, suchas the Xilitla-Tamapatz area. If groundwater originating from the western portionof the Platform resurges in the “reducedregion,” the effect would be to decrease thevalue given above, though the correctioncould only be a small percentage of the to-tal. The fact that the long term average dis-charges may be greater than the eight yearperiod selected for analysis would demandsomewhat greater values for the long termunitary discharges than those given above.The mean annual precipitation of the “re-duced region” during the eight year studyperiod was calculated to be about 1500 mm.The runoff was 41% and the evapotranspi-ration was 59% of the estimated rainfall,assuming that the change in storage wasnegligible.

Some support for the analysis above maybe obtained from S.R.H. data for the RíoTempoal basin, which drains predominantlyclastic sedimentary rocks along the easternSierra Madres about 100 kilometers south-east of Cd. Valles. It has an area of 5275km2 and lies outside the study area. Thelong term average precipitation for the ba-sin was calculated to be roughly 1650 mm/year (data from S.R.H. Boletín No. 19) andthe unitary discharge for the years 1955–1968 was 0.017 m3/sec/km2 (S.R.H. BoletínNo. 32). The yields are compared in Table4.12.

The calculated unitary discharges for thefour areas are reasonably consistent withone another, but their numbers are few.Quantitative comparisons of the efficiencyof the surface and subsurface routes havenot been accomplished yet. A much greaterpercentage of the rainfall on shale valleysis converted into runoff during wet yearsthan during dry years. This may also be truein the karst areas, but changes in ground-water storage must be taken into account.From the tables of unitary discharge, it is

clear that in this region, as has been foundin many other karst areas, there is little orno relationship between the topographicsurface basins and the subsurface karstdrainage basins. Principal objectives offuture work in the region should be thedetermination of the source areas of thesprings and the detailed study of the karstflow systems.

4.4. Ground Water in theSierra de El Abra

In previous sections of this chapter,some of the gross characteristics of the re-gional hydrology were discussed, i.e., thewater budget and aspects of the temporalvariations of precipitation and of discharge.In this section, some field observations ofthe karst ground water, principally in theSierra de El Abra, will be presented. Fur-ther discussion of the hydrology of the ElAbra will be given in Chapter 7.

4.4.1. Distribution of ground-water,storage

For six to eight months of every year,recharge of the El Abra aquifer in the Si-erra de El Abra is negligible. Thus, the flowof the springs must be maintained byground-water storage.

Observations in the caves yield impor-tant information concerning the distribu-tion of karst ground water. Approximately45 kilometers of cave passage have beenexplored (most of them have been sur-veyed) in the Sierra de El Abra. More thanone-half of this total has been seen by thewriter. Eighty percent of the known passagelength lies within the western margin caves,which are predominantly swallet caves (seeChapter 6).

In the wet season, particularly duringand shortly after a storm, points of vadoseseepage or flow into the caves are very ac-tive. Within a few days following a storm,most of these vadose inlets are reduced toa moderate or slow rate of drip from cracksor stalactites. During the dry season, nearlyall the vadose inlets are inactive. The ac-tive sites are drips, except for a very fewcontinuous flows. There may be some biasin the above observations because so muchof the known cave passage length is alongthe western margin and because the rela-tively impermeable Méndez Formationcovers the majority of the passages of theswallet caves. However, the vadose inletsin the caves in the El Abra range do notseem to be more active than those in the

Table 4.12Unitary discharges of various basins

UBasin m3/sec/km2

Total region 0.00844

Valles valley, minimum 0.00604

Valles valley, maximum 0.00955

“reduced region” 0.0193

Tempoal region (14 year average) 0.017

67


swallet caves. Unless the preponderance ofthe vadose percolation water bypasses thelarger conduits (passages) and reaches thephreatic zone unobserved, then judgingfrom the cave drips the total seepage rateis perhaps two orders of magnitude lessthan the estimated 10 to 15 liters/sec/km2

needed to supply the base flow of El Choy.One of the notable features of the swallet

caves is the numerous lakes and pools. Datacollected during this study (to be presentedlater) indicate that all of the cave lakes thusfar encountered in the Sierra de El Abra,with the exception of the lake in Sótano deSoyate and possibly the terminal sump inCueva Chica, are perched water bodies. Tothe explorer the lakes seem large, and theyoften pose exploration problems. For ex-ample, a lake in Sótano del Tigre is 270 mlong, 15 m wide, and (fortunately for theexplorer) only 1.4 m deep for the most part.Excluding the lakes in spring caves suchas El Choy, all of the known large lakesoccur in the western margin caves, againwith the exception of the lake in Soyate.From maps, estimates were made of thevolume of water in the lakes in the Arroyo,Tinaja, Los Sabinos, Tigre, and Jos caves,which contain 40% of the known passagelength and most of the large lakes. The to-tal calculated volume is 48,000 m3. Thelakes in most of the remaining swallet cavesare generally smaller, and the known poolswithin the El Abra range do not amount tomore than a few tens of cubic meters. Thus,the total volume of the known vadose cavelakes is roughly 80,000 m3. This is enoughto supply the Choy spring for only one thirdof one typical dry season day—if all thewater from the lakes were drained. Cer-tainly most of the interior of the El Abrarange has not been explored. Yet, becausemost of the lakes do not lose much waterduring the dry season and because theirtotal volume is believed to be relativelysmall, it does not seem possible for the ElAbra springs to be supplied by the changein storage of the vadose lakes.

Only four dry season cave streams havebeen found in the El Abra. They are locatedin Cueva de El Pachon, Sistema deMontecillos, Sótano de la Tinaja, andSótano del Arroyo. There is a good chancethat the stream in Arroyo is the same as theone in Tinaja (see descriptions and maps,Chapter 6). Because of the limited numberof observations, it is not known whetherthe streams are really permanent, but inMay 1973, near the end of a severe dry sea-son, a tiny trickle still flowed in Arroyo. Inreality, all of these “streams” are merely

trickles of a few liters or less per minute,depending on the date during the dry sea-son. Thus, they do not constitute a signifi-cant discharge of water.

From the above data, it is believed thatthe total dry season vadose flow that passesthrough cave passages of explorable sizedoes little to support the discharge of theEl Abra springs. If it is true that a signifi-cant portion of the vadose seepage watersare integrated into the major conduits whilestill in the vadose zone, then most of thebase flow of El Choy and El Mante is drawnfrom the phreatic ground water body.

An estimate of the storage supportingthe flow of a spring may be made from ananalysis of the base flow recession. Burdonand Papakis (1963) give the equationsQt = Q0e–Kt and St = Qt / K, which representthe exponential decay of the discharge Qand the storage St of the aquifer at time t; Kis the recession constant. If Q and t are plot-ted on semilog paper, the data will form astraight line where the equations are appli-cable. Inspection of the Choy daily dis-charge data from 1960–1971 indicated thatthe most stable long period of flow occurredduring the unusally severe dry season of1970–1971 (see Figure 4.7). A semilog plot(not shown) of the data yielded an excel-lent straight line fit for the period January1, 1971, to May 1, 1971. Deviations fromthe fitted line were less than the expectederrors of the discharge measurements. Fromthis graph,the recession equation was de-termined to be Qt = Q0e–(0.001643/day) t. Thestorage was calculated to be 200,000,000 m3

on January 1 and 164,000,000 m3 on May1. Thus the volume of phreatic storage be-hind El Choy in the dry season is about thesame as its average annual discharge of172,000,000 m3 (5.45 m3/ sec). The stor-age value given above is the volume ofwater that theoretically could drain to ElChoy due to the force of gravity, and doesnot include the phreatic storage below thespring point.

Similar calculations were not attemptedfor El Mante because of the complicationsof the reservoir and distributary canal sys-tem, but were made for the Coy spring. Aspreviously noted, the hydrograph of the RíoCoy at Ballesmí gauging station is morevariable than El Choy during the winter andspring months because of greater and morefrequent rainfall on the surface basin andon the recharge area. Reasonably long re-cession periods are not available for analy-sis. A straight line can be fitted to the baseflow recession from December 20, 1962 toFebruary 6, 1963. This gives a gravity

storage of 327,000,000 m3 on January 1,1963. The Coy has a greater storage thanEl Choy, but it also has a larger base flow.Calculations for a short period in April1963 and for a 30 day stable period inApril–May 1970 give slightly over400,000,000 m3 of storage. According tothese calculations, early in the dry seasonthe gravity storage for the Coy spring isprobably about one half of the averageannual discharge (18 year record) of923,000,000 m3 measured at the Ballesmístation, which also includes 194 km2 ofsurface drainage. However, the storagegiven here should be considered a mini-mum, because a long recession limb couldnot be examined and because the dry sea-son base flow has so little variation fromyear to year that a larger storage may exist.

4.4.2. Ground-water level fluctuations

Cave lakes exist over a wide range ofelevations within the Sierra de El Abra.Previous workers (Bonet, 1953a, and Wil-liam H. Russell, personal discussions) havebelieved that most of these lakes, particu-larly those which “terminate” passages (i.e.,where the passage becomes filled with wa-ter), lie at the water table. An important partof the hydrological field studies of this the-sis was aimed at testing this hypothesis,mapping the water table, and collectingquantitative data about water level fluctua-tions. This section will discuss field mea-surements of water level fluctuations. Otherdata and interpretations concerning waterlevels are contained in Chapter 7.

Two cave lakes representing contrastingsituations were selected. One was the ter-minal sump in Sótano de Jos (map and de-scription, Chapter 6), deemed the swalletcave in which a stage recorder could mostfeasibly be placed in a satisfactory position,although many of the other swallet cavesare larger and have greater catchment ar-eas. The known part of the cave is about350 meters long, and it has a surface basinof 4.3 square kilometers. A continuous re-cording Ott recorder was bolted to a boardspanning the gap between two ledges 6 mabove the static level of the lake. A quasi-flexible plastic pipe, secured at the bottomby a board bolted to the wall and along itslength by wires to bolts in the cave walls,was emplaced to protect the float and coun-terweight during floods. The deep lake inSótano de Soyate (Chapter 6) was the otherlake selected. It is a deep shaft on the west-ern flank of the Sierra de El Abra. The cavehas a negligible surface catchment, though

68


a small vadose inlet is active during storms.It descends via a deep shaft and two shorterdrops to a large lake room 150 m long and20 m wide. The water depth was measuredin many places, the maximum being 54 mand the minimum 34 m; thus the bottom ofthe lake is near sea level and well belowthe base level of the El Abra range. An Ottwater level recorder was placed over a holeabout 12 m above the general dry seasonlevel of the lake, which gave a free drop tothe lake. The results of these projects areshown in Figures 4.13–4.15. Also shownin these figures, for comparative purposes,are the Choy discharge hydrograph andappropriate precipitation data.

The important features of the Joshydrograph may be seen in Figure 4.13,although these particular floods are verysmall. The hydrograph was perfectly flatfor two days until a storm occurred between8:00 A.M. on July 11 and 8:00 A.M. on July12—probably very early in the morning ofJuly 12. A flood pulse entered the cave,causing a response at the Lake Room be-ginning at 9:25 A.M., and 50 minutes laterthe crest stage of 7 cm was reached. An-other shower on July 12 caused a secondpulse to build on the first, reaching a crestof 15 cm. A third rainfall-flood event oc-curred on July 14, after which the waterlevel rapidly returned towards its normallevel. There seems to be a close relation-ship between the events recorded on the Josand Choy spring hydrographs, although theamount of water which passed through Joscould not have had much effect on theChoy’s discharge. The Jos basin apparentlyreceived less rainfall than some other partsof the Choy’s source area. Also, the flood

Figure 4.13. Soyate and Jos cave-lake hydrographs compared with rainfall and Choy discharge, July and August 1971.

Figure 4.14. Soyate cave-lake hydrograph and correlation with rainfall andChoy discharge, June 1972.

69


hydrograph is substantially broader at theChoy than in Sótano de Jos. An additional4.5 days of record in August was com-pletely featureless and at static level.

The hydrographs obtained in Sótano deSoyate are very different from those ob-tained in Jos. Unfortunately, the Soyaterecord began on July 16, 1971, missing thefloods of the preceeding four days. Agradual recession, totaling 92 cm, occurredover the next 21 days, in harmony with theChoy discharge recession. The water levelfluctuations during the July 12–16 floodsmust have been much larger at Soyate thanat Jos. Three small flood events were re-corded in early August. Because the dis-tance from the stage recorder to the watersurface was not measured for this particu-lar record, only the change in elevation ofthe water surface is known. However, mea-surements at other times suggest that thereference line shown in Figure 4.13 wasabout 11 ± 0.5 m below the support (da-tum) of the stage recorder.

Figure 4.14 shows a shorter, but moreactive hydrograph of Soyate in June 1972.The first showers of the wet season in lateMay and early June were already causingthe water surface to rise when the recorderwas started on June 5. An afternoon rain-fall on the western flank of the El Abra,including the Soyate area, probably causedthe steepening of slope at point A on theChoy and Soyate hydrographs (note thatthis observed shower was scarcely felt atCd. Valles or El Choy). Two to three dayslater, heavy rainfall again caused the waterlevel to rise rapidly. Although no record ofthe crest of the flood was obtained, the stagerecorder was inundated by leakage aroundthe pulley axle bearings. Thus, the waterlevel rose more than 9 m after the recordbegan on June 5, and probably more than12 m from mid-May.

Attempted stage recording operations inSeptember 1971 provided information con-cerning the most spectacular flood eventsfor which some data were obtained. Figure4.15 shows an interpretative sketch of theevents based on the limited data and on thecharacter of the previously discussedhydrographs. Personal communicationsfrom Steven Bittinger and Don Broussard,who along with Frank Binney and DianaPorter went to service the stage recorderson September 11, have provided importantinformation concerning the events. FromJuly 16 through September 8, only a fewvery minor floods occurred at El Choy;none exceeded 10 m3/sec (see Figure 4.7).From the data obtained during the first half

of this long stable period, it is reasonableto assume that the terminal lake in Jos wasat its static level and that the Soyate lakesurface was about 11 m below the datumboard. The ground surface and vegetationwere undoubtedly dampened by severalmoderate rainfalls during early September.On September 8, about 40 mm of precipita-tion fell, which produced a sharp 34.7 m3/secflood crest early on September 9 at El Choy.On September 11, both recorders werefound full of water, presumably havingbeen flooded out by the pulse on Septem-ber 9. In Jos, the flood wave crested lessthan 2.5 m above the recorder, but its forcehad torn loose one end of the board at thebottom of the pipe. The lake surface hadreturned to its static level, and the flowthrough the cave was estimated to beequivalent to that of “a few garden hoses.”In Soyate, the water level appeared to beback down to approximately its normal rangewhen the Choy is at base flow. Fortunately,

the Jos recorder worked for the next thir-teen days. A single storm caused a sharplypeaked flood pulse in Jos and a muchbroader pulse at the Choy on the 17th. Thecrest of the flood of slightly more than 1 mwas reached 2 hours after the stage beganto rise and approximately 7 hours beforethe crest at El Choy. Then it subsided al-most as rapidly.

At the end of the month, a major stormhit the southern El Abra region. When therecorders were removed in January 1972,both were found to have been disastrouslyinundated. This must have occurred onSeptember 29–30, and possibly again dur-ing the second week of October (see Fig-ure 4.7 for the complete Choy hydrograph).More damage had been done to the still pipein Jos, and water in a can placed 2.5 mabove the recorder demonstrated that thecrest had reached at least that level. InSoyate, the water rose so high and soquickly above the recorder that the water

Figure 4.15. Flood events of Jos, Soyate, and Choy, September 1971

70


pressure popped the glass window out ofits stiff rubber frame and into the case. Inearly January 1972, the water level was11.5 m below the Soyate recorder board,and the Jos lake was at its normal staticlevel.

Inspection of all the above Figuresshows a close temporal relationship ofevents recorded at Jos, Soyate and El Choy.Furthermore, each event is in response to aparticular precipitation event. A one-to-onecorrelation is attempted in Figure 4.14, butsuch correlations are more easily seen whenthe storms are isolated events. It is also clearthat the amount of precipitation receivedat El Choy or at Cd. Valles (Santa Rosa sta-tion), both of which lie off the El Abra rangeand peripheral to the source area of theChoy spring, is often not a good predictorof the magnitude of response that is ob-served at Jos, Soyate, or El Choy.

Although year-round observations in Josare not available, it appears that the watersurface at the Lake Room siphon stays at aconstant static level except when wet sea-son flood pulses of very short durationenter the cave. Possibly, a few centimetersare lost because of evaporation or leakageduring the dry season. The hydrograph issimilar to a desert arroyo which occasion-ally has flash floods, but maintains only atrickle or no base flow at all. The characterof the hydrograph combined with elevationand geomorphic data to be discussed laterlead to the conclusion that the lake is aperched water body. It is predicted that

more air filled passage exists on the otherside of a water trap (inverted siphon).

In contrast with the Jos lake hydrograph,the water surface in Soyate is always mov-ing—up when it rains, and down when itdoes not. These movements occur in almostperfect temporal correlation with the Choyhydrograph. Note in particular the similartime span (breadth) of the flood hydro-graphs of Soyate and El Choy, the correla-tions of Figure 4.14 and during more subtlefluctuations in August 1971, and the cor-related break in slope on July 17, 1971,followed by an increasingly gradual de-cline. Figure 4.16 is an attempted compari-son of the stage at Soyate with the dischargeat El Choy. Although based on sparse andsometimes shakey correlations, a broad re-lationship is evident. The letters are keyedto the hydrographs. Vertical lines representthe three times the level of the recorder wassurpassed. Event Y probably only slightlyexceeded the height of the board; floodssuch as Z likely cause the water surface torise 20 m or more.

Based on the hydrographs, the lake’selevation of approximately 59 m (or 24 mabove the spring at base flow), and the lackof known air filled passages below that el-evation (except at inappropriate places suchas near the Río Tampaón) the Soyate lakeis interpreted to be a “water table” lake. Thecave in reality is a rather large piezometer;the elevation of the water surface reflectsthe local piezometric surface of the pres-sure conduits (water filled caves) which

carry water to the Choy spring.To date, the Soyate lake is believed to

be the only “water table” lake known in theSierra de El Abra. This is unfortunate, be-cause Soyate may connect with the largeconduit(s) coming from the swallet cavesto the west to El Choy. This is suggestedby the thick mud deposits at and above therecorder, by the large population of blindfish, and by their need for organic debrisfor food. The smaller fluctuations in Soyateare probably the result of local infiltration,but the larger fluctuations may be influ-enced by sharp pulses from swallet caves.Discovery of more lakes such as that inSoyate will be very important in achievinga better understanding of the aquifer.

4.4.3. Water tracing

A large number of water tracing experi-ments are needed in the El Abra range todetermine sink to rising connections, drain-age divides, ground water velocities, andpossible dispersion of the tracer into sev-eral flow paths and to more than one spring.In practice, such experimentation in thisregion is difficult because very few of thesinking arroyos can be observed from thehighway to know when they are running.Furthermore, when large rains come, ac-cess to the sink points involves walkingseveral kilometers on muddy tracks, andaccess to most of the smaller springs is gen-erally impractical.

It was decided to conduct a lycopodiumspore tracing test of Sótano de Japonés (seeChapter 6) to its most likely discharge pointthe Choy spring. Nets of 25 micron meshnylon were made approximately in the man-ner described by Drew and Smith (1969),though somewhat smaller, and five wereplaced in the Choy stream about 100 mdownstream from the spring on July 15,1971. They were attached to a wire andfloated in mid-stream with about 90% oftheir catchment area below the water sur-face. Twenty-four pounds of dyed sporeswere injected at 6:15 P.M. on July 18 intothe arroyo leading to Japonés where itcrosses the highway 1.6 km upstream of thesink. Local showers during the day hadcaused a small flood. The discharge wasestimated to be about 0.3 m3/sec at the timeof injection.

The nets were sampled on July 22 and28 and August 5 and 14. A very few sporeswere found in the first sample and none inthe others. Such a low catch might be ex-plained by the fact that only 2 to 3% of thestream’s cross section was sampled and byFigure 4.16. Relationship between Soyate cave-lake stage and Choy discharge.

71


the large amount of sediment caught, whichtends to block flow through the nets.Unfortunately, the possibility of contami-nation cannot be completely ruled out;hence, the test is only tentatively positive.The straight-line distance from sink tospring is 22.9 km and the sample time was92 hours after injection. If the test is posi-tive, an average ground water velocity ofat least 250 m/hr is calculated. This valueappears reasonable when compared with alist of velocities determined for karst flowsystems having a wide range of relief anddistance (Prof. D. C. Ford, personal com-munication of unpublished list derived fromthe literature and his own data).

4.5. Summary

The El Abra and Tamasopo Formationsplus suspected Lower Cretaceous platformfacies limestones form a karstic aquifer thatis continuous over most of the Valles–SanLuis Potosí Platform. Other rocks have lowpermeability or are impermeable and yieldsurface runoff. The El Abra limestone(geologic mapping has generally not dis-tinguished between the El Abra and theTamasopo Formations) crops out over alarge portion of the Platform, and karstcaves and major springs have developed inthis high relief region. The S.R.H. (Secre-taría de Recursos Hidráulicos) has estab-lished many hydrologic gauging andclimatologic stations in the region to col-lect data for water resources control anddevelopment. A surface drainage networkreaches about two-thirds of the Platform,but no surface runoff leaves the northwest-ern area. Rainfall is strongly temporallydistributed into wet and dry season (June–October and November–May, respectively)and spatially distributed into a wet easternzone (800 to 2500 mm/year) and a semi-arid western zone (250 to 600 mm/year).Very large precipitation events are commonduring the wet season.

Runoff from Platform basins closely

correlates with annual precipitation, withthe season (wet or dry), and with individualprecipitation events. This statement is truefor springs as well as surface catchmentson impermeable rocks such as the Vallesvalley. There are several very large karstsprings in the region—two are among thelargest reported in the literature. Springsprovide essentially all the dry season baseflow from the Platform; and 70 to 80% ofthe total runoff is estimated to have passedfor at least a short distance through sub-surface karst channels. Most of the largesprings occur at or near the eastern edge ofthe Platform or along the lower part of theRío Tampaón, the largest river in the region.The Coy spring has a minimum recordedflow of 13.0 m3/sec (459 cfs; 18 year pe-riod) and discharges about 20% of the en-tire base flow of the Platform; its long termaverage flow is estimated to be about 24m3/sec (847 cfs), including the nearby smallintermittent springs at the Coy dome. (Theaverage at the Ballesmí gauging station,which includes a surface basin of 194 km2,was 29.28 m3/sec, and there may be othersmall springs near Aquismón). The Fríosprings, a cluster of springs in a small areathat form the headwaters of the Río Frío,have an estimated long term average dis-charge of about 28 m3/sec (988 cfs), includ-ing some ungauged springs below theprincipal spring guages. The Choy and theMante are the largest springs in the Sierrade El Abra and have average flows of 5.4and 11.9 m3/sec, respectively. The karstground-water flow systems and the flowsof springs are extremely dynamic in re-sponse to large wet season precipitationevents. Most springs exhibit dischargevariations (Qmax/Qmin) of 25 to 100 timesor more. In 1951, the Frío springs produceda flood crest of 554 to 700 (?) m3/sec, thegreatest from a karst spring known to thisauthor. This is particularly remarkable be-cause recharge, as for most of the springsin the region, occurs at myriad localitiesrather than a very few discrete points. The

Choy, which was the most intensivelystudied spring in this work, usually yieldsseveral pulses of 30 m3/sec or more eachyear, and occasionally surpasses 100 m3/sec.The Choy (and probably all the othersprings except Media Luna) shows a re-sponse within a few hours of heavy rainfalland usually crests roughly 24 hours after amajor precipitation event; some largersprings may take slightly longer. The S.R.H.discharge data for the Santa Clara and ob-servations in this study show that the manysprings having small base flows are alsoextremely dynamic and yield a large partof the runoff from the region. Studies ofthe spatial variation of discharge, as mea-sured at the gauging stations, have locatedareas of deficient and excess runoff andshowed some differences between the dis-tribution of base flow and flood flow. Theareas having deficient runoff have zones ofkarstic losses, and the areas having excessrunoff have karst springs that receiveground water from areas not related to thesurface basins.

Some observations and field work con-cerning the hydrology of the Sierra de ElAbra are reported in this chapter, but abroader discussion is given in Section 7.3.The water in dry season gravity storagebehind the Choy spring was calculated tobe equivalent to about one average year ofits discharge. Based on observations incaves all across the range, it was concludedthat the water must be stored in the phreaticzone, although some of it might exist asvadose seepage if it bypasses the largerconduits. Stage recorders placed over theterminal siphons in Sótano de Soyate andSótano de Jos (a swallet cave) producedrecords that indicate the former is a genu-ine water-table lake, whereas the latter is aperched lake. It is believed that all of thesiphons yet discovered in the western mar-gin caves are perched siphons. Both record-ers were inundated, showing that largewater level fluctuations often occur.

72


5

HYDROGEOCHEMISTRY

5.1. Introduction

The objectives of the water chemistrystudies in this thesis are (1) to determinethe chemical character of the karst watersin the region, especially the springs, andmake comparisons with the karst waterchemistry of other regions; and (2) to usethe water chemistry as a tool for hydrologicanalysis, particularly hydrograph separa-tion. This chapter will examine the chemi-cal composition of karst waters in variousenvironments and some variations that oc-cur during the wet season. Further analysisand modeling will be given in Chapter 7.

Four classes of karst water have beensampled. They are large streams on shalecatchments which are pirated underground(swallet water), cave drips, cave lakes, andsprings. Figure 5.1 shows the schematicrelationship of these karst waters in the Si-erra de El Abra. The first two classes arerepresented by just a few samples; hencethey serve only as an indication of theircharacter. Many more samples from cavelakes were obtained. Most of these lakesare perched water bodies in the westernmargin swallet caves. However, the sam-pling plan was principally aimed at differ-entiating the types of springs and studyingtheir response to flood pulses. Sampling ofthe first three water classes was confinedto the Sierra de El Abra, whereas samplingof the fourth type included most of themajor and several of the smaller springs inthe region. Three classes of water whichwere not sampled are the main surfacedrainage systems, such as the Río Tampaón,and the diffuse or unconcentrated surfaceand vadose trickle input water on top ofthe El Abra range.

A statement of the sampling procedures,the analytical methods, and the errors isgiven in Appendix 1. Here it is sufficientto say that the analyses of samples collectedduring the summers of 1972 and 1973 areregarded as high quality data (the Ca, Mg,and SO4 listings are the average of at leasttwo determinations, and the temperature, pH,and alkalinity were carefully measured); the

analyses of the summer 1971 and Decem-ber 1971–January 1972 are less reliableprincipally because the pH determinationwas less precise and the Ca, Mg and alk-alinity analyses were generally performedonly once. A few samples simply have un-explained errors; however, most of theanalyses are believed reasonably good. Thedata have been processed on a CDC 6400computer using Watchem II, a programwritten by John Drake and modified by T.M. L. Wigley (1972). The program calcu-lates the saturation conditions of a samplewith respect to calcite, dolomite, and gyp-sum, the theoretical PCO2, and several otherresults such as the percent error, the Ca/Mg and Ca/SO4 ratios, and the ionicstrength. Saturation conditions are calcu-lated using temperature dependent equili-brium expressions and take into account theion pairs CaSO4

0, MgSO40, MgHCO3

+,CaCO3

0, and MgCO30. Only the first three

ion pairs are significant in the area, andCaHCO3

+ is ignored completely (afterLangmuir, 1971). The results agreed within± 0.04 log units of the calculated calciteand dolomite saturation indices given byboth references, and within ± 0.02 for thepPCO2 on three samples by Langmuir. Theslight differences are probably the result ofvariations between authors as to what ionpairs are important and what values of the

equilibrium constants are used, especiallyfor dolomite.

In this study, only the saturation state ofthe water with respect to calcite, dolomite,and gypsum is of interest. The degree ofsaturation may be evaluated by the satura-tion index, or SI, which is calculated asfollows:

calcite:SIc = log(aCa

++·aCO3– –) – log Kc

dolomite:SId = 0.5[log(aCa

++·aMg++·a2

CO3– –) – log Kd]

gypsum:SIg = log(aCa

++·aSO4– –) – log Kg,

where the a’s are activities and the Ks arethe appropriate equilibrium constants. Notethat the saturation index for dolomite (SId)is calculated for the stoichiometric form sothat it may be compared directly with thevalues for calcite and gypsum. A positive(+) number is supersaturated, and a nega-tive (–) number is undersaturated. Sampleshaving a value within ±0.10 of zero areapproximately saturated with respect to thatparticular mineral. More information aboutthermodynamic calculations is available inGarrels and Christ (1965), Stumm andMorgan (1970), Langmuir (1971), andBack and Hanshaw (1970). Because of theimportance of ion pairs in the waters studied

Figure 5.1. Hydrochemical environments in the Sierra de El Abra: (1) swalletwaters from shale or flaggy limestone; (2) cave drips; (3) cave lakes, 3aperched and 3b water-table; and (4) springs.

73


Table 5.1aChemical analyses and derived geochemical measures (best analyses)

Temp Ca* Mg* Alkalinity* SO4* % Ca/Mg Q†

Location No. Date °C pH mg/l meq/l mg/l meq/l mg/l meq/l mg/l meq/l error molar pPCO2 SIc SId SIg m3/sec

Springs

Choy 1 5/21/73 26.4 6.91 204 10.16 44.3 3.64 300 4.91 422 8.79 +0.36 2.79 1.45 +0.02 –0.11 –0.73 2.85

2 6/07/73 26.4 6.92 203 10.14 44.5 3.66 303 4.97 422 8.79 +0.15 2.77 1.46 +0.03 –0.09 –0.73 2.84

3 6/13/72 25.1 6.74 98.1 4.90 5.6 .46 276 4.53 37 .77 +0.56 10.6 1.30 –0.35 –0.78 –1.84 flood

4 6/15/72 25.0 6.69 97.3 4.86 5.6 .46 271 4.44 35 .73 +1.43 10.6 1.26 –0.42 –0.84 –1.86 flood

5 6/17/72 24.9 6.70 97.3 4.86 5.6 .46 273 4.47 41 .85 0.00 10.6 1.27 –0.41 –0.83 –1.80 flood

6 6/19/72 24.9 6.80 98.5 4.92 6.1 .50 267 4.38 44 .92 +1.12 9.83 1.38 –0.31 –0.72 –1.76 flood

7 6/23/72 24.9 6.83 95.3 4.76 5.8 .48 270 4.43 40 .83 –0.19 9.90 1.40 –0.29 –0.70 –1.81 ~6.5

8 6/28/72 25.1 6.87 98.1 4.90 6.6 .54 265 4.34 53 1.10 0.00 9.06 1.45 –0.25 –0.64 –1.69 4.4

9 7/07/72 25.2 6.93 112 5.60 10.7 .68 267 4.37 90 1.87 1.89 6.37 1.51 –0.15 –0.46 –1.44 4.2

Mante 10 5/25/73 26.5 6.88 228 11.38 44.8 3.68 309 5.06 487 10.15 –0.50 3.09 1.41 –0.04 –0.11 –0.65 11.32

11 7/11/72 25.8 6.92 152 7.58 25.5 2.10 291 4.77 240 5.00 –0.46 3.61 1.47 –0.06 –0.24 –1.00 ~23

Coy 12 5/17/73 24.6 7.06 291 14.52 46.7 3.84 253 4.15 693 14.42 –0.57 3.78 1.70 +0.17 –0.04 –0.45 29.1

13 6/04/73 23.9 7.02 262 13.06 40.6 3.34 235 3.85 614 12.78 –0.67 3.91 1.69 +0.06 –0.16 –0.52 17.2

14 6/17/72 21.6 7.10 58.1 2.90 9.2 .76 187 3.06 26 .54 +0.83 3.81 1.86 –0.40 –0.63 –2.15 115

15 6/20/72 21.5 7.11 56.9 2.84 9.0 .74 184 3.02 25 .52 +0.56 3.83 1.87 –0.41 –0.63 –2.17 68.8

16 6/27/72 21.6 7.08 76.1 3.80 10.5 .86 206 3.38 61 1.27 +0.11 4.42 1.80 –0.30 –0.55 –1.71 24.4

17 7/07/72 22.4 7.08 182 9.08 29.2 2.40 232 3.80 366 7.62 +0.26 3.78 1.76 0.00 –0.22 –0.79 21.9

Poza Azul 18 5/25/73 22.6 7.10 125 6.26 17.7 1.46 262 4.30 169 3.52 –0.64 4.29 1.72 –0.01 –0.25 –1.17(Frío) total

Nacimiento 19 5/25/73 22.5 7.09 122 6.08 16.8 1.38 264 4.33 158 3.29 –1.06 4.40 1.70 –0.03 –0.28 –1.21 5.40+(Frío)

Sabinas 20 5/25/73 21.0 7.35 65.3 3.26 11.2 .92 241 3.95 12 .25 –0.24 3.55 2.00 –0.01 –0.22 –2.45 0.3e

Tanchachin 21 6/20/72 24.6 6.79 118 5.88 8.8 .72 251 4.12 111 2.31 +1.30 8.18 1.40 –0.31 –0.68 –1.34 flood

Canoas 22 6/18/72 21.8 6.92 84.9 4.24 5.1 .42 269 4.41 13 .27 –0.21 10.1 1.52 –0.27 –0.70 –2.32 0.4e

Media Luna 23 6/18/72 29.6 6.74 340 16.95 61.8 5.08 304 4.98 839 17.46 –0.92 3.33 1.26 +0.02 –0.12 –0.37 ~4

Taninul sulfur 24 5/17/73 39.0 6.82 137 6.86 21.2 1.74 437 7.17 47 .98 +2.69 3.94 1.06 +0.16 +0.06 –1.74 0.04e

pool

25 6/13/72 40.5 6.32 138 6.88 20.2 1.66 417 6.83 50 1.04 +4.08 4.14 (0.55) (–0.32) (–0.40) (–1.67) 0.05e

26 6/15/72 32.0 6.60 124 6.18 8.3 .68 360 5.90 32 .67 +2.16 9.09 1.01 –0.20 –0.53 –1.86 0.34e

27 6/17/72 31.6 6.62 128 6.38 10.7 .88 389 6.38 36 .75 +0.90 7.25 (0.98) (–0.16) (–0.45) (–1.80) 0.57e

28 6/28/72 36.0 6.76 133 6.62 12.6 1.04 386 6.32 36 .75 +4.01 6.37 (1.08) (+0.04) (–0.18) (–1.80) –—

Tantoan thermal 29 5/18/73 31.0 6.88 97.3 4.86 6.1 .50 297 4.87 15 .31 +1.71 9.71 1.36 –0.11 –0.46 –2.23 ~.006e

Well

Rancho Peñon; 30 5/18/73 25.5 7.22 125 6.24 12.6 1.04 410 6.71 17 .35 +1.46 6.00 1.62 +0.38 +0.09 –2.13 —in shale

* Milligrams per liter and milliequivalents per liter; alkalinity in mg/l assumes all is HCO3–

† The Coy, Choy, Mante, Frío, and Media Luna values are by the Secretaría de Recursos Hidáulicos; all other are from this study; e = estimated

74


Table 5.1bChemical analyses and derived geochemical measures (other analyses)



Springs

Choy 31 5/20/72 26 — 199 9.92 47.2 3.88 319 5.23 438 9.12 –1.95 2.56 — — — –0.73 3.07

32 6/03/72 26.4 ‡6.82 200 10.00 47.4 3.90 ‡278 4.56 424 8.83 +1.87 2.56 ‡1.39 ‡–0.11 ‡–0.22 –0.74 ~5.5

33 8/06/72 25.0 6.90 117 5.82 10.2 0.84 264 4.32 82 1.71 +4.96 6.93 1.49 –0.17 –0.50 –1.47 flood

34 12/23/71 26.0 6.92 184 9.20 48.6 4.00 322 5.27 401 8.35 –1.57 2.30 1.43 +0.02 –0.07 –0.79 4.04

35 1/15/72 26.3 6.87 198 9.90 43.8 3.60 306 5.02 405 8.43 +0.19 2.75 1.40 –0.02 –0.14 –0.76 3.64

36 6/05/71 26 — 200 10.00 48.6 4.00 — — 430 8.96 — 2.50 — — — — 3.00

37 6/12/71 26.4 6.9 204 10.20 46.2 3.80 298 4.88 430 8.96 +0.57 2.68 — — — –0.73 2.90

38 6/12/71 26.4 6.9 204 10.20 48.6 4.00 315 5.17 — — — 2.55 — — — — 2.90

39 6/15/71 26.4 6.9 200 10.00 48.6 4.00 319 5.22 — — — 2.50 — — — — 2.96

40 6/18/71 26.4 6.9 200 10.00 43.8 3.60 319 5.22 445 9.28 –3.20 2.78 — — — –0.72 7.67

41 6/20/71 26.4 6.9 196 9.80 46.2 3.80 325 5.32 — — — 2.58 — — — — 4.83

42 6/21/71 26.4 6.8 196 9.80 38.9 3.20 339 5.56 — — — 3.06 — — — — 16.7

43 6/22/71 26.5 6.7 184 9.20 35.3 2.90 312 5.12 365 7.60 -2.50 3.17 — — — –0.81 19.7

44 6/24/71 26.4 6.7 168 8.40 34.0 2.80 312 5.12 — — — 3.00 — — — — 18.1

45 6/29/71 25.2 6.6 104 5.20 7.3 0.60 283 4.63 52 1.08 +0.78 8.67 — — — –1.68 27.0

46 6/30/71 25.2 6.5 102 5.10 6.1 0.50 276 4.52 39 .82 +2.38 10.2 — — — –1.80 30.0

47 7/07/71 24.9 6.5 104 5.20 2.4 0.20 254 4.16 30 .62 +6.09 26 — — — –1.90 6.0

48 7/14/71 25.1 6.5 104 5.20 8.5 .70 251 4.12 — — — 7.43 — — — — 25.0

49 7/24/71 — — 112 5.60 8.5 .70 — — 88 1.83 — 8.00 — — — — 5.58

50 8/05/71 25.4 6.7 132 6.60 17.0 1.40 265 4.35 184 3.83 –1.11 4.71 — — — –1.13 4.91

51 8/14/71 26.0 6.90 168 8.40 38.9 3.20 292 4.78 325 6.77 +0.22 2.63 1.45 –0.06 –0.17 –0.88 25.0

Mante 52 5/29/72 ‡26.5 ‡6.73 218 10.90 34 2.80 ‡300 4.92 — — — 3.89 — — — — 11.7to 14

53 1/17/72 26.2 6.78 184 9.20 52.3 4.30 303 4.97 414 8.62 –0.33 2.14 1.32 –0.15 –0.22 –0.78 14.52

54 6/06/71 — — 236 11.80 53.5 4.40 — — 535 11.14 — 2.68 — — — –0.60 11.85

55 8/17/71 25.6 6.90 — — — — 265 4.35 180 3.76 — — — — — — ~25

Santa Clara 56 1/07/72 24.0 7.00 88.1 4.40 4.9 .40 271 4.44 7 .15 +2.24 11.0 1.58 –0.14 –0.58 –2.57 0.38m

57 6/06/71 24.1 — 86.5 4.32 2.9 .24 — — — — — 18.0 — — — —base flow

Arroyo Seco 58 1/07/72 23.6 6.90 85.7 4.28 2.4 .20 247 4.04 7 .15 +3.34 21.4 1.52 –0.29 –0.88 –2.57 0.007e

(Rancho Peñon)

59 6/05/71 23.8 — 85.7 4.28 1.9 .16 — — 10 .21 — 26.8 — — — — 0.0008e

Presa Riachuelo 60 6/15/71 27.5 7.05 100 5.00 14.6 1.20 327 5.36 56 1.16 –2.52 4.17 — — — — 0.007e

San Rafael de 61 5/25/73 27.4 6.91 127 6.36 17.5 1.44 305 5.00 140 2.92 –0.76 4.42 1.42 –0.06 –0.28 –1.25 0e

los Castros

Coy 62 6/04/72 22.5 ‡7.31 107 5.34 15.3 1.26 ‡186 3.05 162 3.37 +1.38 4.24 ‡2.07 ‡–0.01 ‡–0.25 –1.23 33.1

63 7/22/72 23.2 6.98 190 9.48 30.9 2.54 231 3.78 454 9.46 –4.83 3.73 1.66 –0.09 –0.32 –0.71 54.5

64 12/23/71 23.5 6.90 249 12.40 52.3 4.30 268 4.40 605 12.60 –0.89 2.88 1.52 –0.03 –0.18 –0.55 22.2

65 6/08/71 25 — 341 17.00 51.1 4.20 — — 836 17.41 — 4.05 — — — –0.34 16.2

66 6/29/71 22 6.8 88.1 4.40 12.2 1.00 168 2.75 120 2.50 +1.41 4.40 — — — –1.40 30.2

continued on next page

75


Table 5.1bcontinued



Springs continued

Coy 67 8/05/71 23.0 6.85 204 10.20 31.6 2.60 250 4.10 430 8.96 –1.01 3.92 1.50 –0.15 –0.38 –0.70 37.1continued

Poza Azul 68 5/29/72 ‡21.9 ‡7.15 77.7 3.88 11.4 .94 ‡226 3.71 54 1.12 –0.10 4.13 — — — –1.76 18.95+(Frío)

69 12/27/71 — 7.05 102 5.08 15.6 1.28 275 4.50 92 1.92 –0.47 3.97 1.64 –0.09 –0.31 –0.47 ~13.2+

Nacimiento 70 6/15/71 23.0 7.1 122 6.08 20.9 1.72 257 4.22 180 3.75 –1.08 3.53 1.72 –0.03 –.023 –1.16 ~8–9+(Frío)

Nacimiento 71 1/06/72 20.0 6.95 47.3 2.36 5.6 .46 177 2.90 4 0.8 –2.76 5.13 1.74 –0.66 –0.96 –3.00 3.70m

Huichihuayán

(Río Huichi- 72 6/27/71 21.7 7.4 44.1 2.20 8.5 .70 — — 5 .10 — — — — — — —huayán)

San Juanito 73 6/02/72 19.7 ‡7.38 47.7 2.38 4.4 .16 ‡154 2.52 5 .10 +2.24 6.61 — — — –2.89 6–8e

(Mujer de Agua)

74 1/06/72 19.6 6.90 43.3 2.16 6.3 .52 146 2.40 1 .02 +5.10 4.15 1.77 –0.83 –1.08 –3.62 2.53m

Pimienta 75 6/04/72 22.5 ‡7.21 127 6.34 21.1 1.74 ‡210 3.44 238 4.96 –1.94 3.64 — — — –1.04 —

76 6/27/71 22.5 6.9 152 7.60 21.9 1.80 186 3.05 300 6.25 +0.53 4.22 1.67 –0.32 –0.56 –0.90 7e

Agua Clara 77 1/01/72 24.0 6.70 229 11.40 43.8 3.60 272 4.45 500 10.41 +0.47 3.17 1.31 –0.23 –0.40 –0.63 1–2e

Tanchachin 78 5/29/72 ‡22.0 ‡7.28 81.7 4.08 12.2 1.00 ‡229 3.75 61 1.27 +0.59 4.08 — — — –1.70 —

79 1/01/72 24.4 6.70 237 11.80 36.5 3.00 317 5.19 447 9.30 +1.06 3.93 1.23 –0.13 –0.35 –0.66 2.21m

80 7/11/71 24.4 6.5? 216 10.80 24.3 2.00 308 5.05 380 7.91 –0.62 5.40 1.04? –0.36? –0.64? –0.73 3e

Puente de Dios 81 7/11/71 22.6 6.7 168 8.40 18.2 1.50 203 3.32 300 6.25 +1.69 4.67 — — — –0.87 7e

(Tamasopo)

Los Bañitos 82 5/17/73 32.8 6.79 176 8.78 41.1 3.38 360 5.89 321 6.69 –1.66 2.60 1.19 +0.02 –0.05 –0.88 .00005e(Covadonga)

83 6/03/72 32.8 ‡6.65 175 8.72 41.8 3.44 ‡353 5.79 311 6.48 –0.39 2.77 — — — –0.90 —

84 6/27/72 32.8 6.70 176 8.78 39.9 3.28 354 5.79 314 6.54 –1.08 2.93 1.10 –0.08 –0.15 –0.89 —

Tantoán thermal 85 1/07172 31.2 6.90 100 5.00 6.8 .56 302 4.95 13 .27 +3.05 8.93 1.37 –0.07 –0.40 –2.29 0.014m

Taninul 86 6/03/72 39.4 ‡6.52 139 6.94 20.9 1.72 ‡431 7.06 52 1.08 +3.05 4.02 — — — –1.65 —sulfur pool

87 1/15/72 34.8 6.6 129 6.44 15.1 1.24 414 6.79 32 .67 +1.47 5.19 0.90 –0.13 –0.31 –1.88 0.07e

88 6/18/71 40.3 — 136 6.80 19.4 1.60 451 7.39 — — — 4.25 — — — — 0.17e

89 6/21/71 30.9 — 124 6.20 9.7 .80 400 6.55 — — — 7.75 — — — — 0.85e

90 6/22/71 29.9 — 124 6.20 13.4 1.10 401 6.58 — — — 5.64 — — — — 0.71e

91 6/24/71 30.6 — 128 6.40 12.2 1.00 388 6.36 — — — 6.40 — — — — 0.35e

92 6/29/71 30.1 — 120 6.00 4.9 .40 334 5.48 — — — 15.0 — — — — 0.17e

93 7/07/71 33.5 — 136 6.80 7.3 .60 351 5.75 46 .96 +4.82 11.3 — — — — 0.08e

94 8/05/71 35.1 — 134 6.70 15.8 1.30 404 6.62 — — — 5.15 — — — — —

Taninul 95 6/04/71 — — 132 6.60 2.4 .20 — — — — — 33.0 — — — — sinkinglimestone pool trickle

96 6/18/71 33.7 — 140 7.00 12.2 1.00 — — — — — 7.00 — — — — sinking0.05e


76


Table 5.1bcontinued



Springs continued

Taninul 97 6/21/71 29.7 — 124 6.20 7.3 .60 — — — — — 10.3 — — — — ?limestone pool continued

98 6/24/71 27.7 — 120 6.00 4.9 .40 — — — — — 30.0 — — — — rising0.1e

99 7/07/71 27.9 — 112 5.60 2.4 .20 — — — — — 28.0 — — — — rising0.01e

Cave lakes

Sótano de Jos 100 1/14/72 24.5 7.3 84.9 4.24 3.4 .28 256 4.19 15 .31 +0.18 15.14 1.90 +0.12 –0.38 –2.25 —

Sót. de Soyate 101 7/16/71 — — 84.1 4.20 2.4 .20 247 4.05 — — — 21.0 — — — — —

102 1/03/72 — 7.0 73.7 3.68 3.9 .32 220 3.60 6 .12 +3.5 11.5 1.66 –0.28 –0.72 –2.69 —

Sót. del Tigre 103 12/21/71 — 7.2 83.3 4.16 5.8 .48 272 4.47 23 .48 –3.2 8.67 1.77 +0.04 –0.33 –2.09 —

Cueva de los 104 7/15/72 24.1 7.40 87.0 4.34 1.9 .16 236 3.87 14 .29 +5.7 27.1 2.04 +0.19 –0.28 –2.27 —Sabinos

105 7/15/72 23.5 7.40 92.6 4.62 1.7 .14 239 3.91 11 .23 +6.2 33.0 2.04 +0.21 –0.46 –2.25 —

Sótano de la 106 1/04/72 — 7.2 82.5 4.12 4.6 .38 256 4.19 4 .08 +2.5 10.8 1.80 +0.02 –0.40 –2.83 —Tinaja

107 1/04/72 — 7.15 94.5 4.72 4.9 .40 304 4.99 6 .12 +0.04 11.8 1.67 +0.09 –0.35 –2.63 —

108 1/04/72 — 7.05 88.1 4.40 7.3 .60 289 4.73 4 .08 +1.9 7.33 1.60 –0.06 –0.39 –2.83 —

109 7/26/71 — — 92.2 4.60 3.6 .30 — — — — — 15.3 — — — — —

110 7/26/71 — — 94.2 4.70 3.6 .30 — — — — — 15.7 — — — — —

Sót. del Arroyo 111 12/31/71 — 7.2 67.3 3.36 3.6 .30 210 3.44 3 .06 +2.1 11.2 1.88 –0.14 –0.57 –3.01 —

112 12/31/71 — 7.3 66.5 3.32 3.4 .28 204 3.34 3 .06 +2.8 11.9 1.99 –0.06 –0.49 –3.01 —

113 12/31/71 — 7.3 85.7 4.28 4.9 .40 264 4.33 10 .21 +1.5 10.7 1.88 +0.14 –0.27 –2.43 —

114 1/12/72 — 7.0 121 6.04 10.2 .84 396 6.49 12 .25 +1.0 7.19 1.42 +0.14 –0.20 –2.28 —

Cave stream (between lakes)

Sót. del Arroyo 115 1/12/72 — 7.55 80.1 4.00 4.9 .40 263 4.31 6 .12 –0.30 10.0 2.14 +0.36 –0.04 –2.67 trickle

Cave drips

Gurtas de 116 6/14/72 — 7.32 99.4 4.96 1.7 .14 287 4.70 — — — 35.4 — — — — —Quintero

117 6/14/72 — 7.4 83.3 4.16 1.7 .14 247 4.05 — — — 29.7 — — — — —

118 6/14/72 23.9 7.4 101 5.06 1.9 .16 — — — — — 31.6 — — — — —

Sót. del Arroyo 119 1/12/72 — 7.05 133 6.64 6.8 .56 434 7.12 4 .08 –0.11 11.9 1.43 +0.27 –0.17 –2.72 —

Sótano de Jos 120 8/12/71 — — 80 4.00 4.9 .40 — — — — — 10.0 — — — — —

Sótano de la 121 7/26/71 — — 100 5.00 4.9 .40 — — — — — 12.5 — — — — —Tinaja

122 7/19/71 — — 136 6.80 7.3 .60 423 6.94 — — — 11.3 — — — — —

123 7/19/71 — — total hardness 6.20 meq/l — — — — — — — — — — —

124 7/20/71 — — 88.2 4.40 1.2 .10 — — — — — 44.0 — — — — —

125 7/20/71 — — 102 5.10 1.2 .10 — — — — — 51.0 — — — — —


77


spring, the concentrations of the last threeconstituents were found to be unimportant(see Table 5.2). Many of the characteris-tics of the karst waters in the area are shownin Figures 5.2–5.5, which will be referredto in the following pages. Figure 5.2 showsthe chemical composition of representativesamples from each class of karst water sothat they may be easily compared.

5.2.1. Swallet streams

The synclinal valley along the westernside of the Sierra de El Abra contains a flu-vial system developed on shale and flaggylimestone, with a thin veneer of soil. Aspreviously described, many of the arroyoshave cut down to the El Abra limestone andsubsequently have been pirated under-ground. The streams are seasonal; they arecompletely dry except for small pools dur-ing the dry season, but carry large floodsduring major wet season storms. A greatdeal of sediment and organic material, in-cluding logs, is swept into the caves dur-ing floods. The five samples listed werecollected at very low discharges and havea range of 85 mg/l to about 200 mg/l totalhardness (as CaCO3). The total hardnesswas mostly calcium carbonate. For thesesamples, there was considerable difficultyin obtaining precise determinations of Caand Mg, possibly because of organic com-pounds in the water. However, the alkalin-ity roughly balances the measured cations.Thus, these samples serve as an indication

Table 5.1bcontinued



Surface streams

Yerbaniz 126 7/25/72 25.1 7.28 32.0 1.60 1.2 .10 96.5 1.58 — — — 16.0 2.28 — — — 0.15e

arroyo north branch

Arroyo 127 7/25/72 26.3 7.48 61.7 3.08 1.0 .08 271 4.44 — — — 38.5 2.04 — — — 0.04e

Limoncito

Arroyo arroyo 128 7/25/72 26.4 7.95 72.1 3.60 0.5 .04 236 3.86 — — — 90.0 2.59 — — — 0.06e

(Sótano del Arroyo)

Japonés arroyo 129 7/18/71 — 7.4 46.5 2.32 1.5 .12 130 2.13 — — — 19.3 — — — — 0.03e

Taninul sulfur 130 7/18/71 — 7.2 total hardness 2.64 meq/l 148 2.42 — — — — — — — — 0.03e

* Milligrams per liter, milliequivalents per liter; alkalinity in mg/l assumes all is HCO3–

† Coy, Choy, Mante, Frío, and Media Luna values are by the Secretaría de Recursos Hidráulicos; all others from this study;

m = measured by current meter, e = estimated, usually by measurement of surface velocity and rough channel area

‡ pH or alkalinity measured by Russell S. Harmon and of uncertain accuracy; any error will affect pPCO2, SIc, and SId

here, the total concentration (analyticalvalue) of a species in solution will be indi-cated by a symbol without charge (e.g., Ca,or Catot for emphasis), and that part that isionized will be indicated as such (e.g.,Ca++).

5.2. Chemistry and Environment of theKarst Waters

In this section, the chemistry of the karstwaters will be examined and compared. Forthe springs, the discussion will be orientedtoward their dry season chemistry in orderto show their character during that longperiod of stable aquifer conditions andspring discharge. For other classes of wa-ter, all available data will be examined.

Nearly all the analyses performed andsome calculated parameters are listed inTable 5.1. The variables generally measuredwere Ca, Mg, alkalinity, SO4, temperature,pH, turbidity (qualitative), and discharge(where applicable). These data have beendivided into two groups according to thequality of the determinations: Table 5.1a,high quality, and Table 5.1b, less reliable(but in most instances believed reasonablygood). A few good analyses were placed inthe second category for reasons other thanthe quality of the analytical measurements(e.g., the Covadonga thermal spring LosBañitos because the actual source could notbe sampled). In addition, many springswere checked for Na, K, and Cl. Exceptfor the Taninul Sulfur Pool, a thermal

of the dissolved load entering the swalletcaves. The concentrations seem surpris-ingly high, but the Méndez shale evidentlycontains a large amount of CaCO3 (con-firmed by J. Carrillo Bravo, personalcommunication). It seems likely that thedissolved load will be considerably dimin-ished during flood stage, which is volu-metrically much more important.

5.2.2. Cave drips

Cave drips (vadose seeps) and vadoseflows (vertical shaft or open crack waters)are the predominant recharge water typesin the Sierra de El Abra. Ten drip sampleswere collected, seven of them in swalletcaves. The total hardness ranged from215 mg/l to 360 mg/l as CaCO3. In all thesamples, the Ca/Mg molar ratio was 10 orgreater. Because this is such an importantclass of water in the El Abra range, it wouldbe desirable to greatly expand the samplingprogram. The few samples collected havean average total hardness of 274 mg/l,somewhat higher than the small springsalong the east face of the El Abra. This isnot surprising, inasmuch as the area appearsto be in a minor phase of calcite deposi-tion. All except one of the samples werecollected during the wet season, since manysources are inactive, or nearly so, duringthe dry season. At present no relationshipbetween the dissolved solids and the dripor flow rate can be given.

78


5.2.3. Cave lakes

The cave lakes in the Sierra de El Abramay be distinguished in two ways. Oneconsideration is the source of the water, i.e.,whether it comes from vadose seeps andflows or from swallet flood waters. Nearlyall the samples are from swallet cave lakesin which the stored water is flushed outprobably four to eight times each year byfloods, counting successive days of rain-fall and flooding as one event. In manylakes, water is slowly exchanged betweenflood events by trickles. The second con-sideration is the hydrologic position of thecave lake. All of the lakes sampled exceptone are perched above the zone of perma-nent saturation. The lake in Sótano deSoyate is the only known water-table lakein the Sierra de El Abra.

Fourteen different lakes were sampled.The analyses are placed in Table 5.1b be-cause, with the exception of the Cueva deLos Sabinos lakes, the pH of the sampleswas measured at the field laboratory (assoon as possible) rather than in the cave,and because most of the lake samples werecollected during the earlier part of the sam-pling program. The pH of samples from thetwo lakes in Cueva de los Sabinos wasmeasured in the cave, one day later in thefield house, and ten days later in the fieldhouse. The pH increased by 0.02 units af-ter one day for both samples and 0.08 forone sample and 0.10 for the other samplesafter ten days. Thus, the true pH of the othersamples may be within 0.05 units of thevalue listed, including the reading error. ApH error affects the calculated PCO2 and SIcvalues. If the pH error is 0.05 units, the

error in the pPCO2 and SIc values would beabout 0.05.

The cave lakes are calcium bicarbonatetype water, as would be expected from a lime-stone region. The average concentrationsof calcium and magnesium were 4.34 meq/land 0.36 meq/l, respectively, for a totalhardness of 235 mg/l. These cations arebalanced by the alkalinity, which wouldeffectively be bicarbonate, and by an aver-age of about 0.2 meq/l of sulfate. The Ca/Mgratio ranged from 7.19 to 33.0 (Figure 5.3).The data indicate that the water has notcome in contact with a significant amountof sulfate bearing rock and has dissolvedlittle dolomite or magnesium from magne-sian calcite. Figure 5.2 shows four selectedanalyses for comparison with other watertypes.

The environment of the cave lakes has

Figure 5.2. Comparison of chemistry of water samples from various environments in the Sierra de El Abra.

79


considerable spatial and temporal variation.Few data are available yet, but some obser-vations have been made. In the precedingparagraph it was noted that the principalsource rock is a relatively pure limestone(also see Carrillo Bravo, 1971). The CO2reservoir that the water is in contact withwill ultimately exert a control on theamount of dissolved calcite. Some sites arerelatively open and an exchange betweenthe cave air and the atmosphere occurs,whereas other sites are relatively closed.For example, the two lowest hardness con-centrations measured were collected in theMain Passage of Sótano del Arroyo (seecave description, Chapter 6) when therewas a noticeable air current blowing intothe cave through this passage. Using aDraeger kit, the CO2 in the cave air wasmeasured to be 0.03 to 0.035%, the sameas the average atmospheric CO2 content.

By far the highest hardness measured oc-curred in a lake at the bottom of a 30 meterdrop in Strickland’s Bad Air Passage, alsoin Sótano del Arroyo. This is a downstreampassage, the lowest explored point in thecave, and very near the water table. The airwas stagnant because of restricted circula-tion; the CO2 was measured to be 3.8%,identical to the calculated PCO2 (and dan-gerous as well!). The average calculatedpPCO2 for twelve sample sites is 1.80(1.6%), which is considerably greater thanthe CO2 content of the cave air at the sampletime (see Table 5.3). Also, it should benoted that the samples were approximatelysaturated with respect to calcite (assumingthe pH drift was not large). At the end ofthe dry season in early June 1973, many ofthe lakes in the lower part of Arroyo had alayer of calcite completely covering theirsurfaces.

The chemical processes that occur in theperched lakes and the phreatic zone of theSierra de El Abra are probably very sig-nificant. Floods wash huge quantities oforganic material into the swallet caves. Sig-nificant quantities must also be entrainedin the smaller vadose flows of the El Abrarange, and bats introduce waste in bothenvironments. In the swallet caves, Cuevade los Sabinos, Sótano de Soyate, and manyother caves, large populations of blind fishand other aquatic animals survive on thisinflux of material. Mitchell et al. (1977)discuss the occurrence of the cave fishes.Reducing conditions evidently exist in thebottom sediments of many of these lakes,for as one wades across them gas bubblesrise to the surface and may then be lightedwith a carbide lamp. The gas is almost cer-tainly methane, created by the decom-position of vegetal debris. From these dataand hydrologic and geomorphic observa-tions, two important hypotheses are made:that during intense storms, swallet floodwaters on the western margin and vadoseflows on the El Abra range introduce intothe phreatic zone and the perched lakeslarge quantities of water with considerablecapacity for calcite solution, and that theoxidation of organic matter in these sub-surface environments may provide a sig-nificant increase in the CO2 available forsolution, especially in sites where air cir-culation is restricted. Vadose trickles mightbecome saturated with respect to calcitebefore they reach the phreatic zone, butflood waters almost certainly would not.

5.2.4. Cave streams

Dry season vadose streams in the ElAbra are almost non-existent in the ap-proximately 45 km of explored passage.Only four trickles are known, and each hasan estimated flow on the order of 1 l/minor less. One sample of a trickle in the Wal-lows in Sótano del Arroyo is listed. Thecomposition of this class of waters shouldnot be very different from the cave lakes,because the streams are usually flows be-tween lakes.

5.2.5. Springs

Karst springs are the integrated outputof karst hydrologic systems. The physicaland chemical character of the output de-pends on the type of recharge, aquifer hy-draulics, soil cover, climate, relief, rockstructure, mineralogy, reaction kinetics, andother geologic factors such as intrusive

Table 5.2Analyses for Na, K, and Cl

Sample in Na+ K+ Cl–

Spring Table 5.1 mg/l mg/l mg/l

Choy 1 3.6 0.7 4.8

35 3.3 0.8 4.2

3* — — 6.9

Mante 10 6.1 1.5 4.2

Coy 12 4.2 1.2 3.3

Taninul sulfur pool† 24 88.4 10.2 225

87 55.0 6.2 —

26* 11.0 1.5 30.0

Los Bañitos (Covadonga) 82 9.1 1.3 3.8

Tantoán thermal 29 9.0 0.7 —

Arroyo Seco 59 2.3 0.7 <7.5‡

Poza Azul (Frío) 18 4.8 1.0 —

Sabinas 20 1.3 0.2 1.5

Media Luna 23 13.9 5.7 —

San Juanito 73 0.7 0.3 —

Agua Clara 77 3.3 0.9 —

Canóas 22 — 1.2 —

Santa Clara 57 — — 7.5–15‡

Taninul limestone pool 95 — — 144–152‡

96 — — 136–144‡

97 — — 23–30‡

* In flood.

† See Figure 5.11 for other chloride measurements.

‡ Measured by Hach kit; others by Cl– specific-ion electrode.

80


bodies and heat flow.For this study, the original plan was to

sample only the Sierra de El Abra springs.However, the chemistry of the Choy andMante springs indicated that the hydrogeo-logy was more complicated than initiallyenvisioned. Hence, the sampling programwas expanded in an effort to determine theorigin of the high sulfate-magnesium wa-ters. Two types of studies have been con-ducted: the determination of the majorchemical constituents of a large number ofsprings in the region during the dry seasonbase flow period, and the examination ofthe chemical and physical changes whichoccur during wet season flood pulses atselected springs. This section will considerthe first type of study.

Twenty springs were sampled at leastonce, including all of the large springs thatoccur near the eastern margin of the Plat-form. Only a few springs in the interior ofthe Platform were sampled. The data givenfor the small San Rafael de los Castros andthe Riachuelo nacimientos may not repre-sent the ground water that supplies them,because there was no visible flow when theformer was sampled (the sample was takenfrom the pool directly at the spring) andthe latter rises into a small artificial reser-voir, thus making it difficult to obtain adirect sample. As described in Section 4.2,many of the small springs cease flowingsometime during the dry season. Thoughthe Media Luna spring rises from the bot-tom of a 75 meter deep lake, its flow is sogreat that the chemistry could not changesignificantly. Each of the remaining eigh-teen springs (including Media Luna) arerepresented in Figure 5.2 by a dry seasonchemical analysis, except for the Pimienta,the Puente de Díos, and the Canóasnacimientos, which were at an intermedi-ate flood stage. The twelve large springsare grouped in the upper row and arrangedapproximately from the southernmost onthe left to the northernmost on the right.Two other groups are also shown, threethermal springs and three other relativelysmall springs. The variety of chemical typesof spring water in the region as comparedto the internal waters of the Sierra de ElAbra is apparent.

Figure 5.4 presents the observed tem-perature regimes for the springs. The topmark is the highest or nearly the highesttemperature likely to be measured at eachspring during the year. Where other tem-peratures have been measured, the observedrange is given by a vertical line. For theCoy, the Choy, and the Taninul Sulfur Pool,

Figure 5.3. Ca/Mg against SO4 for El Abra region karst waters. The graphdistinguishes the calcium-bicarbonate type waters from the calcium-magnesium sulfate type waters. Also shown are the effects of flooding atthe Choy and Coy springs (discussed in Section 5.3).

Table 5.3CO2 of soil and cave air

Location % CO2

Sótano del Arroyo (December 1971–January 1972)

1. top of first drop inside cave 0.03

2. lake at bottom of Big Triangular Room 0.035

3. junction of The Wallows and Neal’s Passage 0.035–0.04

4. The Wallows near First RIght Hand Lead 0.045

5. The Wallows at Second Right Hand Lead 0.045

6. 60 m before fissure on other side of flowstone block in 0.05 Strickland’s Bad Air Passage

7. top of fissure (30 m drop) at end of Strickland’s Bad Air Passage 2.7

8. 16 m down the fissure in no. 7 above 3.5

9. bottom of fissure in no. 7 above 3.8

Sótano de Soyate (January 3, 1972)

1. jungle atmosphere 0.6 m above ground at Soyate, 11:00 A.M. 0.025

2. soil air near Soyate at 10 cm depth; reasonably loose humus and earth 0.080

3. bottom of 195 m entrance pitch 0.060

4. at 220 m depth; 14 m above the lake 0.065

Sótano de la Tinaja (December 1971)

1. cave air above vegetal debris and guano pool just before Traverse Lake 0.08

2. 10 cm down in debris at no. 1 above; sticks and leaves 0.31

3. air at Second Lake, Sandy Floored Passage 0.10

Trail to Cueva de Tanchipa (January 1971)

1. in loose vegetal debris on main range 0.055

81


this must represent nearly the full range ofvariation that occurs at these springs, sincethey were studied in detail over a widerange of discharge. The three thermalsprings stand out clearly, especially theTaninul Sulfur Pool. The Media Luna tem-perature would appear to be anamolous(measured 10 cm below the surface). Awarm surface layer might be partly respon-sible, but this seems unlikely because thedischarge is so large (about 4 m3/sec.) Thefourteen other springs exhibit a range from19.6 to 26.5°C, for a difference of 6.9°C.More information will have to be devel-oped before the spring water temperaturescan be properly understood; hence an ex-planation will be given in Chapter 7.

Some of the springs have relativelypure calcium bicarbonate type water, witha Ca/Mg molar ratio greater than 10. Oth-ers have somewhat greater amounts of Mg,but contain very little SO4. Many of thelarge springs have Mg concentrations

between 40 to 60 mg/l, with Ca/Mg ratiobetween 2.5 to 4.5, and have very high Caand SO4 concentrations, reaching as muchas 340 and 840 mg/l respectively. These re-lationships are best seen in Figure 5.3.Many springs were tested for Na, K, andCl. Only the Taninul Sulfur Pool has sig-nificant concentrations of these constitu-ents, which are not normally important inkarst regions. Analyses from six springswere plotted on a trilinear diagram, Figure5.5. All of the points except the TaninulSulfur Pool lie near the line which wouldbe produced if calcite, dolomite, and gyp-sum were the only dissolved minerals.

The thermal springs do not seem to havea well defined chemical pattern. Three havebeen sampled, and two others are known.The Tantoán thermal spring has nearly thesame composition as other small El Abrasprings, but has an elevated temperatureand a slight odor of H2S. The Los Bañitos(Covadonga) source has a very small flow

and a chemical content somewhat similarto the Choy. To the contrary, the Sulfur Poolat the Hotel Taninul is the hottest, has byfar the largest discharge, and Na plus Clcomposes nearly one half of the dissolvedload. Visibility in it is normally only a fewfeet. A spongy precipitate of complex sul-fur compounds collects on the pool surface,and there is a strong smell of H2S.

The factors that control the concentra-tions of dissolved species are the mineralsexposed along the flow paths, solution ki-netics (and time), saturation with respectto various minerals, partial pressure ofcarbon dioxide and other gases, and tem-perature. Where the data are of sufficientquality, the saturation state with respect tocalcite, dolomite, and gypsum is given inTable 5.1.

All of the springs are undersaturatedwith respect to gypsum. The Coy and theMedia Luna were the most nearly saturated,reaching an SIg of about –0.35. With the

Figure 5.4. Temperature regime of springs in the region.

82


possible exception of the Taninul SulfurPool, all of the springs were also under-saturated in dolomite. Springs such as theMante, the Coy, and the Choy, which con-tain a large amount of dissolved gypsum,also have the highest Mg concentrations.Late in the dry season, they have an SId ofabout –0.10, only slightly undersaturated.Springs such as the Sabinas and the Fríogroup, in which carbonate minerals providemost of the dissolved load, have an SId be-tween –0.20 and –1.0. The springs forwhich good quality data were obtained atthe end of the 1973 dry season are approxi-mately saturated with respect to calcite.Unfortunately, not enough samples were col-lected from springs containing principally

dissolved carbonate minerals to determineif, as a general pattern, they tend to becomesaturated with respect to calcite without theassistance of “excess” Ca from dissolvedgypsum (many of the springs cease flow-ing late in the dry season). Because theSanta Clara and Arroyo Soco de Peñonwere found to be only mildly undersatu-rated early in the dry season and the Sabinasand Frío springs were saturated in May1973, it seems likely that by the end of thedry season these “normal” karst springwaters approach equilibrium with respectto calcite. (The Frío springs are actuallytransitional in composition; see Figure 5.2.)Calculated carbon dioxide partial pressuresranged from 10–2.00 (1%) to 10–1.26 (5.5%)

and averaged about 10–1.5 (3%) for thesprings. The El Abra springs have a higherPCO2 than those discharging water fromnearby higher mountains. Also, the PCO2of the El Abra springs is much higher thanthe cave air CO2 measurements at all butone cave locality (Table 5.3; another badair passage that is small, with restricted aircirculation, is known in Sótano del Tigre,but was not measured). This suggests thatmost of the cave sample sites are relativelyopen and that fissures and passages nearthe phreatic zone must generally have re-stricted air circulation and higher CO2 con-centrations. Comparison of the chemistryof these springs with other regions will begiven in Chapter 7.

Figure 5.5. Trilinear plot of samples from six springs. The effect of flooding on the Taninul Sulfur Pool is shown.

83


Normally, the karst spring waters havea bluish cast and are clear. The best exampleis the Media Luna where, with the aid of adiving mask, one can easily see the bottom75 meters below. On occasion, a very slightmilkiness has been observed at the Coy, theMante, and the Choy, all high sulfatesprings. This may be due to precipitationof calcite in the highly turbulent flow be-low the exit point as the water begins toequilibrate to surface conditions. However,no tufa deposits exist in these channels,although some are found at waterfallsaround the region. The thermal waters havea pronounced milkiness.

5.3. Flood Pulse Behaviorof the Springs

Ashton (1966) was one of the first work-ers to apply the “black box” model to karsthydrologic networks. The basic concept isthat a flood wave that enters the subsur-face undergoes a transformation in itsphysical and chemical character, dependenton the nature of the karst underground sys-tem and the properties of the water, beforeit is discharged back to the surface. Thetrick is to interpret what is in the box bymeasuring the characteristics of the output.Brown (PhD thesis, 1970), Brown and Ford(1971) and Atkinson et al. (1973) have sub-sequently made contributions to this field.Also, there have recently been a greatnumber of studies on the related problemof the relationships of various chemical andisotopic parameters to discharge and sepa-ration of the hydrograph of surface streams.In this thesis the water chemistry is used asa hydrological tool to study the dynamicsof the El Abra region karst systems. TheChoy, the Coy, and the Taninul Sulfur Poolwere selected for detailed studies duringflood pulses. A few other springs were alsosampled during floods to see if they mightbehave in a similar manner.

Figure 5.6 shows the data for El Choyspring for the period June 1 to August 14,1971. The first samples are typical of thedry season character of El Choy: clear blu-ish water at 26.4°C, with very high Mg, Ca,and SO4 concentrations (see the May 1973samples for the most accurate values). TheCa and SO4 values may be read directly inmilliequivalents per liter (meq/l), and Mgis obtained by subtraction of Ca from thetotal hardness curve. An estimate of thedepth to visible rocks in the water was madeas a crude measure of its turbidity (labelledvisibility in Figure 5.6). Though it may bepossible to see more than 30 m in the water

in the dry season, all values greater than 6m are plotted as 6. The first major stormyperiods of the rainy season began in mid-June (see Figure 4.7) and lasted about twoweeks. Each peak of the discharge curverepresents a distinct rainfall event, ratherthan the successive arrival of flood wavesvia different channels for the same rainfall,as was common in the works by Ashton(1966, 1967). Though water analyses werenot made more than once daily at best, theyappear to adequately define the concentra-tion curves, losing only the fine detail. Thefirst pulses were relatively small, causinga gradual decrease in dissolved solids andan increase in very fine suspended sedi-ment, which may have been remobilizedfrom within the aquifer rather than indicat-ing swallet water. Brown, muddy water,

probably from the swallet caves such asSótano del Arroyo, first appeared duringthe large pulse on June 27 (oral communi-cation, foreman of Rancho Florida near ElChoy). The dissolved load decreased untilthe water had nearly the same compositionas the Santa Clara and Arroyo Seco dePeñon springs: principally calcium bicar-bonate. The unequal flood-water dilutionof chemical species shown in Figure 5.6 isdifferent from that reported by Shuster andWhite (1971) in their study of spring waterchemical variations. The chemical responseseems to lag slightly behind the onset offlooding, reaching its lowest values late inthe storm period or after it, and then gradu-ally returning to its previous state. A smallerflood, July 12–16, had little effect on thechemistry or temperature because nearly the

Figure 5.6. Effect of flooding on chemistry of Choy spring, June–August 1971.

84


maximum effect had already been pro-duced. It is important to note that the physi-cal response (discharge) of the aquifer issharp and strongly peaked, quickly return-ing almost to its pre-flood state, whereasthe chemistry, temperature, and turbidityresponses are not as sharp and take abouttwo months to return to their pre-floodstate. This means that a substantial volume,probably several million cubic meters, ofphreatic water was flushed out by theground-water recharge, and it took twomonths for the chemistry of the newlystored water to become “normal” again.

In the summer of 1972, both the Coyand the Choy springs were studied during

flood pulses (Figure 5.7). Though lackingthe detail during the early stages of theflood compared with the previous year’sstudy, the same phenomena seem to havebeen repeated at El Choy. Here the turbid-ity data are of special interest. Dark, muddywater was flowing on June 13 (during aperiod of high flow, about 35 m3/sec, main-tained by smaller pulses following thelargest on June 9). This indicates that thearroyos of some swallet caves had run(probably on June 8, and on following daysas well) and some of the water was beingdischarged on June 13. By June 15, the tur-bidity had decreased somewhat, though thedischarge was similar, and the water was a

murky olive-green color. Probably most ofthe water from the swallet caves had alreadypassed, and a large part of the dischargewas being drawn from the main range. Theturbidity continued to improve over the nexttwo weeks, but quite a lot of sediment musthave been placed in suspension in the aqui-fer. As before, the chemical wave was muchbroader than the physical wave. Unfortu-nately, three moderate floods occurred atthe Coy during May 1972, before samplingbegan in June (refer to Figure 4.7). Thesefloods probably reduced the dissolved sol-ids in the discharged water a little belowthe values recorded at the end of the dryseason in May–June 1973 and June 8, 1971,so that these data may not reflect only theeffects of the June flood. However, the Juneflood was much larger than the others;hence most of the chemical changes areascribed to it.

The Coy chemistry and temperaturebehave in a manner similar to El Choy, de-creasing during flooding to a calcium bi-carbonate type water with a total hardnessof only 180 mg/l, but returning toward dryseason concentrations and temperature re-gime more rapidly than El Choy. This maybe due to the fact that the Coy has a verystrong base flow, the largest of any springin the region. One point of difference is thatthe Coy spring was never observed to carrymuddy water.

Three chemical analyses of Coy waterswere made during the summer of 1971(Table 5.1). They are sufficient to show thatthe Coy can change from its dry seasonchemical character to its flood-diminishedcontent in less than two weeks. On June 8,1971, Coy water contained the highest dis-solved solids of twelve samples collected—341 mg/l Ca, 51.1 mg/l Mg, and 836 mg/lSO4. Undoubtedly there was little changeuntil flooding began on the 18th. On June29, the water had only 88.1, 12.2, and 120mg/l of Ca, Mg, and SO4, respectively. ByAugust 5, at the beginning of the next ma-jor flood, the concentrations were up to204, 31.6, and 430 mg/l, respectively.

Obviously, there would be a good cor-relation between the Ca and SO4 contentover the range of discharges for the Coy orChoy, because the Ca excess over that de-rived from carbonate rock solution and theSO4 are derived from gypsum or anhydrite.Figure 5.8 shows a very strong correlationbetween the Mg and SO4 content of thesprings for all the flood pulses studied. Thechemistries rapidly pass diagonally down-ward during floods from the dry seasonvalues at the upper right to the minimum

Figure 5.7. Effect of flooding on the chemistry of Choy and Coy springs,May–July 1972.

85


values at the lower left, and then slowlyreturn toward the upper right after the flood.There may be some slight looping, but thedata are not sufficient to detect it or sea-sonal variations. Note that the paths of theChoy and Coy cross. This may be explainedby the lower dolomite content of the El Abralimestone in and near the reef than of thelimestone farther from the reef (CarilloBravo, 1971) that is the source of the Coyfloodwater. This may also be seen in Fig-ure 5.3, where the chemical paths are like-wise shown. The three summer analyses ofEl Mante do not show the full range ofvariation to be expected, but plot near theChoy line in Figure 5.8. A similar plot,

Figure 5.9, of temperature against SO4 con-tent also shows a very strong correlationfor all three springs. El Choy and El Mantegive nearly identical results. The Coy lineis approximately parallel to the others, but3.5°C lower. This must be due, at least inpart, to the cooler temperature of its re-charge water.

From the data and correlations discussedabove, it is inferred that there are twosources of water for the springs. (It wouldbe helpful for the following explanation torefer to Figure 2.2 and the lithologic de-scriptions in Chapter 2.) The first source,referred to as the local source, is water whichcirculates entirely within the limestone of

the El Abra or Tamasopo Formations. Thissource is a nearly pure calcium bicarbon-ate type water near the reef, but passes tocalcium-magnesium bicarbonate water (Ca/Mg ~ 4) farther west of the reef, where theMg content in the rock increases. It is be-lieved that the local source supplies all thedischarge of the smaller springs such as theSanta Clara (Figure 5.2), as well as someof the base flow and nearly all of the floodflow of the large springs. The second sourceprovides a water that contains much moredissolved Ca, Mg, and SO4 than the localsource, but about the same amount ofHCO3. It is also warmer than the localsource, as shown by the temperature de-crease during flooding. If the geology ofthe area as described by Pemex geologistsis correct, particularly with respect to thedistribution of gypsum in the subsurface(Carrillo Bravo, 1971), then the secondsource is a regional deep flow system origi-nating in the mountain ranges far west ofCd. Valles. The regional system must pickup its Ca, SO4 load from the GuaxcamiFormation. Most of the Mg probably is dis-solved from the dolomite at the base of theEl Abra Formation in the interior of thePlatform or from the Lower (?) Cretaceousdolomite encountered in the Tamalihualewell. The two sources mix to produce theoutput of the Choy, the Coy, the Mante, andseveral other large springs. Further analy-sis of this model will be given in Chapter7.

The changes that occurred in the degreeof saturation with respect to calcite, dolo-mite, and gypsum of the Coy and the Choywaters during the 1972 flood pulse studiesmay be seen in Figure 5.10. The leftmostpoint for both springs is a representativedry season value from May 1973, becauseno completely pre-flood samples were col-lected in 1972. Unfortunately, the SIc andSId of the second point for both springs areconsidered unreliable (pH and alkalinitywere measured by R. S. Harmon). Thoughthe interpretation is also complicated bysome small floods at the Coy spring duringMay of 1972, it is apparent that the satura-tion indices decrease not more than a fewweeks from the beginning of flooding. Infact, the saturation states seem to vary in amanner similar to the chemical constituentsshown in Figure 5.7, i.e., they respond rap-idly to each major flood and gradually re-turn to normal. This is clearly so for El Choy.

Several data in this chapter may bedrawn together here to make an importantconclusion: that a significant portion ofthe water circulating within the El Abra

Figure 5.8 (top). Relationship between Mg and SO4 for Choy, Coy, and Manteduring flooding. Figure 5.9 (bottom). Relationship between T (°C) and SO4 forChoy, Coy, and Mante during flooding.

86


limestone has a hydrologically short resi-dence time, ranging from perhaps one dayto several months. In fact, a large portionof the discharge during a flood event prob-ably comes from the rainfall which stimu-lated the flood. The large conduits devel-oped in the El Abra are capable of passingwater from sink to spring very rapidly. Theabove conclusion is supported by the fol-lowing data and interpretations:

1. The fact that the Coy and Choy be-came considerably undersaturated with re-spect to calcite, SIc about –0.4, during theserelatively modest floods suggests that someground-water recharge from the Junestorms passed completely through the sys-tem in a few days. The discharge was amixture of extant phreatic water and thenew recharge. This result is particularlysurprising for the Coy spring because ofits great distance from any possible re-charge area (at least 10 km). It should alsobe noted that the dry season SIc values ofthe carbonate springs—Sabinas, SantaClara, Arroyo Seco de Peñon, and Canóas—are higher than the SIc values of the floodflow of the Choy and the Coy.

2. The muddy water discharged from ElChoy is believed to come directly from theswallet caves such as Arroyo and Tinaja onthe west side of the El Abra range. Basedon the Choy 1971 turbidity observations(Figure 5.6), the swallet waters may travelto the Choy at a velocity on the order of1 km/hr. The flood water at these sinks is ahighly turbid mass of water, mud, and or-ganic debris. During the early stage offlooding the changes in color and turbidityof El Choy might be due to remobilizedsediment from within the aquifer, ratherthan swallet water or soil washed in fromthe top of the range. Neither muddy waternor suspended fine matter was ever ob-served at El Coy, though the water seemedto lose a little of its clarity during floods.This may be explained by the fact that thekarst west of Aquismón, which is believedto be the local source for El Coy, does nothave extensive pirated streams from shalecatchments as the southern El Abra does.

3. If the lycopodium test from Sótanode Japonés to El Choy is correct, then theconclusion is proven, at least for the swalletcaves of the southern Sierra de El Abra(Section 4.4.3).

Undoubtedly, some of the water has alonger residence time than the flood wa-ters described above. The springs obviouslycontinue to flow throughout the long dry

season. Some of the recharge water maypass into deep or restricted flow paths thatmight take as much as a few tens of yearsor more to traverse. The high SIc values ofthe carbonate springs during the dry sea-son also suggest a longer residence time.

One other sequence of observations,concerning the Taninul Sulfur Pool spring,is of interest. The data and interpretativecurves are shown in Figure 5.11 for June1971. The physical situation is that the ther-mal spring has been developed into an arti-ficial pool as a health spa, and about 80 m tothe north (parallel to the east face) lies astagnant pool, approximately 15 m acrossand at least 5 m deep (Figure 6.2). Thewater surface of the Sulfur Pool (thermalspring) is 1.5 meters above that of the“limestone” (stagnant) pool. Normally,the base flow of the Sulfur Pool, 0.05 to0.1 ml/sec, flows in a channel past the lime-stone pool, but often a tiny trickle is lostfrom the channel to the stagnant pool, ason June 4, 1971 (see top line of Figure5.11), when roughly 0.5 l/min entered thepool. As previously described, during thedry season the Sulfur Pool has a milkyopaqueness, is hot, has a strong smell ofH2S, has an abundance of sulfurous coagu-late, and has a high NaCl content. Chloride

contamination shows up in the limestonepool samples (Table 5.2).

By June 18, the discharge had increasedto about 0.17 m3/sec in response to the firstrains of the wet season (compare the dis-charge curve with that of El Choy, Figure5.6). Curiously, the temperature increasedslightly, and the visibility increased greatly.More importantly, a lot of Sulfur Pool wa-ter had entered the limestone pool; the flowwas about 0.05 m3/sec into the pool at thetime of observation. Many fish were gulp-ing air at the pool surface. Others weredead. The entire water body was movingin a giant clockwise vortex; it was sinkingback into the El Abra in the center of thepool

During the next eleven days, severalflood pulses passed through the sulfurspring system. The high discharges shownare estimates based on surface water ve-locity and approximate channel size. OnJune 21, the Sulfur Pool was at flood stage.The artificial chute to the normal channelcould not handle all the water. Up to 0.1meter depth of water was spilling over thenorthwestern edge of the pool and flowingoverland directly into the “limestone” pool.About 0.1 to 0.2 m3/sec was flowing intothe limestone pool from one channel and

Figure 5.10. Effect of flooding on degree of saturation with respect to calcite,dolomite, and gypsum at Choy and Coy, May–July 1972.

87


out the other, and the overland flow ap-peared to have been higher before the timeof observation. It is not known if there wasa small loss back into the El Abra, as onthe 18th. The effect on the chemistry isreadily apparent from the graphs. The flowfrom the Sulfur Pool had become almosttotally dominated by El Abra calcium bi-carbonate type water. The Cl content de-creased an order of magnitude, the totalhardness and sulfate content decreasedsomewhat, the temperature decreased bymore than 10°C, though it was still 5°Cabove normal El Abra water, the water be-came clear, and it had lost most of the H2Sodor. During the next two days, the dis-charge decreased somewhat. There wasonly a small overland flow into the lime-stone pool. Other conditions remainedabout the same.

By the late afternoon of June 24, thedischarge from the Sulfur Pool had de-ceased substantially. There were slightchanges in the hardness and temperaturetoward normal conditions, and the visibil-ity was much less. There was no flow intothe limestone pool, but it was dischargingabout 0.1 m3/sec of relatively clear waterat 27.7°C. Thus, the limestone pool is re-ally an estavelle. Sometime during the pre-vious few days, the water level inside theSierra de El Abra, or more particularly in-side the cave system behind the limestonepool, must have risen sufficiently to reversethe flow.

Residents at the Hotel Taninul reportedthat a very large flood had occurred on the27th, but the flow had waned on the 28th.By the 29th, the discharge and visibilitywere approaching base flow conditions.

Thus, the Sulfur Pool yields a rapid, sharpresponse to rainfall events. The chemicaland temperature curves had a broader re-sponse pattern, like the Coy and Choysprings. During the period from June 18 toJune 29, the crude hydrograph of theTaninul Sulfur Pool appears to be very simi-lar to the Choy hydrograph (Figure 5.6). InJune 1972, flooding at the Sulfur Poollagged behind the Choy (see Table 5.1 andFigure 5.7). On June 13 (also June 4), theSulfur Pool was at base flow, hot and sul-furous, and there was no sign of previousflooding, whereas the Choy had been atflood stage for several days. However, onJune 15 and 17, the Sulfur Pool was inflood, and its chemistry and temperaturewere again severely affected by the El Abrawater.

Figure 5.11. Effect of flooding on chemistry of Taninul Sulfur Spring, June–July 1971.

88


5.4. Summary

An extensive program of water chemi-cal analysis had the following objectives:to determine the chemical character of karstwaters in the region, especially the springs,and to use the water chemistry as a tool forhydrologic analysis, particularly hydro-graph separation. Samples of swallet wa-ters, cave drips, and cave lakes were con-fined to the Sierra de El Abra, whereassprings were studied wherever they couldbe located, but especially along the easternedge of the Platform. The El Abra cavewaters were found to be of calcium bicar-bonate type, with only small amounts ofMg and SO4. However, three types ofsprings were found: (1) small to largesprings of calcium-bicarbonate type, gen-erally having only small amounts of Mgand SO4, (2) large springs of calcium-magnesium sulfate type waters (with bicar-bonate also) that are somewhat warmer than

the type 1 springs; and (3) thermal springs,each having a different chemistry. The Ca,Mg, and SO4 concentrations at the Coysprings were 341 (as Ca, not CaCO3), 51.1,and 836 mg/l for a sample late in the dryseason. Several spring samples were testedfor Na, K, and Cl, but these were found tobe important constituents only for theTaninul Sulfur Pool (thermal spring).Hence, several of the springs discharge alarge quantity of water that has had pro-longed contact with calcite, dolomite, andgypsum (or anhydrite), but apparently thereis little or no salt in contact with the ground-water flow systems. The chemical analysesobtained in this study (Table 5.1) indicatethat all of the spring water analyses (andperhaps the other types) published byHarmon (1971) are suspect: many of thebicarbonate and sulfate values he reportedare erroneous, some by as much as 300%;the very high bicarbonate concentrationscaused the calculated carbon dioxide par-

tial pressures and the saturation with re-spect to calcite to be excessively high;hence most of the conclusions are incorrect.

Wet season floods cause large changesin the physical and chemical characterisitcsof the high sulfate springs. The responseof the chemical waves lags somewhat be-hind the discharge hydrograph or floodwave because some water in the phreaticzone must first be expelled. The concen-trations of dissolved species become greatlyreduced and approach those of the calciumbicarbonate type springs. It takes severalweeks to a few months, following a largeflood, for the water chemistry to approachits preflood state, whereas the dischargereturns to base flow much more rapidly. Thecorrelated behavior of magnesium, sulfate,and temperature suggest that there are twosources of water for the high sulfate springs.Several lines of evidence indicate that someof the discharged flood water was derivedfrom the causal precipitation event.

89


6

THE CAVES

6.1. Introduction

Caves in the middle latitudes have beenextensively explored, surveyed, and stud-ied scientifically. There are many thousandsof caves known in the United States, andthe karst and caves of Europe are famous.In the tropics, however, although caves areknown to exist in many areas, there are rela-tively few data available on their numbersand general character, and even less knowl-edge of their geomorphology and genesis.

The objectives of the cave studies aretwo-fold. The first is to provide maps, fairlydetailed descriptions, and geomorphicanalyses of tropical caves in a high reliefarea. The second is to study the hydrologyof the presently active caves and to developa model for all the caves of their differentforms as a function of their hydrology orpaleohydrology and geomorphic setting.With only two exceptions, the caves stud-ied were in the Sierra de El Abra. However,the author is familiar with a number ofcaves in the remainder of the region, andbrief mention of their character will be given.Each cave will be treated as a separate en-tity; a map, a general physical description,and substantial geologic and genetic infor-mation will be given. The caves will begrouped into the same physiographic as-semblages as the surface geomorphic fea-tures in Chapter 3, i.e., the east face, thecrest, and the west flank and west margin.In Chapter 7, a summary and analysis ofthe caves in each physiographic zone isgiven, and they are placed in the context ofa broader hydrologic-geomorphic model.

Scientific study of the caves of Mexicohas been very sparse. Previous work in theregion of interest is limited, but greater thanany other region in Mexico, with the pos-sible exception of Yucatan. A few earlyobservations were made in connection withbiological studies, especially of blind fishcaves. Bonet has published two reports,one on the Xilitla area (1953b) and theother on the Sierra de El Abra (1953a), thatcontain relatively simple maps, tempera-ture data, descriptions, and geomorphic

interpretations of some horizontal caves (aswell as the surface). These papers providedthe beginning of cave and karst research inthe area, and Bonet correctly interpretedmany morphologic features and controls.However, his studies were hampered bylack of equipment and techniques for de-scent of pits so that the caves beyond couldbe observed, by too great an emphasis onDavisian concepts of deep phreatic solu-tion and the karst cycle of erosion (Davis,1930), and by a relatively small sample ofcaves. In the Sierra de El Abra, he observedcaves along the east face, along the west-ern margin, and in the passes; thus most ofthe range remained unknown. In the early1960s, members of the Association forMexican Cave Studies began exploring andmapping caves in Mexico. By the time theirfirst bulletin was published (edited byRussell and Raines, 1967), most of Bonet’scaves had been visited and some of theswallet pits had been descended and exten-sive passages found, but there was notmuch new surveying or many new caveslocated. The bulletin contained eight maps,brief descriptions of twenty-four caves inthe El Abra range (caves in other areas werealso described), and a simple geomorphicmodel of the caves.

Since 1967, exploration has greatly in-creased the total length of known caves inthe El Abra to its present value of about 45km, most of which have been surveyed.About 19 km were mapped as part of thisstudy, and a few existing maps were im-proved by resurveying or additionalsketching of detail. In addition, Neal Mor-ris (personal communication) is preparinga book that will give an account of explo-ration activities in recent years and manymaps and descriptions not included here;he has made available maps of some caveswhich were studied in this research.[Morris’s book was never published—ed.2003.] Two caves that were not visited areincluded because of their importance, andsome others that were visited are excludedbecause they were similar to others describedhere. The cave surveys were conducted by

measuring the bearing, inclination, and dis-tance between successive stations, whilesketches were made of the passage plan andcross section and sometimes the profile. Alegend of map symbols is given in Appen-dix 2. The survey data were normally pro-cessed and plotted by a CDC 6400 com-puter. Elevations were obtained for manyof the caves with a Pauling altimeter thatwas graduated in 5 foot (1.52 m) increments.Some of the elevations were measured in areconnaissance manner in conjunction withother work where very long traverses werenecessary, and they are considered onlyapproximate; others were measured moreaccurately. Mitchell et al. (1977) have re-cently obtained elevations for most of theblind fish caves and claim great accuracy.The data collected here agrees in some in-stances within a few meters, but others havelarger differences. Where available, theirelevations will generally be used. Somecaves that are particularly important forwater-level information should be locatedby a transit survey.

The cave studies have emphasized planand profile characteristics, hydrology if thecave is active, geomorphic features, pas-sage shape, the more important structuraland stratigraphic features, chemical andclastic sediments, and relation of the caveto surface. Because it was necessary to sur-vey so many kilometers of passages, verydetailed studies generally have not beenpossible. The general geomorphic and hydro-logic aspects of the caves are documented,but insufficient structural, lithologic, andgeomorphic information have been gath-ered to propose detailed speleogeneticmodels for individual caves. Aside from theobservations and maps, the results will bea characterization of cave features and de-velopment in each geomorphic setting inthe El Abra, and a hydrologic model. Someprevious studies which have been useful forinterpretation of cave geomorphology areDavis (1930), Bretz (1942), Ford (1965,1968, 1971), Palmer (1969, 1972, 1975),Pohl (1955), Lange (1960), Ewers (Ph.D.thesis in preparation [1982]), Jennings

91


Figure 6.1. Map of Cueva del Nacimiento del Río Choy.

92


(1971), Davies (1960), and G. Deike(1960). The general locations of caves andsprings are shown in Figure 3.1.

6.2. Caves of the East Face of theSierra de El Abra

6.2.1. Cueva del Nacimiento del RíoChoy

The Cueva del Nacimiento del Río Choyis a short but grand cave 3.5 km north ofthe south El Abra Pass. Though oftenthought of as two caves, it is in reality one,having an upper and a lower chamber (orpassage) connected by a 62 m shaft (Fig-ure 6.1). The upper cave, especially the topof the shaft where the river may be heardbelow, is used as a religious site by somelocal people.

This cave is particularly important be-cause it is the site of a first magnitude springand because its geomorphic features pro-vide considerable information about karstflow systems in the range. Normally, the flowis 2 to 4 m3/sec (Plate 6.3). Occasionally,

and deposition of very minor amounts ofchemical sediments and guano. Phreaticfeatures extend up to about 10 m below thePit Entrance (elevation approximately 132m), where any higher evidence has been re-moved by collapse. Depth measurementsof the lake show that the deepest point co-incides with a water boil seen during floods(Figure 6.1). The floor of the lake slopesrather uniformly down to the deepest point.Thus, water presently rises at least 26.5 mto resurge. Geomorphic observations sug-gest that in the past, when the coastal plainstood at higher levels, water climbed to theMiddle Entrance and the Upper Cave En-trance to resurge. If this interpretation iscorrect, then a phreatic rise of lift of at least120 m occurred. A group of flat-toppedgravel capped hills of Méndez shale lie afew kilometers east of the Choy spring. Thetops of the undisturbed hills are about 77 melevation and are remnants of a paleocoastal plain which probably correlateswith the middle resurgence level.

6.2.2. Las Cuevas de Taninul Numeros 1 and 2

Figures 6.2 and 6.3 show the plans andprofiles of springs and caves near HotelTaninul, two kilometers south of the southEl Abra Pass. Two springs occur at the baseof the scarp. The Taninul Sulfur Pool is athermal, sulfurous spring with a permanentflow. It is an artificial pool except for the

Plate 6.1. View from inside Choy cavelooking out the lower or Resurgence Entrance(note people for scale). Water from the risepool flows over a boulder dam created byceiling collapse to form a skylight. The wallsdisplay broad phreatic pockets and muchsmaller reef cavities (?). Note the steeplyinclined joint.

Plate 6.2. Phreatic ceiling pocketing and high phreatic domes at El Choycreate a highly irregular ceiling. This view from Middle Entrance also slowsthe rise pool, which is at least 26.5 meters deep, and a major “facies” planethat probably provided the route for water to reach Middle Entrance when itfunctioned as a spring.

floods crest at more than 100 m3/sec, completely filling the lowerentrance (river exit; Plate 6.4).During flooding, the lake surfaceoften rises more than 2 m. Organicdebris is often trapped against theback wall by the circulation pat-tern. Collapse of part of the ceil-ing over the lake has created alarge skylight, and the bouldershave formed a dam which raisesthe lake level (also the base level)about 1.5 m. Boulders may alsohave partially blocked the sourceconduit, which would impede theflow and access for scuba divers.

The smooth, blind domes, thewall pocketing, the uniform wallsof the large shaft, and the irregu-lar solutional ceiling and bedrockfloor are proof of solution underwater-filled conditions (Plates 6.1and 6.2). Joints were the princi-pal loci of ground-water flow andsolutional enlargement in theunbedded reef limestone, butsome surfaces separating sub-facies of the reef may also havebeen used (Plate 6.2). Modifica-tions of the original solution caveare slight and include partial col-lapse of some of the phreaticdomes to form skylights andbreakdown, thin accumulation ofentrance talus in the upper cave,

93

AMC

S Bulletin 14 — C

hapter 6

Figure 6.2. Maps of Cuevas de Taninul Numbers 1 and 2 and the Taninul springs.

94

AMC

S Bulletin 14 — C

hapter 6

Figure 6.3. Profile of Cuevas de Taninul Numbers 1 and 2.

Plate 6.4. The Choy in moderate flood, around 35 m3/sec. The water has risenabout 3 m (compare Plate 6.3). Large floods completely fill the entrance.

Plate 6.3. The Choy spring at base flow (3 to 4 m3/sec). The reef facies isunbedded, but has many fractures, other separation surfaces, and occasional“facies” planes. 95


source area, where water appears to risethrough boulders partly blocking a naturalconduit in the limestone. Nearby, and about1.5 m below the Sulfur Pool level, lies anormally stagnant pool, referred to in thisthesis as the “limestone pool.” This pool isreally an estavelle. During the wet season,flood water from the Sulfur Pool often en-ters it and sinks in the center, back into theEl Abra. Sometime later, it reverses the di-rection of flow to become a spring. It islikely that some El Abra water is then dis-charged, as well as the Sulfur Pool water.No bedrock was observed in the limestonepool. More hydrological information isgiven in Section 5.3.

Cueva de Taninul Numero 1

Taninul Number 1 is an inactive caveoriented perpendicular to the scarp; it is afavored place for geologists to see a crosssection of the Cretaceous reef trend. Thelarge entrance area has been developed intoa bar and dance floor. Natural bridges, walland ceiling pocketing, and domes indicatesolution under phreatic conditions. Thecave gradually rises on a bedrock floor anddiminishes in size, becoming typically 1 mwide and 2 to 3 m high, until a narrow bed-rock squeeze is encountered 135 m fromand 12.5 m above the entrance. The solu-tion features and passage pattern of the backportion of the cave are dominantly phreatic,but some 8 cm scallops and passage mean-ders (?) may be evidence of moderate flowvelocity under pressure or of a vadosephase. This is the only cave visited whichhas a reddish earth fill in some wall andceiling cavities. Small joints have con-trolled the passage location and cross sec-tion in the back half of the cave. Note thatthe passage descends at a lesser angle thanthe boundary plane between two subfaciesof the reef trend.

Near the middle of the cave, there is ajoint-controlled Upper Level complexwhich connects with the main passage inat least two places. One of the passages hasa high dome and a 10 m drop in a fissure,which has one wall composed of bouldersheld in place by dirt and flowstone. Thedip-oriented side passage near the entrance,the joint domes, the Upper Level, and theend of the main passage probably weresources of water integrating to form a smallfossil spring approximately at the presentcave entrance.

Cueva de Taninul Numero 2

Taninul Number 2 lies 120 m north andfarther up the scarp than Number 1 and is aconsiderably larger cave. Part of the roofof an east-trending passage has collapsedto give access via the Main Entrance. Per-haps there was an exit a few meters to theeast for water that used to flow through thecave. A breakdown slope descends 6 m toone end of a moderately large passage at70 m elevation. This 6 m wide, 7 m highpassage trends northward parallel to thescarp for 85 m, then turns westward for110 m, becoming still larger. The north-trending portion contains only a thin sedi-ment veneer and in some places none at all,whereas the fill in the west-trending por-tion thickens until a flowstone bank blocksthe passage. The test pit (for phosphate)shown on the profile penetrated 4.5 m offine-grained sediment with a few boulders.The other test pit penetrated perhaps 2 mof sediment, then 3 m more down a narrowslot which may have been an enlarged jointin the original bedrock floor.

At the bend, an east-trending passageslopes upward over sediment and break-down to a small chamber to the north,which connects by a narrow passage to an-other chamber. A thick breakdown pile inthis second chamber blocks what may havebeen a larger connection with the mainpassage, and probably blocks another fos-sil ground-water exit. Both chambers havehoneycomb or spongework walls as described

by Bretz (1942).There are two other unusual side pas-

sages. Fifty meters north of the entrance across-joint(s) has been dissolved outforming an embayment into the entire pe-riphery of the walls of the passage (sectionA-A′-A″). The bottom slopes down to an18 m fissure drop to a static pool of waterat approximately the same elevation as theTaninul Sulfur Pool. The nature of theembayment and the uniform solution ex-hibited by the walls could only have beencreated by water-filled conditions. Theother passage is a series of short drops fromthe west end of the entrance chamber to amud-floored fissure at the same elevationas the bottom of the other fissure.

The tubular passage cross-section, so-lution sculpture including spongework, andthe fissures show that the cave was in thephreatic or epiphreatic zone during virtu-ally all its formative phase. The lower levelfissures could have been sources or dis-tributaries in this three-dimensional system.Near the bend of the main passage, there isa 1 m high bank of rounded gravel that dem-onstrates a minor vadose phase of deposi-tion and then retrenchment. The relation-ship of the vadose phase to the fissures isnot understood.

6.2.3. Grutas de Quintero

Grutas de Quintero is a locally wellknown cave in the northern part of the Si-erra de El Abra. The entrance is a stoopway

Plate 6.5. The Entrance Passage of Grutas de Quintero. View from point A onthe map toward small passages leading to The Loop. The passage exhibitsphreatic solution features and cross section. Note the prominent facies plane.

96


near the base of the range at an elevationof 179 ± 5 m (Figure 6.4). Inside, the pas-sage enlarges to 9 m wide by 4 m high andcontains some sediment fill. Plate 6.5, takenat point A on the map, shows phreatic walland ceiling solutional forms and a promi-nent bedding plane that probably was thelocus of passage initiation. The small pas-sages lead into The Loop, which is compa-rable in size with the larger part of MainPassage. Main Passage is a large, tubular,paleophreatic conduit typically 10 to 12 mwide and 8 to 12 m high (Plate 6.6). In sev-eral places large travertine dams haveformed, which occasionally create deep wetseason pools. At 180 m and 240 m fromthe entrance, there are low alcoves slopingdownward from the north side of MainPassage. They eventually become blockedwith sediment, travertine, or water. Beyondthe second alcove, the floor of Main Pas-sage rises rapidly over what is probably amassive fill of flowstone and sediment thatcauses a decrease in ceiling height to 5 m.Three hundred meters from the entranceStalactite Chamber is encountered. The left-hand portion of this chamber slopes upwardto a blind alcove and is barren except forone large stalagmite. The ceiling of theright-hand portion is covered with stalac-tites, and the next 60 m of Main Passage iswell decorated. From Stalactite Chamberto The Pit, 200 m beyond, the passage ishorizontal and has a slightly lower averageceiling height because of a deep fill of fine-

grained sediments. The fill is 15 m thick ina test hole dug by miners. The Pit is 12 to15 m deep and marks the end of Main Pas-sage. It is not known how much of it isformed in bedrock versus how much is inthe sediment fill; the back wall is definitelybedrock.

A hole at the bottom of The Pit passesthrough a muddy pool to the back half ofthe cave. Lower Passage is generally muddyand averages 3 m wide and 4 m high. Itextends 530 m, rising slightly along theway, to a small room blocked by flowstone,where the gurgle of flowing water beyonda crack can often be heard. Mud Passage isof similar size and character and is 210 mlong. The floor rises initially about 3 m,declines 4 m in the central area, then climbssteeply over a flowstone slope to a room10 m above the Mud Passage entrance.Three small passages leave the room, oneof which has a 9 m drop ending in a crackand dirt floor.

Unfortunately, Grutas de Quintero wassurveyed by several different groups, andmost of the original notes were lost beforethe elevations were added to the map. Also,the geomorphological studies extend onlyto The Pit and may be impossible to con-tinue because the cave is being mined in aserious way. Nevertheless, many importantaspects of the cave are known and othersmay be hypothesized.

All of the cave back to The Pit is devel-oped in the massive El Abra reef limestone.

Probably all of Main Passage formed ini-tially along the gently dipping undulatingbedding plane shown in the photographsand cross sections and grew to its presentsize in the permanently water-filled zone.Phreatic features include wall and ceilingpocketing, smooth blind solutional domesthat are developed along small fractures,and a tubular passage cross section. An al-timeter traverse of Main Passage was madeto obtain approximate elevation data. Fromthe entrance, the passage rises 8 m up thedip in 100 m, then turns north along thestrike for 80 m, and then turns up dip againto Stalactite Chamber 18 m above the en-trance. The remaining passage to The Pit isapproximately along the strike at about +15m. Thus, the bottom of The Pit is about thesame elevation as the entrance. The pas-sages beyond are thought to be within theelevation range of the first half of the cave.The dip is 5 to 6° E in the Entrance Pas-sage, and becomes less in Main Passage.

With the exception of a drip basin belowa dome at point C, any vadose solution fea-tures that may be present are buried. As thewater table gradually lowered below thecave, the cave’s hydrologic functionchanged from a phreatic trunk conduit toan intermittently active vadose channel ofsmall discharge. The principal vadose ac-tivity has been the deposition of traver-tine dams, flowstone, and fine-grained sedi-ments. At points B, H, and J, “phreatic”pocketing of stalagmite and flowstone showthat the details of the cave’s later historyare complicated. In the area of point J, thefollowing phases or events are believed tohave occurred: deposition of massive sta-lagmites, severe phreatic (or paraphreatic)re-solution of the stalagmites, another pe-riod of growth of large stalagmites and sta-lactites, and finally a later (present?) phaseof deposition of small stalactites. From Dto E, the bedding plane has supplied thewater for the flowstone deposits.

The relationship of Grutas de Quinteroto the topography of the El Abra range isshown in Figure 6.4. From near the base ofthe scarp, the cave penetrates 240 m per-pendicular to the face to a point directlyunderneath the top of the scarp, then fol-lows the strike of the range southward for600 m. It was almost certainly a spring-cavesystem, whose exit was approximately atthe present entrance. Phreatic features inStalactite Chamber reach an estimated 30 mabove the entrance, or an elevation of about210 m. This suggests that water may haveflowed upward 40+ m from the presententrance along joints or along the reef front

Plate 6.6. The Main Passage in Quintero, looking into the cave from point F. Itis an excellent display of phreatic features, including tubular passage shape,pocketing, and blind domes. The bedding plane probably controlled the locusof passage development.

97


to an old coastal plain level at 220+ melevation (indicated by the capped hill ofMéndez shale directly in front of the cave,which is capped by a freshwater limestone;see Figure 6.4 and Plate 3.1). The patternof the cave is interpreted to be that of anintegrating system. Water must have cometo Mud Passage and Lower Passage fromsinks and crevices on top of the range 150m above, flowed up The Pit, down MainPassage, and up to the fossil spring. Thethree alcoves on the north side of MainPassage may also have been sources ofwater. Alternatively, the alcove off Stalac-tite Chamber may be just part of a loop tothe middle alcove. The contrast of passagesize between the front and back halves ofthe cave is puzzling and demands furtherstudy. Among the possible explanations area change of limestone facies or othersources that may be buried below The Pitor the deep sediments in Main Passage. Theback half of the cave is generally wet andmuddy, indicating it is still somewhat ac-tive hydrologically, but the water now evi-dently seeps down to unknown passages.Drips and possibly the alcove sources (?)create very small intermittent flows in thefirst half of the cave.

6.2.4. Cueva de las Cuates

This cave is located in the high centralportion of the El Abra range. It is namedfor its twin entrances, which are near thetop of the east face at an estimated 500 to600 m elevation. A steep climb over boul-ders leads to the entrances, which are in arecess formed by collapse of the cave (Plate3.6).

The main cave is short (only 102 m),but high. It appears to be formed along asingle large, nearly vertical joint. As indi-cated by the pocketing and smooth solu-tional forms, smooth blind domes, andirregular ceiling profile (Figure 6.5), thecave was formed under water-filled condi-tions. The floor is covered with dry guano.At the back of the cave, a small hole leadsto a drop (unclimbable?) to a small roomwhich was not explored. A flowstone natu-ral bridge at the entrance indicates that ahigher sediment floor existed in the pastand that material is somehow being re-moved, possibly by vadose trickles.

6.2.5. Cueva de la Ceiba

This description is adopted from a de-scription plus map by John Bassett (1971),and from personal discussions with him. Figure 6.5. Map of Cueva de los Cuates.

The cave was not visited for this study, butis included because it has an important hy-drologic implication.

The cave is located about half way upthe eastern scarp in the central portion ofthe range at an estimated 300 m elevation.The high arched entrance is typical of theeast face caves. A 30 m long passage leadsto the main part of the cave, which consistsof a large chamber 160 m long, varyingfrom 6 to 23 m wide and having an irregu-lar ceiling with many high domes (Figure6.6). This chamber is developed along asingle large fracture parallel to the east face,which may be seen extending at least 30 mvertically at the north end of the chamber.The floor in the first half of the chamber isrelatively level except for a broad bedrockmound. The floor of the back half risesalong a bedrock incline to 31 m above theentrance (see the cross sections on the map,which are arranged to serve as a crude

profile). The cave appears to end abruptlyin a crack a few centimeters wide, but a 40m dome may be a continuation.

In the middle of the chamber, a pit se-ries descends 88 m to a steeply slopingroom which is blocked at the bottom bysediments. An eroded mudbank containsseveral layers of large mineralized bones.The pit series is developed along a swarmof microfractures, rather than large joints,and displays smooth forms characteristicof uniform solution under phreatic condi-tions.

Vadose solution features are completelyabsent, and current-deposited sedimentarystructures are minor. The only changes thathave occurred since the cave was drainedare the formation of some flowstone andstalagmites beyond the pit, a minor amountof collapse, and deposition of a smallamount of sediments, mostly washed downthe pit.

98


6.2.6. Ventana Jabalí

Ventana Jabalí is located 18 km NW ofTamuín. The large entrance is the most con-spicuous of the east face caves (Plate 6.7)and lies about two-thirds of the way up thescarp at about 350 m elevation. The cavehas been mined extensively for phosphateand guano, and it is well known among cave

explorers for its spectacular size and sky-lights. It deserves to be equally well knownamong karst geomorphologists for its deepphreatic solution features. Figure 6.7 is amap of Ventana Jabalí, modified slightlyfrom one published in Russell and Raines(1967, p. 77), Figure 6.8 is a profile drawnin this study, and Plates 6.7–6.10 showvarious views of the cave.

The cave is a single immense passage380 m long and 15 to 20 m wide, orientedgenerally perpendicular to the scarp. It liesentirely within the unbedded reef facies ofthe El Abra limestone. An extraordinaryceiling profile was created by solution up-ward along joints under completely water-filled conditions (see profile). Near themiddle of the cave a phreatic dome extends

Figure 6.6. Map of Cueva de la Ceiba.

101


Figure 6.7. Map of Ventana Jabalí.

Figure 6.8. Profile of Ventana Jabalí.

102


Plate 6.7 (upper left). View out the high archedentrance of Ventana Jabalí to the coastal plain 200+ mbelow. The equipment is left from mining operations.The ceiling and walls are of phreatic origin.

Plate 6.8 (lower left). The immense chamber ofVentana Jabalí. A very high phreatic dome hascollapsed to allow summer sunbeams to enter from153 m above. A person may be barely visible to theleft of the lighted floor. One beam strikes a naturalbridge 70 m above the floor. Sharp exterior phreaticsolution edges may be seen 35 m above the floor, andthere is a high fracture to the right of the bridge.

Plate 6.9 (above). The high dome and bedrock naturalbridge in Ventana Jabalí. Phreatic features extendupward about 140 m.

103


floor. If detailed sediment studies were car-ried out, there would probably be much in-formation of geomorphic and climaticsignificance (Plate 6.11). Underneath theskylights, rocks and washed-in soil accu-mulates in mounds. Away from the sky-lights the sediments are almost as thick,and may include remobilized entrance orskylight-facies sediments (moved by animalsor vadose trickles), breakdown, guano,windblown dust, and possibly chemicalprecipitates in the sediment matrix. Beyondthe big dome, the floor climbs steeply overwhat appears to be partly a bedrock slopeto 52 m above the entrance and then de-scends slightly. Beyond the rise, guano de-posits have accumulated to a depth of at

Plate 6.10 (above). The rear portion ofVentana Jabalí, showing irregular ceilingprofile, sharp exterior angles, and other water-filled solution forms. There is at least 10 m ofguano fill.Plate 6.11 (below). A 5 m sediment sectionexposed in a test pit near the entrance ofVentana Jabalí. There probably is a veryinteresting history recorded in the sedimentshere and in other caves in the range.

upward about 140 m and is span-ned by a bedrock natural bridgeabout 70 m above the floor (Plates6.8 and 6.9). The few meters ofoverburden above the dome hascollapsed in several places to formsmall skylights 153.4 m high. Thephreatic ceiling profile of the caveis mostly extant, because there hasnot been much collapse.

In the front half of the cave,sediments have accumulated sothat the floor varies only a fewmeters in elevation. Mining opera-tions have removed a large volumeof the fill, up to 7 to 8 m in places,and have not reached the bedrock

Figure 6.9. Map of Sótano de Escalera.

104


least 10 m. Near the end, there is an un-usual bedrock partition that rises 6 m abovethe floor and completely crosses the pas-sage. The cave terminates abruptly at a highbedrock wall.

The source(s) of the water that made thisimmense cave are unknown, but it is likelythat the conduits are buried. The solutionpocketing, the sharp exterior corners oredges (Plate 6.10), the bedrock bridge andpartition, and the highly irregular solutionceiling are conclusive evidence of forma-tion under water-filled conditions. Phreaticwaters circulated through a vertical rangeof at least 140 m.

6.2.7. Other east face caves

There are many other cave openings atall elevations along the east face that havenot yet been examined. The caves betterknown locally are included here, but it iscertainly possible that important discover-ies and scientific information will be foundin the unexplored caves. Also, there areseveral active springs not described here;some of these will be referred to whereappropriate in the discussion in Section 7.4.

6.3. Caves on the Crest of theSierra de El Abra

6.3.1. Sótano de la Escalera

Sótano de la Escalera lies about 1.5 kmnorth of the south El Abra Pass, on the rela-tively flat top of the range at an estimated

Plate 6.13. View toward the entrance of Cueva Pinta from near the 15 m deeptest pit in sediments. The photo shows smooth phreatic sculpturing, sharpexterior ceiling angle, flat sediment floor, and the entrance debris ramp.

Plate 6.12. The entrance to CuevaPinta, showing the original solutionmorphology and the thin roof whichhas collapsed.

300 m elevation. The entrance is a 28 mdeep pit created by ceiling collapse into alarge void (Figure 6.9). A rubble slope de-scends into the center of the cave, whichcomprises a single chamber 75 m in diam-eter and 35 m high. The upper portion ofthe walls and the ceiling appear to havebeen formed by upward stoping. They showslight solutional modification of thestairstep shape up to about the level indi-cated on the profile. The lowest part of thefloor is flat and composed of washed-insoil, guano, and some very large stalag-mites, several at least 12 m tall. They showthat the floor of the chamber has been stablefor a long time. No signs of stalagmite re-solution were observed.

6.3.2. Cueva Pinta

Cueva Pinta (Figure 6.10) is located eastof Sótano de Soyate, very near the top ofthe monoclinal western flank of the range.The entrance lies at an elevation of roughly465 m, and, as in so many of the caves onthe top of the range, it has been altered orwas initially formed by collapse into an oldphreatic chamber (Plate 6.12). A breakdownramp slopes down to a large L-shaped roomthat has more than a hundred Indian paint-ings, mostly handprints. The floor is flat,aggraded by sedimentary processes thoughtto be principally the washing-in of soil fromthe entrance area. Guano and possibly windblown dust may be important componentsof the fill. At the north end of the chamber,

miners have dug a test pit downwardsthrough at least 15 m of sediment; hencethe room is actually much larger than itappears. Plate 6.13 shows solution featuresin the room that were created by water-filledconditions and also shows part of the en-trance debris cone in the background. Thecave contains abundant stalagmite and sta-lactite deposits, many of which have beenpartly to almost totally redissolved. Fromthe east end of the room a crawlway overfill leads to 90 m of small passage that endsat a 27 m blind pit. Other than size, thepassage has the same features as the room.

6.3.3. Sótano de la Cuesta

Sótano de la Cuesta is located at thewestern edge of the crest in the high cen-tral portion of the range. This cave has notbeen personally investigated, so only a briefdescription can be given. In the bottom ofa small doline, two small holes and a largerone 18 m across drop 174.3 m into an enor-mous chamber (Figure 6.11; map providedby Neal Morris of Austin, Texas; note allmap data are in feet). The chamber mea-sures 327 m long, about 95 m wide in thecentral area, and for the most part 165 to200 m high! In map view, it is approxi-mately trapezoidal. The total depth is 217m. The floor is covered with flowstone,breakdown, and finer-grained sedimentsthat probably include guano. From photo-graphs (by Peter Strickland, Austin, Texas),it appears that the upper walls and the

105


ceiling exhibit only angular forms causedby breakdown processes, but it was not pos-sible to tell whether the lower walls wereformed by solution or breakdown. Note thatthe southwest wall is fault-controlled. It issuggested that a major active or formerlyactive cave channel lies somewhat belowthe present floor level. The fault and prob-ably associated fractures formed a localizedweakness in the cave roof, which has un-dergone gradual upward stoping for a verylong time. Presently, the roof is withinabout 15 m of the land surface. In a shorttime, geologically, this chamber will be-come a collapse doline, similar to theCaldera (Figure 3.4) and many others inthe El Abra range. Knowledge of this cavenow leaves no doubt about the origin ofmost of the large El Abra sinkholes. Theroom has a volume estimated to be between1.5 and 2.5×106 m3, by far the largest yetdiscovered in the El Abra range. The entirecentral floor area beneath the main cham-ber may still be actively subsiding.

6.3.4. Sótano de los Loros

Sótano de los Loros lies about 750 mwest of the crest of the east face, near LaCueva de Ceiba. The 15 by 30 m entrance

is located on the top of a low hill and wasformed by collapse of the thin roof over alarge phreatic dome or chamber. At thebottom of the 50 m entrance pit, a passage,typically 6 m high and 10 m wide, extends200 m in a north-northwesterly direction(Figure 6.12). Wall and ceiling solutionpocketing and rounded passage cross sec-tion indicate phreatic development. Thecave is profusely decorated with forma-tions, particularly the floor, which is mostlycovered by flowstone and some fine-grained sediments. The original passagemay have been much larger, but the bed-rock floor is hidden entirely by sedimen-tary fill that may be quite thick. Possiblythis cave is related to Hoya de Zimapán,which lies about 300 m to the west. Thepassage is terminated at a point where theceiling lowers and the clastic sediment-flowstone fill appears to thicken.

6.3.5. Cueva de Taninul Number 4

Taninul Number 4 is located on the northside of the south El Abra Pass and 200 to300 m from the east face. It is composed ofone large horizontal passage about 140 mlong plus many smaller passages and fis-sures, and the cave is well known for its

numerous entrances and skylights (Figure6.13). This cave is believed to be represen-tative of the conduits that carry or havecarried slowly moving phreatic waters inthe El Abra range. Phreatic solution formsare ubiquitous. The cave is developed atthe western margin of the reef facies wheresome faint horizontal bedding begins toappear. Water sought every available joint,fracture, and minor sedimentary parting,creating a highly irregular conduit cross-sectional area and blind fissurelike alcoves.Structural control by NNE-SSW and E-Wjoint sets may be seen on the map. Most ofthe cave is developed in one massive bedat least 6 m thick. Along the main conduit,solution has proceeded upward to the topof this bed, and many low arches have beencarved into a 1 m thick and less solubleoverlying bed. Where joints have permit-ted (which is quite often), the latter bed hasbeen breached, and domes have beenformed in another massive bed that extendsto the surface. The ceilings of several ofthe larger domes have collapsed, creatingmany skylights with piles of breakdownbelow them. Sediments cover all the floorexcept in the small bypass loop, so the char-acter of the original floor is unknown. Acareful study of the sediments and pollen

Figure 6.10. Map of Cueva Pinta.

106


here and in many other east face or crestcaves could be very informative. Includedare the collapse skylight facies, fines withsome stones near the entrances, and thickguano deposits with rodent or bat bones inthe rear. The cave may be older than andgenerally unrelated to the adjacent valley.The continuations of the conduit are prob-ably buried.

6.3.6. Cueva de Tanchipa

Tanchipa is located marginally below thewest crest in the central part of the range, afew kilometers south of Sótano de laCuesta. The ceilings of two rooms of thecave have collapsed to form sinkholes (Fig-ure 6.14; map provided by Neal Morris;note all data are in feet). North from onesink, there is an old phreatic room thatslopes downward over a thin veneer of

talus to a complex series of fissures. Thesefissures (Plate 6.14) form a vertical mazeand have well developed phreatic pocket-ing. They are controlled by longitudinaljoints related to the N-S fold at the westernedge of the crest. It is not known whetherwater formerly went down or up the pas-sages, but they were clearly part of a deepphreatic flow system of at least 157 m ver-tical extent. For some reason, the fissuresare smaller near the bottom of the cave.Perhaps the system diffuses into many smallpassages. The cave is well decorated withformations, but has very little clastic sedi-ment.

6.3.7. Hoya de Zimapán

Hoya de Zimapán is located 1 km westof the top of the east face of the El Abra, inthe high central portion of the range. Here,

as in many other places along the range,the eastern portion of the crest has twostrike oriented ridges and an interveningvalley estimated to be 75 m deep in thisarea. The entrance to the cave lies in thenorthwestern side of a doline, about 50 by70 m across, near the western base of thewesterly ridge. The doline is believed tohave formed by ceiling collapse of a largephreatic passage (Plate 6.15).

The view into the entrance passage thatgreets the explorer is breathtaking (Plate6.16). A steep breakdown slope leads downto an almost perfectly round tunnel fully30 m in diameter (Figure 6.15). Two mas-sive formations partially block the passage.The second infills the lower 20 to 25 m ofpassage (Plate 6.17); water ponded by ithas opened a narrow drain at the base.

Immediately beyond the second stalag-mite, the passage turns on end and descends

Figure 6.11. Map of Sótano de la Cuesta (provided by Neal Morris; all data are in feet).

107


room, here named the Zimapán Room(Plate 6.18). Initially, there is a mud-flooredpassage or alcove 10 m high and 40 m widebefore the main chamber is reached. Thetotal room is 225 m long, up to 95 m wide,and an estimated 40 m high. As may be seenin the photograph, the lower walls havesmooth solution surfaces, whereas thehigher walls or roof exhibits both solutionpocketing and the steplike form caused bythe breakdown process. There may be asmall dome or inlet shaft in the center ofthe ceiling, but to date no one has had suf-ficient light to determine if this is correct.The floor is covered with gour pools cre-ated by a drip source near the middle ofthe room. The pools contain some mudwashed in from the entrance. There are alsoseveral large stalagmites, some of whichhave been extensively redissolved (plates6.19 and 6.20). These chemical depositsand the mud have aggraded the floor, con-cealing all information there may have beenconcerning the original floor and theamount of breakdown. Also, the continua-tion of the cave has unfortunately beenburied.

The volume of the Zimapán Room isestimated to be between 175,000 and250,000 m3. Thus, it may be comparablein size to the famous Big Room in CarlsbadCaverns. The total depth of the cave is 320m, and the horizontal length is about 480m. A detailed discussion of the history ofthe cave is given in Section 7.4.2.

Figure 6.12. Map of Sótano de los Loros (provided by Neal Morris; all dataare in feet).

Plate 6.14. Phreatic fissure in Tanchipa. It isnow well decorated.via three large shafts to a depth of 300 m.The shafts maintain the smooth, tubularphreatic shape and sculpturing of the en-trance passage, except where thick flow-stone deposits have formed. The lowershafts may be a little smaller in diameterthan the entrance tunnel. Between the firstand second drops, there is a long slopewhere the flowstone accumulation is so

thick that it nearly blocks the passage atone point. Other thick deposits drape overthe top of all three drops and cover the floorbetween the second and third drops. Thewalls of the shafts and connecting passagesshow no sign of past vadose activity, butthe flowstone slopes have been extensivelyredissolved by vadose flows.

The third pit drops into the end of a huge

Plate 6.15. The entrance to La Hoya de Zimapán. Collapse of thepassage produced a doline and the breakdown ramp (person at top forscale). Note the broad phreatic pockets on the wall and ceiling.

108


6.3.8. Sótano and Cueva de los Monos

Cueva and Sótano de los Monos are lo-cated east of Sótano de la Tinaja (LosSabinos area) near the middle of the crest.They lie in the northeastern corner of aclosed basin of about one-fourth squarekilometer area. The datum for the survey(Figure 6.16) has an elevation of about 465m.

Cueva de los Monos is an old phreaticcave at the top of a big pit. It is explorablefor about 60 m and consists of two rooms.The lower one, Crystal Chamber, has sev-eral pockets of large calcite crystals (6 cmacross) for which the room is named. Itappears that the cave was formed underwater filled conditions, then calcite depos-ited on the walls and in pockets (possiblyunder water because of the size and clarityof the crystals), only to be later partiallyredissolved, forming “phreatic” pockets inthe crystals. Some time later, large massesof flowstone and stalagmite were deposited.Some of these have petroglyphs of mon-keys, coatis, deer, and people. The namemono means monkey, but it also meansmimic—the glyphs mimic animals andpeople. Holes in the floor below the glyphsconnect to the bottom of the main pit. Up-per Chamber is reached by climbing up asteep slope at the back of Crystal Chamber. Itis well decorated by large old stalactites thathave superb examples of re-solution (Plate6.21) and by a younger generation of smallstalactites. The room is 2 to 3 m high andmay continue beyond the sediment blockat the far end. There are several domes (in-lets?) that may have supplied the materialfor at least 1.5 m of clastic fill which has Figure 6.13. Map of Cueva de Taninul Number 4.

Plate 6.16. The nearly circular entrance passage toZimapán. It is 30 m in diameter. A large column partiallyblocks the passage.

Plate 6.17. A 20 m high stalagmite barrier across the end ofthe entrance passage in Zimapán. Ponded water hasdissolved a hole through the base.

109

AMC

S Bulletin 14 — C

hapter 6

Figure 6.14. Map of Cueva de Tanchipa (provided by Neal Morris; all data are in feet).

110


Figure 6.15. Map of La Hoya de Zimapán.

111


Plate 6.18 (above). The Zimapán Room, looking westacross 140 m of the room from the northeast corner. Thelower wall show solution pocketing, but the high ceilingexhibits both solution pocketing and steplike forms caused bycollapse.

Plate 6.19 (below left). Large stalagmites that haveundergone substantial “phreatic” resolution. There is aprominent solution notch around the base. Most of thefloor of the Zimapán Room is covered by gour pools (dryat this time), as shown here.Plate 6.20 (below right). Another stalagmite that has beenpartially redissolved in the Zimapán Room and solutionfeatures on the wall.

112


thin interbedded flowstone layers and per-haps some guano.

Sótano de los Monos is thought to be amuch younger cave than Cueva de losMonos. Water collected on a surface catch-ment of soil and was pirated down a groupof N-S joints, possibly before a closed ba-sin had developed. The surface of the basinhas gradually lowered, and the arroyo isnow entrenched an estimated 30 m belowthe back wall of the pit (see the profile).The spectacular entrance pit, 15 m acrossat the level of entry and 141.5 m deep (464feet; about 171 m or 560 feet from the highside) is by far the largest feature of the cave.The bottom is larger than the top, and natu-ral bridges partition the pit. It is believedthat the entrance pit is a group of coalesc-ing vertical shafts formed under vadoseconditions. The intersection with Cueva delos Monos may have been fortuitous, orperhaps the same joints also supplied wa-ter to the cave when it was active. Smallbreakdown covers the floor of the pit. Inan alcove at the east end of the bottom, aclimb over a 6 m high bedrock partitionleads via a crawlway to a succession ofsmaller vadose shafts (the Entrance Series)which have many natural bridges. Somewood jammed in domes indicates pastfloods. This route may not be used pres-ently, because most of the water probablysinks in a depression of boulders on thefloor of the entrance pit. Where the watergoes is unknown.

At the bottom of the Entrance Series, at–222 m, the horizontal part of the cavebegins. This portion of the cave is very in-teresting, because the passage forms andprofile are not what one expects in a va-dose cave. Rather than a relatively uniformtunnel or pot-holed canyon passage with acontinuously descending profile, FirstLevel and Second Level often are more likea series of water-filled solution formedrooms connected by short crawlways. Inmany places, solution upward has createdblind chimneys. Another anomaly of theprofile is The Loop, where water first mustrise up a 3 m high bedrock wall to a smalltube with a canal (Figure 6.16), next flowdown into The Loop and back up about 4m, and then continue as a free-surfacestream down The Chute. The strata arenearly flat lying. A combination of struc-tural and stratigraphic features such as the5 cm shale band marked on the profile ap-parently controlled passage initiation inthese levels. Second Level is much largeron the average than First Level and isslightly concave up in profile. It probably

continues beyond the present end, wheresediments rise to fill the passage. It is sug-gested that these levels derived their fea-tures from water ponded in the irregular orconcave up profile. Floods which werelarger than the passage could handle mayalso have had some effect.

In the past, there was a major fill ofcobbles and sand (see cross section of ce-mented fill bridge). Much of the fill hasnow been removed. There also is a poorlydeveloped lower passage, Misery Route,which begins as a hole low on the wall atthe lowest point in Second Level. AtBroussard’s Misery, the passage dissipatesto a few holes along a vertical joint, so thatit is necessary to raise the explorer and pushhim through with his arms outstretched.The passage ends in narrow cracks and asmall flowstone bank. This passage prob-ably was created by a piracy of SecondLevel water. Subsequently, the piracycaused the entrenchment of the SecondLevel fill up to the entrance to Misery Routeand slightly beyond. There continue to besmall flows that enter via the known caveand from several inlet domes. Under presentconditions, the First and Second Levelsmust not flood violently, because delicatesoda straws that are greater than 1 m longfound near the entrance to Misery Routewould not survive. If major floods enter thecave, they must be taken to new lower levels.The Second Level is abundantly decoratedwith gour pools, flowstone, stalagmites,stalactites, and soda straws, especially

where there are active inlet domes.At the present time, Sótano de los

Monos is the only known vadose cave ontop of the El Abra range that has extensivepassages. As a simplified model, it is simi-lar to the shaft-drains of the MammothCave area (Pohl, 1955; Brucker et al.,1972), where the vertical shaft is manytimes larger than its horizontal drainageconduit because of differences in the rateof solution. The total surveyed passagelength of the Monos cave system is 571 m,and its vertical range is 291 m. It is still along way from El Choy, where its watersprobably reappear.

6.3.9. Other caves on the crest

From the results of this study and otherrecent explorations, it is certain that thecrest of the range contains a very large num-ber of caves. Analysis of the caves has pro-vided information concerning the nature ofground-water circulation in the El Abrarange and important findings for theoriesof cavern genesis and karst hydrology (dis-cussed in Section 7.4). There are severalcaves known that are similar to ones de-scribed here, hence their inclusion wouldhave only been repetitious. Other thanMonos, one type not mentioned is the va-dose shaft or lapiez well. The majority, per-haps, of caves known on the range aremerely blind shafts ranging from 10 to 200m deep. They are blind in the sense that nocontinuation that is humanly explorable

Plate 6.21. A very broad and deep pocket in Monos dissolved without disconti-nuity into bedrock and flowstone by a “phreatic” re-solution phase.

113


was found, but water surely flowed onwardthrough cracks or through passages thatmay be blocked. These shafts have not beenstudied closely. Some of them may be oldphreatic fissures or pits, but most probablyare vadose shafts. In reality, joints and bed-ding planes opened by vadose zone solu-tion which are too small to permit humanentry are much more common than the pitsthat can be explored.

6.4. Caves on the Western Flankand the Western Marginof the Sierra de El Abra

As described in Chapter 3 (surface geo-morphology), the western flank of the rangedips relatively gently compared with theeastern escarpment. The removal of theoverlying rocks has exposed the El Abralimestone to solutional processes over a beltas much as 6 km wide. However, the caves,the structure, and the geomorphology ofthis zone are poorly known. A few cavesare included here.

At some places along the western mar-gin of the El Abra limestone outcrop, thestructural geology and the geomorphologyare such that arroyos having catchments onMéndez or San Felipe beds are pirated un-derground a short distance after they passonto the limestone (Section 3.1.3). Thelargest number occur in the Los Sabinosarea, and several of these caves weremapped and studied. An isolated cluster ofthree swallet caves is located about 8 kmnorth-northwest of the village of LosSabinos; one cave from that area is alsodescribed. The location and relationship ofthe caves to surface geology, drainage andgeomorphology can be seen in Figure 3.2.

6.4.1. Sótano de Soyate

Sótano de Soyate lies about half way upthe western flank of the El Abra range, 2km east of the Los Sabinos area swalletcaves. Figure 6.17 is a map and a profilemodified from ones by Elliott (1970) by anew cave depth survey, including use of awire and the inclination for the entrance,by water depth measurements, and by mi-nor alterations of detail. The entrance is avery deep shaft along a major joint or jointsinclined at 83°. It descends 103 m, typi-cally 6 m by 12 m in cross section, to aledge or restriction of the pit, then another92 m as a long fissure to the floor 195 m(639 feet) below the surface. There are anumber of places where the fissure andother recesses extend beyond the reach of

the light from a carbide lamp. Vadose filmshave left a 1 m thick deposit of flowstonefor the lower 150 m of the pit on the “foot-wall,” or wall descended by explorers. Thefloor is covered by small angular break-down that has come mostly from the top20 m of the pit, and by some mud. At thesouth end of the floor of the entrance pit, aseries of travertine dams and slopes drop20 m into the end of a large chamber. Atthis level, the walls and the floor are cov-ered by a thick deposit of mud, and thereare some cobbles washed down from above.A small hole in the floor drops another 11or 12 m (under normal water conditions)to a deep lake that fills the lower part ofthe chamber. The lake is 147 m long and20 to 25 m wide for the most part. It wasplumbed in a number of places approxi-mately along a lengthwise center line, at a

time when the lake was 11 m below the topof the hole. The deepest point measured was53.4 m (175 feet), and no depth less than33.5 m was found. Thus, the bottom of thelake is nearly at sea level and well belowthe El Abra base level. From the ViewPoint, the chamber appears immense, pos-sibly higher than the 34 m (110 feet) givenby Elliott (1970). The total depth to the lakesurface was surveyed to be 233.9 m (767.3feet).

Soyate is an important cave because ofits morphology and because of its presenthydrologic activity. The entrance is situatedat the base of a hill on the east side of amajor valley, now a closed karst basin,which formerly drained southward in amanner similar to the Río Sabinos (see Sec-tion 3.1.3) before the latter was pirated bythe swallet caves (Figure 3.2). Near the

Figure 6.17. Map of Sótano de Soyate.

114


surface in the entrance pit, vadose watershave cut many small drip pockets and sharppinnacles and edges common to verticalshafts and have formed a vadose chute thatreaches about half way down the footwallof the shaft. Rock fall has removed theoriginal solution features of the upper ap-proximately 30 m of the overhanging wall,but for the rest of its depth, the wall exhibitssmooth, usually equidimensional solutionpockets up to a meter or more in diameterand 0.1 to 0.3 m deep. Where they werestudied, the pockets show some beddingcontrol. They also are often asymmetric,like scallops, having a greater deviationfrom the vertical on the higher side, and onthe lowest few meters of the shaft they arestrongly elongated, resembling flutes. Af-ter considerable debate, it is believed thatthese pockets were made by phreatic solu-tion rather than by vadose films. Further-more, the overhanging wall lacks vadosefeatures such as pothole remnants andpebbles or clay in niches and, in contrastwith the footwall, has virtually no flow-stone. The walls above the lake aresmoothly sculptured, there are partitionswith solution holes, and the ceilings abovethe south end of the lake and above the trav-ertine slopes contain phreatic domes.

Studies of the cave during the thesis re-search have shown that it is very activehydrologically, even though it has no di-rect surface catchment (see Section 4.4.2).A stage recorder was placed over the holethat drops to the lake to record water-level

Figure 6.18. Map of Cueva del Prieto.

fluctuations. It was found that dramaticwater-level oscillations occur several timesa year, and that the hydrograph correlatedvery well with the Choy hydrograph.Mitchell et al. (1977) give 293 m for theelevation of the entrance. Hence, the dryseason lake surface must have an elevationof about 59 m. The elevation of the lakesurface, the depth of the lake,and the char-acter of its hydrograph are strong evidencethat it is a water table lake, i.e., part of thepotentiometric surface of the aquifer.

At the north end of the bottom of theentrance shaft, there is a small intermittentsource of water that can occasionally beheard gurgling down to a lower level be-tween rainy spells. Flood water from thissource has cut a very small channel in thefloor, but it is believed that it would haveneither the correct hydrograph shape nor asufficient discharge of water to cause theobserved lake level variations. Instead,flood pulses passing through cave passagesbelow the lake surface cause its level to rise.The lake surface undoubtedly reflects thelocal hydraulic head of the aquifer. How-ever, the mud deposits and the large num-ber of blind fish, which require organicdebris for food (not supplied by bats in thiscave), suggest that one or more of the LosSabinos area swallet caves is (are) also con-nected with Soyate. Thus, the lake hydro-graph may be the complex result of bothlocal and distant events.

There is a lack of evidence to demon-strate that Soyate has undergone significant

vadose enlargement. There is neither anactive nor a dry valley leading to the cave,no degradation of the entrance pit by streamentrenchment as found in the Los Sabinosswallet caves, and no major assemblage ofvadose solution forms within the cave, suchas stream potholes or features of verticalshafts as in Kentucky (Pohl, 1955). Rather,the cave seems to have reached nearly itsfull size under waterfilled conditions. Thus,Soyate is believed to have been part of adeep phreatic flow system. The asymmet-ric pockets on the overhanging wall sug-gest that water leaked in from the surfaceor from side passages along the joint andthen flowed downward to depths more than200 m below an ancient water table. Thissource diminished as karstification of thesurface increased. The hypothesized con-nection with the swallet caves would haveoccurred later in the cave’s history. How-ever, Soyate must always have been con-nected directly with the water table, andmay owe a part of its enlargement to solu-tion upward from below or, more likely, amixing of water from above and below.

6.4.2. Cueva del Prieto

Cueva del Prieto is located in the ex-treme southern part of the El Abra range,about 1 km northeast of El Pujal and per-haps 1 km north of the Río Tampaón. Thiscave is probably the same as one placedinto the blind fish literature as Los Cuatesby Mitchell et al. (1977), but it was called

115


as given here by the ranch owner. The en-trance lies in a collapse sinkhole (Figure6.18) on top of a broad hill or knoll anddoes not capture a surface stream. El Abralimestone is exposed over the surroundingarea. The cave is very strongly structurallycontrolled by N-S joints, and the traces ofthese joints can be easily followed on thesurface by the presence of solution andcollapse features. There are actually twopits in the sink. The easternmost is merelya fissure-like pit that descends to water. Theother pit is actually an entry to the top of afissure cave (see plan and profile, Figure6.18). At the bottom, 22.2 m below, thefloor slopes downward underneath the en-trance sink to a deep pool, where the passageis blocked by breakdown. In the oppositedirection the cave extends at least 100 m.Initially, there is a climb over breakdownboulders, followed by a short drop to a mudfloor. The passage continues relatively hori-zontally and contains abundant formations.Near the end the ceiling lowers, and sev-eral pools are encountered. This passagewas not explored beyond a 3 m drop into adeep pool, but it may continue under amassive formation. As shown by solutiondomes, natural bridges, and other features,the cave is of phreatic or epiphreatic originand was initiated when water levels werehigher in the past. Elevation data given byMitchell et al. (1977) show that the lowestpools are very near the elevation of the RíoTampaón and thus may be at the top of thephreatic zone. Their surfaces may fluctu-ate during the wet season. Re-solution fea-tures in stalagmites were observed up tonearly the level of the entrance.

6.4.3. Sótano de Japonés

Sótano de Japonés is the southernmostof a group of three large swallet caves thatlie 8 km north-northeast of the village ofLos Sabinos and 18 km north of Cd. Valles.Japonés captures a west-flowing arroyo thathas breached a structural high of the ElAbra limestone. The drainage area is about8.6 km2, including a small tributary north-east of the entrance. The initial capture wasdown a large N-S joint in the Main Sink(Figure. 6.19; “S-I”is for shaft number 1,“Level I” for the highest passage level, etc.).Since the piracy, the upper parts of the westand north walls have retreated by collapseand slope processes, so that the sink is nowrather large. However, the El Abra has beenexposed for only a few hundred meters inthe streambed, and the arroyo has en-trenched no more than about 10 m below

the top of the formation at the cave en-trance. These facts and the passage dimen-sions indicate that this cave is muchyounger than the larger of the Los Sabinosarea swallet caves. Presently, another entrance(S-II), developed on a N-S joint completelycrossing the arroyo about 25 m upstreamfrom the Main Sink, captures the majorityof flood waters except for perhaps the larg-est pulses. Primitive captures occur stillfarther upstream.

As can be seen from the map, Japonésis very complex, and there may be quite alot more passage yet unexplored and un-surveyed. Offsets of overlapping passagesand a projected line profile are shown asaides. Nearly all the cave is developed onfour levels (see profile in Figure 6.19).Level I is the highest and least extensive,but includes the present main flood chan-nel (about 3 m in diameter), a smaller up-dip tributary from upstream piracies, and amaze-like alcove off the Main Sink. LevelI quickly plunges down a 19 m drop (S-III)to Flood Route (Level II). This 4 to 5 mhigh flood water passage has a somewhatovate form caused by joints and minor va-dose trenching. Sediments include bouldersto mud and logs. After a gradual descentalong 300 m, a 32 m deep fissure (S-V)descends to Level III, the Maze Level. Thislevel is reached more directly by S-I in theMain Sink. It comprises a labyrinth that isonly partly explored. Passages are com-monly only 0.5 to 1.0 m high, but some ofthe principal routes reach 3 m diameter.Although the idea has not been carefullychecked, it appears that most of this levelis developed at one stratigraphic horizon.The initial flow control may have been abedding plane, but a bed or beds havingsmall, closely spaced joints have had a pro-nounced effect. Special symbols are usedon the map to indicate that in many placesthe walls of the passages are poorly defined.Flood waters have sought every penetrablejoint and enlarged it. Often (more often thanshown) the walls are merely a collection ofbedrock columns remaining between thejoint openings. In other places, loose rocksand mud form the wall. Several extraordi-nary low rooms have developed, such asRoom V, which is more than 100 m in di-ameter, but only 1 m or less high. The netflow direction is southwest, and the largerpassages are oriented in that direction. Theygenerally slope downward in the directionof flow, but elevation data are not avail-able for most of this level; hence a dashedline is shown on the profile as an approxi-mation. There are at least five shafts that

descend from the Maze Level to still lowerpassages. Three of them (S-VI, S-VIII andS-IX) are aligned beyond S-V, the shaft thatconnects Level II to Level III. They descendabout 35 m to an underlying passage thatalso connects them together (see detailedprofile). This is the beginning of Level IV,the lowest known level in the cave. Atpresent, Level IV is composed of only twopassages. Main Drain is the largest knownpassage in the cave and is so named be-cause it is thought to re-collect much of thedistributary waters from the Maze Level.Initially, it trends horizontally 230 m south-east along the strike, before it turns south-westward like the principal passages inhigher levels. Bedrock floor is exposed forabout 100 m of the strike-oriented passage,but there are also large lakes having muchclastic sediments. From the turning point,the Main Drain gradually descends 22 mover a distance of 450 m to a dome-likeroom with a small siphon lake 139.5 mbelow the datum. This part of Main Drainis large, but varies considerably in cross-sectional shape, area, and height. The ef-fect of large NE-SW joints can be clearlyseen in both parts of the passage. Water,mud, coarser sediments, breakdown, veg-etal debris, and formations are much moreabundant in Level IV than in higher levels.About the middle of the strike passage,there is a large joint-controlled passagenamed Mystery Passage that is entered byclimbing up a sediment bank. It trendshorizontally northeastward for 330 m,reaching all the way back to the overlyingsurface channel. Near the middle, there isa high vadose dome (S-XI) that probablybrings water in from an undiscovered pitin the Maze Level. At its end, there are sev-eral unexplored pits, where water appar-ently escapes to another drainage route.

The morphology and genesis of Sótanode Japonés are complex and warrant fur-ther study, particularly after the cave ismore fully known. A preliminary interpre-tation is given here. Water was first piratedat Main Sink and was divided, part goingdown S-III to Flood Route (Level II) andthe remainder going down a large joint (inS-I) to the Maze Level (Level III). The verygreat amounts of water provided by floodsfrom the large surface basin produced ex-treme local hydraulic gradients, which incombination with the jointing and otherhydrogeologic characteristics of the bedsat Level III caused a distributary maze ofpassages to develop from the base of S-I.The larger channels may be dip tubes, butintermediate size joints were significant in

116


controlling passage orientation and en-largement. Several large joints providedtargets where water was taken down toLevel IV. The Main Drain is believed tohave re-integrated some of the distributedflood waters. Abundant active formationsand high domes or fissures, as well as theseveral known connecting shafts, substan-tiate this point; however, other lower leveldownstream passages might be found, asin Sótano de Yerbaniz (Mitchell et al.,1977). The basic skeleton of the cave wasestablished at an early stage. Level II flowmay originally have extended over to thetop of the large dome pit, S-IX, which isknown to reach much higher than Level III(see detailed profile), and subsequently waspirated down S-V. All of Flood Route(Level II) may have been temporarily aban-doned as S-I became large enough to cap-ture all the flow. Later, the development ofS-II and other upstream holes has capturedthe majority of flow for Flood Route.

In summary, it is believed that the pat-tern or skeleton of the cave was establishedvery early in the cave’s development. Allof the levels were formed in a singlephase—none are related to a water table.The two basic processes are developmentof a complex set of distributary channelsin the initially (presumably) unaltered bed-rock in an attempt to transmit the watersupplied, and regressive piracy down jointsbetween levels and in the surface arroyo.The first process can be especially impor-tant for floodwater caves. Palmer (1972)has described many characteristics of flood-water caves, including maze developmentaround passage restrictions or blockages(e.g., breakdown) where high local hydrau-lic gradients are generated, that make thema distinct class. An analogous situation oc-curred in the primary development ofJaponés, when far more water was providedby floods than could be passed. Water wasforced into a great many structurally con-trolled routes simultaneously, and conduitsformed. As the conduits enlarged to a smallfraction of their present size, the dry sea-son or between-flood water level was low-ered, probably below the terminal siphonin Main Drain. During floods, water wouldback up in the system inversely with itscapacity to quickly transmit the flood. Pres-ently, moderate floods may not cause wa-ter to fill the larger passages, except wherewater may have backed up in the lowestparts of the cave. Although it is difficult toassign a length to a complex cave such asJaponés, about 4 or 4.5 km are shown onthe map, and its depth to the water trap is

139.5 m. There is likely to be more air filledpassage on the other side of the water trap(siphon).

6.4.4. Sótano de Jos

Sótano de Jos is a very active swalletcave near the southern end of the LosSabinos group of caves. It has a catchmentof 4.3 km2, which is pirated down the joint-controlled entrance slot. The cave must bevery young relative to the large swalletcaves nearby, because its passages are com-paratively small, and because the arroyo hasentrenched only about 3 m below the topof the El Abra limestone, which is exposedfor only a few tens of meters upstream ofthe capture point. The cave lies in a valleyperhaps a few hundred meters wide and isa recent capture of most of the drainage thatformerly flowed to Sótano de las Piedrasand Sótano de Palma Seca.

Jos is a very simple cave (Figure 6.20)compared with the larger swallet caves. Theentrance is 27 m deep and displays potholesand vertical shafts common to vadose wa-ter flow. A dome that intersects the side ofthe entrance pitch descends another 34 mto a high vadose canyon containing largerounded boulders. After 40 m, a small tribu-tary passage joins. The main passage turnsnorthward, gradually descends, and gener-ally follows NNE joints away from El Choyspring for 275 m to a deep permanent si-phon 84.5 m below the entrance. In TiptoeLake, vein calcite occurs along a fracture.

Bedrock is exposed on the floor in manyplaces, and where it is covered, the sedi-ments are believed to be very thin. The cavecontains abundant flowstone deposits.These deposits and numerous inlet domesare undoubtedly caused by leakage downjoints and in the bed of the arroyo and downthe stratal dip. On September 11, 1971, twodays after a moderate storm, some waterwas observed flowing in the arroyo, but notdown the entrance drop (D. Broussard, per-sonal communication). In the cave, thewater entered via the domes and the inletpassages. A particularly large flowstonemass is located almost directly underneaththe arroyo (see map).

A stage recorder was placed over theshallow end of the siphon lake to monitorthe hydrologic activity of the cave (Section4.4.2). The location of the outlet from thislake is unknown, but it is believed that theceiling of the siphon will rise sufficientlyto form an air-filled passage again. Thecrests of large floods that pass through thecave must be damped at several places

where the passage cross section is reducedby flowstone or resistant bedrock. For ex-ample, Tiptoe Lake has a tubular cross sec-tion including a smoothly arched solutionceiling created by frequent flooding to theroof. The cave is not yet large enough toimmediately transmit all of the water pro-vided by major floods; thus water backs up,causing large rises of the siphon lake sur-face and possibly flooding the cave backto the canyon at the base of the entrancepitches. Fast flow is indicated by small scal-lops and by the lack of sediment in manyplaces, although much sediment is washedinto the cave. Sediment is transportedthrough the known cave; some of it may bedeposited within the aquifer, but probablymost of it is discharged at El Choy spring.

Some indication of the complicated his-tory of flowstone deposition and erosionin the El Abra region, even in this veryyoung cave, is given by one flowstone mass(marked “eroded flowstone” on the map)which formerly blocked the bottom 1.5 mof the passage. It has been sliced in half byflood waters. A round boulder one-halfmeter in diameter, previously includedwithin the flowstone, now bridges the topof the slot; the exposed flowstone crosssection reveals several distinct periods ofdeposition which were interrupted by peri-ods (or events) of erosion. It is not knownwhether the erosion phases were the resultof a few unusual rainfall-flood events orlonger climatic changes.

In summary, Jos is a young floodwatercave.

6.4.5. Sistema de los Sabinos

Sistema de los Sabinos is the name givento a complex of three large caves that arebelieved to be connected, but as yet haveshort distances remaining between them(Fish, 1974). The system is named after thevillage of Los Sabinos. The three caves areCueva de los Sabinos, Sótano del Arroyo,and Sótano (Cueva) de la Tinaja. A line mapof the system is shown in Figure 6.21.

Cueva de los Sabinos

Cueva de los Sabinos is a large caveabout 2 km southeast of the village of LosSabinos and about 700 m east-northeast ofthe sink of Sótano del Arroyo (Plate 3.9).The entrance is in the north wall of a largesink (Figure 6.22) that was at least partlycreated by collapse into the cave. The por-tal (Plate 6.22) is 13 m wide by about 8 mhigh and is prominently marked with

121

AMC

S Bulletin 14 — C

hapter 6

Figure 6.20. Map of Sótano de Jos.

122


phreatic domes and pockets. A ramp of col-lapse debris from the sinkhole descends 21m to a room dubbed the Temple of Karnakfor the magnificent bedrock pillars on theeast side. The pillars are remnants of solu-tion along a number of joints. Straightahead, First Level (passage) is split into anupper and lower route for about 70 m to alarge shaft that reconnects it. The lowerroute initially is a wide crawlway with adirt floor, followed by a room with forma-tions; the upper route is roughly equi-dimensional in cross section and containsabundant flowstone, guano, and severaldomes. Several small passages displayingphreatic forms and strong joint control meetat the top of the connecting pit. On thelower route, First Level continues horizon-tally beyond the pit as a large dirt-flooredpassage 24 m below the entrance. Itabruptly ends at a small side passage and asteep climb up a bedrock slope to the GuanoRoom. This room appears to be of phreaticorigin. The ceiling of the eastern end of theroom has the highest elevation in the cave,approximately 15 m above the ceiling ofthe entrance.

From the Temple of Karnak, the FirstLevel also continues in the opposite direc-tion as a smaller passage underneath theentrance sink to Flowstone Gallery. Alongthe way, it slowly descends, and there arethree domes on the left that contain abundant

formations and flowstone and, in at leastone case, breakdown from the entrance sinkthat is cemented by flowstone. In FlowstoneGallery, a slope and a pit descend about 25m to Second Level. A wide mud-flooredroom with formations and a vadose chan-nel leads down to a pool 67 m below theentrance, the lowest point in Second Level.

The passage that continues is quite largeand probably has several meters of flow-stone, guano, and clastic fill. It terminatesat a permanent sump only about 30 m fromthe sump at the end of Left Hand Passagein Sótano del Arroyo (see Figure 6.21).These two passages are very similar in size,elevation, and morphology; therefore they

Figure 6.21. Map of Sistema de los Sabinos.

Plate 6.22. Entrance to Cueva de los Sabinos in the north wall of a broadsinkhole (partially collapsed). The ceiling has phreatic domes. A slope overbreakdown leads down to the main cave level.

123


very likely connect. During the wet season,water emerges from the siphon lake andeither trickles down a nearby hole or athigher stage flows to the pool at the lowpoint in the passage. Vegetal matter accu-mulates on the mud banks around this pool.

Beyond a low constriction over theabove mentioned pool is a wet, muddy pas-sage named Elliott’s Swim that trendsnortheast underneath the entrance sink andthen turns back southward. Eventually dryland is reached 1.5 m above the normal lakelevel. Beyond a low inlet, the passage con-tinues horizontally for 150 m as a 4 m high,8 m wide tunnel with cobbles and mud onthe floor. It ends at the Mud Room, wheremud banks may have buried a continuationof the passage. Just before the Mud Roomis the entrance to a smaller passage thatdescends 30 m in its 200 m length. It endsin another siphon (hence DisappointmentPassage) 95.5 m below the entrance, thedeepest point in the cave. The total sur-veyed length is 1502 m.

Sótano del Arroyo

The entrance to Sotano del Arroyo lies1.7 km southeast of the Los Sabinos vil-lage. It is a swallet type sink that has a drain-age basin of 12.4 km2 (Figure 3.2 and Plate3.9). The surface stream was first pirateddown large north-south joints about 70 mabove the bottom of the sink (Plates 6.23and 6.24). Since then, the arroyo has en-trenched rapidly, exposing the El Abralimestone in a gorge for more than 1 kmupstream and forming several waterfalls inthe stream bed. Presently, the large entrancepit is degraded, so that drops of 8 m and 15m in the arroyo bed reach the bottom of thesink, whereas all other walls are 60 m ormore high (see the profile of the entrancesink on Figure 6.23). The datum for thecave survey is taken as the top of the 15 mentrance drop.

Two passages lead from the entrance. Asmall passage, the Thirty Foot Level, trendssouthwest down the 5° dip from a point 10m above the sink floor and directly underthe arroyo. It is nearly blocked by flow-stone, but this can be passed to reach a junc-tion 260 m from the entrance. The ThirtyFoot Level becomes larger and dirt-flooredfor the next 160 m, and then continues as ahigh narrow fissure for at least 150 m be-yond the end of the survey. The southbranch from the junction is also relativelylarge and passes obliquely across the dipto join the Right Hand Water Passage,which is a 4 m by 5 m passage with deep

lakes, interrupted by a 4 m high flowstonedam. This passage has a very irregular floorprofile, with the lowest points being on ei-ther side of the travertine dams in themiddle, 26.5 m below the datum. The pas-sage appears to be a down-dip loop.

The other passage, leading southwardfrom the bottom of the entrance sink, is thelarge Main Passage (Plate 6.25). This is thetrunk channel for the wet season floods thatenter the cave. It contains abundant depos-its of washed-in mud, sand, fine gravel, andvegetal debris (Plate 6.26). There are sev-eral small associated passages, some ofwhich appear to be active distributaries, andothers (higher) which may be fossil dis-tributaries or may be small vadose inlets.The Main Passage is nearly horizontal ex-cept for several piles of breakdown, untilat about 700 m, two 10 m drops are en-countered in a 30 m high canyon. Strongfracture control of the passages is clearlyevident in a number of places in this partof the cave (see map). The floor profile isvery irregular, and the stream has cut down-ward along fractures forming waterfalls, butbeyond each drop the water must rise sev-

across the room, a passage descends to adeep lake 60 m below the datum, having aslittle as 0.3 m of air space. This is the mainflood route, and the majority of the cavelies beyond this point. The passage risesup the stratal dip to Lake Drop, which is a16 m descent on a flowstone chute to themiddle of a deep lake (Reddell’s Lake). Thismarks the beginning of the lower level ofthe cave.

The Wallows is the main trunk channelof the lower level. Typically, it is 2 m highby 15 m wide and filled with lakes andsticky mud. It begins at 72 m below thedatum, but, surprisingly, beyond Neal’sPassage it rises very gradually over its1.1 km length to a terminal sump at about–70 m. The roof is solutional and gener-ally flat except for a few domes or crevicesdissolved along joints. It is believed thatthe right-hand passages off The Wallowsare inlets (i.e., sources), whereas all the left-hand passages and probably the sump ofThe Wallows are downstream passages.During the dry season, water trickles outof the right-hand passages and drains intoNeal’s Passage. Large floods, however, rise

Plate 6.23. The entrance pit of Sótano delArroyo. The walls are 60 m high except wherethe arroyo has cut a canyon down to the view-point 15 m above the floor. The joints extendfrom the top of the El Abra limestone to thefloor. Main Passage leads from the bottom ofthe sink.

eral meters over breakdown andmud. Immediately to the left fromthe base of the second drop is theLeft Hand Passage. A steep climbover breakdown and vegetal debrisascends in a 25 m high canyon al-most to the level of the bottom ofthe entrance sink. The highest partof the floor of this passage is inthe middle, where large stalagmiteand flowstone masses have beendeposited. The last 200 m de-scends slightly in a wide, mud-floored passage to a large siphonlake room. Probing the siphon hasshown that it continues at least ashort distance underwater. It likelyconnects with the Second Levelsump in Cueva de Los Sabinos.The trickles of water that enter LosSabinos via its sump lake probablycome from the small inlet at theend of the Left Hand Passage.Modern flood water from the MainPassage probably rarely uses thispassage.

Straight ahead from the seconddrop, the Main Passage continuesto the so-called Big TriangularRoom, which is over 100 m longand 30 m high and contains room-size breakdown swept clean ofmud by flood waters. Again frac-ture control is evident. Diagonally

124


Plate 6.24 (upper left). View outMain Passage of Sótano del Arroyointo the entrance sink. Wet-seasonfloods enter via the arroyo 15 mabove the floor. (There is a personstanding at the edge of the drop.)Particularly important is the largejoint that initially captured the stream.The leftmost hole on the wall leads tothe Thirty Foot Level.

Plate 6.25 (left). Main Passage inSótano del Arroyo, looking down-stream from 20 m inside the entrance.There are collapse boulders andformations on the left and cobblesand coarse sand on the right.

Plate 6.26 (above). A log in MainPassage in Arroyo. Violent floodscarry in huge quantities of mud andcoarser sediment (see the floor) andvegetal matter including logs.

125

AMC

S Bulletin 14 — C

hapter 6

Figure 6.25. Profile of part of Sótano de la Tinaja.

126


up the entire length of The Wallows, enter-ing all passages.

Neal’s Passage, the first-encountereddownstream passage in the lower level, ishorizontal for 300 m before it descends ashort fissure. It becomes blocked by flow-stone at least 50 m above the local watertable. Water escapes through crevices andthrough a small hole in the flowstone.Strickland’s Bad Air Passage is a very ac-tive flood-water route. It has an irregularbedrock profile, rising slightly up the dipbefore descending several short drops andtight vadose trenches. Beyond a flowstonesqueeze near the present limit of explora-tion, the CO2 content of the air reachesdangerous concentrations (see CO2 mea-surements at several points in the cave).A fissure drops 30 m to a large passageat –134 m (–441 feet), the lowest pointknown in the cave (unfortunately, one ques-tionable survey leg in the Bad Air Passagecould change the depth by ±10 m). Here,the CO2 reaches 3.8%, which has inhibitedfurther exploration. This passage must benear the local water table; hence it may soonsiphon. The CO2 is likely formed by thedecay of organic matter swept into the cave,and is maintained at high concentrationsbecause air circulation is restricted. A smallpassage, Flood Overflow, is used only bylarge floods. Water must rise high enoughto flow up some short pitches and over twobedrock drainage divides (shown as ↔ onthe map) before it also connects with thetop of the fissure. Last Left Lead Passageinitially is divided into two passages,smaller than the previous downstream ones,that join and continue as one. This passageis probably not used as a discharge routeas often as the others because of its profile.It rises 6 m up dip, before it descends al-most back to its initial elevation via shortdrops and finally ascends to 10 m above itsstarting level, where it becomes very lowor may be completely blocked. At the lowpoint of the passage, a 20 m unexplored pitdrops into a large canyon that must be theroute for flood water that enters this pas-sage. It may connect with the bottom (oreven the top) of the fissure in Strickland’sBad Air Passage.

There are about 2 km of known passagesoff the right side of The Wallows. The firstis the 190 m long Methane Passage, namedfor the flammable gas bubbles that rise fromthe lake bottom when it is disturbed. Thispassage is oriented in the same directionas the dip of the strata; therefore it quicklysiphons in a deep lake. A trickle of waterhas been observed to flow from this passage

even late in the dry season. The SecondRight Hand Lead is a very large tributary(?) passage developed along the strike ofthe beds. Like most of the other passagesin the lower level, it is very wet, muddy,and contains some breakdown. A smalltrickle often flows from it in the dry season.After 300 m, it branches. Tracy’s WaterPassage, the right fork, extends 400 m to adeep lake room, where a ceiling slot leadsto a high inlet dome. Access to the left forkis under a low ledge in an alcove. A pas-sage of modest size ascends slightly up-dipover cobbles and sand to connect withSteve’s Surprise Passage, a big canyon 300m long, 8 to 10 m wide, and 10+ m high.This canyon has some massive stalagmitesand a thick mud fill that is dissected by anintermittent stream. There are several largedomes (inlets?), and the passage ends at ahigh dome with a possible continuation atthe top. A large side passage on the eastquickly ends in a siphon. Along the westside of the canyon near the middle, a smallhole leads into the Amazing complex ofpassages. Only the larger ones have beensurveyed and shown on the map. The pas-sages seem to have a totally erratic profileand destination. A high fissure which hasundergone considerable collapse along afracture reaches at least 30 m above TheWallows. It seems likely that either floodbackwater enters the area and escapes tolower levels or vadose dome inlets supplywater that continues on to Tracy’s WaterPassage and Steve’s Surprise Passage or tolower levels.

Sótano del Arroyo is a very complex andactive cave. Its surveyed length is 7200 m,which makes it the longest known cave inMexico [1977].

Sótano (Cueva) de la Tinaja

Tinaja is the southernmost cave of thehypothetical Los Sabinos System. The en-trance sink is about 50 m deep, but en-trenchment of the arroyo and by rockfallhave degraded the sink, so that the cave atthe bottom is easily accessible. The rock ofthe entrance sink is (was) heavily fractured,and, as at Sótano del Arroyo, the large frac-ture that originally captured the streampasses from the surface all the way downto the cave (Plate 6.27). A map of part ofTinaja drawn by David McKenzie (pub-lished by Russell and Raines, 1967) wasconsiderably improved during the fieldwork, several passages were added (Figure6.24), and a profile of part of the cave wasprepared (Figure 6.25). The boulder strewn

floor of the sink continues as a ramp downinto the first part of the Entrance Passage.This passage trends eastward for 550 magainst the 5° WNW stratal dip. It usuallyvaries from 25 to 50 m2 in cross-sectionalarea, and three rooms are encountered. Thefloor is covered by rounded collapse boul-ders, except where mud, vegetal debris, andguano have collected in ponded areasagainst breakdown piles. Traverse Lake at–35.5 m is the lowest point; from it, floodwaters must rise 14 m to reach the 8 mCable Ladder Drop, the end of the EntrancePassage (Plate 6.28). Entrance Passage isa floodwater passage.

Directly east of Cable Ladder Drop, alarge canyon becomes blocked by stalag-mites after only 95 m. A side passage onthe south wall of the canyon leads to Si-phon Pit, a large 19 m deep pit. A climb upflowstone banks on the south wall of thepit leads to the Siphon Pit Series, a groupof moderate to small passages and roomsat about the same elevation as the top ofthe Cable Ladder Drop. These passages arewell decorated and are not an importantflood route. They probably developed astributaries to the main cave. Exploration ofSouth Branch has stopped at a 20 m pitcontaining bad air; there is a high levelpassage on the other side of the pit.

North of Cable Ladder Drop is the SandyFloored Passage, one of the largest andmost interesting in the Sierra de El Abra.The first 100 m is covered with roundedboulders, similar to Entrance Passage. Atthis point, a smaller passage trends to theeast, meeting Linkage Passage in a muddyroom. At the east end of the room, at a depthof 31 m (4.5 m above Traverse Lake), theDownstream Canyon begins. It descends 51m via short drops and mud-coated slopesto a large lake that siphons at –82 m, thedeepest known point in Tinaja. This pas-sage is the major course for flood water inthe cave, but the siphon must be above thelocal water table.

North of the downstream junction, theSandy Floored Passage continues as a largehorizontal gallery (Plate 6.29). Solutionalong cross joints has formed many domesand rooms. Large columns, flowstonebanks, and a profusion of stalactites makethis one of the best decorated passages inthe El Abra. The terrigenous sediments arepredominantly sand and gravel up to cobblesize, with some mud deposited in areas ofslower water velocity such as alcoves. Sanddoes not appear in the cave until just be-yond the downstream junction. The 200 mof passage beyond the Colonnade Room

129


Plate 6.27 (above left). Entrance toSótano de la Tinaja. The entrance sinkhas been sufficiently degraded thatthe cave is easily accessible. Largejoints captured the stream about 50 mabove and have promoted rockfall.

Plate 6.28 (above right). EntrancePassage in Tinaja just above the CableLadder Drop. The photo shows com-mon solution features in the massivebedded limestone of the swallet caves.Here water initially rose up-dip, beforedescending a joint at the drop.

Plate 6.29 (right). Part of the 1.5 kmlong Sandy Floored Passage inTinaja. There is a thick accumulationof sediment ranging from cobbles tosilt or clay size. The sand and mudare derived from the Méndez or SanFelipe formations. Here a recent coatof mud covers the floor.

130


has the best display (best sorting) of thesandy floor. At 575 m beyond the down-stream passage, the Sandy Floored Passagemakes a sharp left turn and bears generallynorthwest. After another 75 m, a bedrockvadose trench is encountered where thepassage shifts up-section about 5 m. Highup on the east wall at this point is Flow-stone Passage. Although a large passage, itis smaller in comparison, and is often nearlyblocked by flowstone. It bears northeast for370 m and ends in a wide mud-flooredroom having several massive columns.

Beyond the junction with FlowstonePassage, the Sandy Floored Passage con-tinues through some of its most spaciousgalleries to the Lake Passage. The passageis renamed here because it changes char-acter. Morphologically, Lake Passage isgenerally similar to the Sandy Floored Pas-sage, but it contrasts in having a series ofsix lakes and more mud. Second Lake isnearly a siphon, and Third Lake is a per-manent sump that may be bypassed by acrawlway over it. Just beyond Second Lakeis a large side passage that appears to be anactive source of water, but it is blocked withmud. In the flowstone mass at the far endof Third Lake, a large re-solution hole witha 0.5 m diameter tree stump in it leads to awide bedding-plane water crawlway thatmay be either a tributary or distributary ofthe main passage. Another 250 m of largehorizontal passage with lakes and sandyfloor reaches a permanent sump. The pas-sage continues quite large, but its roof isjust below the water surface. The siphonof this passage is perhaps only 30 m fromSteve’s Surprise Passage in Sotano del Ar-royo.

At some time in the past, the sedimentfill in the Sandy Floored Passage reachedhigher than it now does. There are somecobble banks (Plate 6.30), and there areseveral places beyond Triple Rooms wherenarrow remnants of the fill rise up to 2+ mabove the floor. Also, there are good ex-amples of stalagmites left hanging by theremoval of the fill. The old fill is composedof sediment sizes ranging from clay tocobbles. The highest occurrence of the fillis between stalactites in Mud Room, 5.5 mabove the present floor (see Figure 6.25).If the passage was uniformly filled to thislevel, First Lake (and possibly other lowpoints) would have been blocked.

Presently, the Sandy Floored Passage–Lake Passage is an inlet. The water is de-rived from a large number of sources, rang-ing from stalactite drips and flows to spout-ing cracks to tributary passages. These

sources were observed to provide moder-ate flows soon after summer storms. (Noobservations were made during stormswhen the flux must have been muchgreater.) They decreased to seeps or dripsover a few days, and most cease flowingduring the dry season. The flow directionis toward the Downstream Canyon, asshown by sand ripples up to 8 cm high, 1.5to 2 cm scallops, cobble imbrications, andother flow markings in the channel. Largeflows occurred at least twice in the period1971–1972, wiping out the footprints ofprevious visits, and, judging from the freshmud veneers, the water reached a depth of1 m in the Sandy Floored Passage. Thus,the two sites marked “flood sump” on theprofile, First Lake, and Second Lake mustform temporary siphons during modestfloods. Standing water was observed in theflood sumps after moderate summer storms.

The modern sediment distribution maybe explained in terms of the sedimentsource, the flow direction, and the passageprofile. Small floods are gradually erodingthe old fill, causing many of the giant sta-lagmites that were resting on it to fall over.Material up to pebble size is being re-moved, leaving cobbles that armor the floorin many places. The highest point on thefloor profile occurs just beyond TripleRooms, and from there to DownstreamCanyon the floor descends gradually. There,the mud is winnowed out by the fast flows,except in alcoves and behind obstructions,

leaving sand, which is transported in thebedload, often as ripples. Beyond TripleRooms, the percentage of mud increasesbecause of ponding behind the high pointof the passage. New sediment introducedvia surface sinks into the Sandy FlooredPassage is probably nearly all mud (andoccasionally pieces of trees). The best in-formation concerning the sequence ofevents occurs at the site marked “dated stal”on the map. At that locality, broken andabraded stalagmites are included in a sandygravel fill, on which rests a 2 m in diam-eter column. The fill probably does notextend downward more than a few meters,because the bedrock floor is exposed atseveral places along the passage.

Near the cave entrance, a small hole inthe north wall drops down through break-down to the Water Passage. It is a low, wide,wet, and muddy passage that takes floodwater through the breakdown. Curiously,this passage drains northward, in the op-posite direction to the Sandy Floored Pas-sage. After 300 m, there is a junction withan inlet passage that continues westwardabove a dome; it is not fully explored. Itprobably comes from an upstream piracyin the bed of the arroyo. It evidently pro-vides a permanent trickle of water that flowsdown the north branch of the junction. TheWater Passage continues horizontallynorthward, true to character, to a multilevelmaze. Beyond a 9 m drop, a passage con-tinues (unsurveyed) through a tight

Plate 6.30. Old fill in the Sandy Floored Passge. It contains cobbles in amatrix of clay (?) and sand. The old fill is being eroded and redistributed orremoved.

131


crawlway (Acupuncture Crawl) to morewater passage and another drop. The DeepBlue Hole Siphon may be another waterinlet. This passage may connect withSótano del Arroyo, providing the perma-nent (?) trickle found in The Wallows, butthere will probably be a sump in between.

The horizontal surveyed length of Tinajais 4502 m and its depth is 82 m.

The system

Individually, the caves described aboveare among the larger caves known inMexico. Sotano de Arroyo is Mexico’slongest [in 1977], at 7200 m. The three arebelieved to be part of a much greater sys-tem named Sistema de Los Sabinos (Fish,1974). The aggregate length of known pas-sages is over 13,200 m, but there unques-tionably is much more passage to be found(see the inset in Figure 6.24 which showsthe caves in relation to the range and theChoy spring; also, see the geomorphic map,Figure 3.2). As yet no connections havebeen made, either by explorers or by dyetracing, although the intervening distancesare probably no more than several tens ofmeters. The passage sizes and relationshipsare shown in a line plot (Figure 6.21),which was made from the cave surveys andair photo measurements of the distancesbetween their entrances. It is believed thatthe Left Hand Passage in Arroyo connectsvia a siphon with the Second Level siphonin Cueva de Los Sabinos, and that the Wa-ter Passage in Tinaja connects (probably viaa siphon) with Methane Passage or SecondRight Hand Lead in Arroyo; however, al-though a connection between the end of theSandy Floored Passage (or Lake Passage)in Tinaja and some part of Arroyo seemslikely, any genetic relationship betweenthese parts of the caves is unclear. It is alsoentirely possible that Disappointment Pas-sage in Los Sabinos connects with one ofthe downstream passages of the Wallowsin Arroyo. It is predicted that large quanti-ties of water infiltrate the limestone out-crops (some of it in explorable caves) eastof the presently known system and on thewestern flank of the range and integratewith lower levels of the system. One par-ticularly likely prospect for connection (aswell as serving as a target during cave de-velopment) is a large sinkhole near the endof Strickland’s Bad Air Passage in Arroyo;a dry valley leads to the sinkhole.

The origin and subsequent history of thesystem is most complex. More field dataare needed before a detailed model can be

proposed, but some generalities will beconsidered here. Not all parts of the cavesare the same age. The First Level in Cuevade los Sabinos has different morphologicfeatures than those found in Arroyo andTinaja, and it is thought to be of phreaticorigin, and thus older than the others. Fur-thermore, there is no deeply entrenchedarroyo leading to the cave, but it appearsthat the main channel or a tributary of theRío Sabinos passed over or near the en-trance some time in the past. Perhaps wa-ter leaked down fractures at the entrance,which is located on some sort of anticlinalflexure (see strikes and dips on Figure6.22). A group of small joint-controlledpassages may also have been inlets, but theymay have been distributaries. The lowerparts of the Los Sabinos cave contrast withthe character of First Level, are more likeArroyo, and, in fact, are probably relatedto it. Sótano del Arroyo was created prin-cipally by its swallet stream. All of the pas-sages on the left side of Main Passage andthe Wallows (viewed into the cave) aredownstream drains; the uppermost, the LeftHand Passage, is now largely abandonedin favor of passages of the Wallows. TheThirty Foot Level is an old down-dip loopwhich was left behind as the entrance sinkdeepened and Main Passage became domi-nant. In Tinaja, at least the Entrance Pas-sage, the Downstream Canyon, and (a largepart of) the Water Passage have been cre-ated by flood waters sinking at the entrance.However, water presently enters the SandyFloored Passage at many points and flowssouthward to the Downstream Canyon, andfaint imbrication of cobbles in the old fillsuggests that this has been the flow direc-tion for a long time. The Sandy FlooredPassage lies underneath the major channelof the ancient Río Sabinos (see Figure 3.2),which cut deeply below the top of the ElAbra limestone, and there are additional dryvalleys (tributaries) and up-dip El Abraexposures to the east. It is suggested thatthis passage was generated by many inletsfrom the overlying valley, forming an inte-grating system that flowed southward. Thenorthern end of the Sandy Floored Passage(actually the name is changed to Lake Pas-sage there) and the complex of passages atthe end of Second Right Hand Lead in Ar-royo are particularly favorably situated toreceive infiltration, because they lie belowthe bed of an arroyo having 2 to 3 km2

catchment at or near the point where itpasses from the impermeable cover onto theEl Abra limestone. No sinkholes could bedetected on the air photos; it appears that

the valley floor contains alluvium. Perhapsa proto-major piracy is developing, but itseems possible that the Sandy Floored Pas-sage is older than the Entrance Passage ofTinaja.

6.4.6. Other caves

There are several other swallet cavesalong the western margin besides the onesdescribed above. They will be mentionedonly briefly here; locations for most areshown on Figure 3.1. Cueva Chica, famousin the blind-fish literature, is a short cavehaving a relatively small catchment andprobably drains to the Río Tampaón. Sótanode Palma Seca and Sótano de las Piedrasare two small, simple swallet caves that liedownstream on the drainage that has beencaptured by Jos (Mitchell et al., 1977).Between Tinaja and Jos lies the Sistema deMontecillos, a 3070 m long, 114 m deepsystem having two entrances (map and per-sonal communication from Neal Morris).A deeply entrenched arroyo leads to a ma-jor swallet of an 8.4 km2 basin; the cavebelow is well developed, but quickly si-phons. An upstream piracy in the bed ofthe arroyo has generated a newer cave ofsmaller dimensions which has one passagethat connects with the older cave and an-other that leads to a large isolated passage,probably created by water pirated from aneastern tributary to the main arroyo. Thisisolated passage apparently contains a per-manent dry season stream (trickle). Part ofthe cave shows very strong joint control.

Approximately 3300 m north of Sótanodel Arroyo lies another large swallet cavecalled Sótano del Tigre. It has a catchmentof about 4.4 km2 and is the highest of theLos Sabinos area swallets. The cave wassurveyed, but the map is not given herebecause 600 m of the passage needs to beresketched. The arroyo has entrenchedabout 20 m into the El Abra limestone, andthe stream is captured down a joint-controlledshaft about 80 m deep. The passages in thevicinity of the entrance form a complexmultilevel maze, formed in the early stageof the cave’s development, when floodwaters forced open many bedding plane andjoint distributaries; some of the routes re-join at a lower level. One route predomi-nated, and after passing horizontally 210 mand descending two drops, it intersects amajor N-S passage. The northern part, theRicinuleid Passage, appears to have a thicksediment fill. The southern part is the maindrain of the cave. It is a large conduit, 20 mwide and 10 m high in places, that gradually

132


descends to a room 120 m long, 45 m wide,and 15 to 20 m high. A siphon occurs in anadjacent, smaller room. This conduit passessouthward about 1400 m towards the LosSabinos system. Altogether, more than3000 m of passages have been surveyed,and the total depth is 160 m.

Near Japonés, two other large swalletcaves have developed. Walsh (1972) pub-lished a map of Sótano de Matapalma whichshows that it consists mainly of one largeconduit that generally traverses southeastfrom the entrance, and it has a length of1722 m and a depth of 86.3 m. Sótano deYerbaniz (Plate 6.31) has recently capturedmost of the Matapalma drainage. It has acatchment of approximately 21.8 km2, thegreatest of the El Abra swallets. No map ofthe cave has yet been published, but a de-scription by Mitchell et al. (1977) indicatesthat it is a complex multi-level maze likeJaponés. Russell (1972) reports a length of1615 m and a depth of 95 m for the cave.

Sótano de Venadito, the only majorswallet outside the southern El Abra area,is a large, incompletely explored system fedby a surface catchment of 7.7 km2. Onelarge western-flank cave, Cueva de Florida,

is located in the northern El Abra range; itwas not observed in this study.

6.5. Caves in Other Partsof the Region

6.5.1. The Coy caves and spring

The Nacimiento del Río Coy and asso-ciated caves are located at the north end ofthe Coy dome, 30 km south of Cd. Valles.The dome is an outcrop of backreef faciesEl Abra limestone in a broad valley ofMéndez shale. It is about 3 km in diam-eter, and at the sharply folded northern endit rises 140 m above the surrounding coun-try to an elevation of 170+ m above sealevel (see topographic map on Figure 6.26).

The Coy spring is one of the largersprings in the world, having a minimuminstantaneous discharge of 13.0 m3/sec andan average of about 24 m3/sec in the pe-riod 1954–1971. It emerges horizontallythrough collapse boulders and cracks alonga 50 m front. At base flow, the water sur-face is normally around 31.4 m a.m.s.l.(S.R.H. data), the lowest major spring sitein the region. During floods, the stage of-

stage data shows good hydraulic connec-tion, with a head loss of about 0.6 m at baseflow and up to 1.2 m for large floods. Thecave contains some stalagmite and flow-stone deposits and a minor amount of guanoand earth.

Fifty meters east of the entrance to LaCueva and a little higher on the hillside isthe steeply sloping entrance to anothercave, Sótano de El Coy. The pit is an in-clined tunnel that descends 26 m to a largeroom, 40 m long, 20 m wide, and 15 m high.The room has predominantly phreaticforms, but several large pieces of rock havefallen in the north part. A climb up througha small hole leads to what was an alcove ofthe room, now littered with talus from acollapse that formed another entrance.Along the north wall of the room, there is alake up to 8 m deep. There appeared to beonly slight movement of the water whenthe cave was visited. The cave and over-land survey put the lake 3.6 m above thespring, which is probably too much.

6.5.2. Other caves in the region

Caves have been observed and exploredby the author and by others in many partsof the region. However, because access isusually difficult, the nature of cavern de-velopment in most of the region is poorlyknown. Many areas have not been recon-noitered. Presently, the local relief is veryhigh, up to 2000+ m, and very deep va-dose cave systems can be expected. Accessto the systems will often be hindered, be-cause so much of the infiltration occursthrough a myriad of small sinks rather thanat major swallets such as the Los Sabinosswallet caves. The Xilitla area has beensubjected to karst solution processes formillions of years, and caves and pits havebeen found throughout most of the eleva-tion range. Blind shafts are very common.There are some small stream sinks whichlead to deep vadose caves. Sótano deTlamaya is 454 m deep and over 1 km long,and Sótano de Huitzmolotitla is 245 m deepand about 3 km long. Some short, fossilphreatic caves have been found, and thereare also collapse rooms. The nearbyAquismón area has caves and some verylarge collapse pits and sinkholes. Sótanode las Golondrinas has approximately theshape of an upright elliptic cone 333 mdeep, with a base measuring 134 by 305 m.The Sierra de Guatemala is also intenselykarstified, and caves have likewise beenfound at all elevations across the range. Onthe western side, there is a small polje at

Plate 6.31. The slot-like entrance toSótano de Yerbaniz. This is a youngstream capture compared to Arroyo andTinaja. The arroyo has entrenchedonly a few meters below the top ofthe El Abra limestone.

ten rises several meters, and wa-ter flows from cracks above theroad as well as smaller ephemeralsprings on the western side of thedome.

Above the spring, two caveshave been found, and there un-doubtedly are more. Figure 6.26shows maps of the caves and theirrelationship to the spring. Directlyup the hillside from the spring, theentrance to La Cueva leads downover a boulder slope to a shortspacious cave, totally of phreaticorigin and development. All thewalls and ceiling are smoothlysculptured and pocketed. The firsthalf of the cave passes directlythrough the 55° dipping beds.Then the cave turns along thestrike and passes over a bedrockrise on the floor to a swiftly flow-ing stream. It is possible to swimand wade upstream 15 m to a si-phon. Although it has not beentested, this stream undoubtedlyconnects to the spring. However,its flow is much less than thespring, so there is at least one, andprobably several other sources(conduits). The S.R.H. has placedstage poles at the spring and in thecave stream. The correlation of the

133


1500 m where a lake often overflows in thewet season into a deep and complex cavecalled Sótano de la Joya de Salas.

It is suggested that significant cave de-velopment in the El Abra limestone will befound wherever the rainfall exceeds about600 to 800 mm in the region. Some cavesand solutional enlargement of fractures and

Figure 6.26. Maps of Las Cuevas and El Nacimiento del Río Coy.

bedding planes should be found in the drierareas. The importance of caves to the hy-drology of the region is indicated by theabove examples. Although more are known,they will not be described here, except tomention one other area. The area north andwest of Jalpan lies near the southwesternmargin of the Platform in a moderate rain-

fall zone (Figure 4.3). Caves and pits arevery well developed there, including onecollapse pit called Sótano del Rancho Barrothat is larger than the Golondrinas shaft.Some additional information on caves inthe region can be found in Bonet (1953b),Russell and Raines (1967), Raines, ed.(1968), and Mitchell et al. (1977).

134


7

ANALYSIS AND SYNTHESIS

Previous chapters have each considered anindividual subject, presenting data, calcu-lations, and some interpretations for eachtopic. This chapter summarizes that infor-mation, analyzes it further, synthesizes itinto models and interpretive discussions,and occasionally compares the results withothers reported in the literature. This taskis complicated by the fact that the subjectsare intimately interrelated; a full under-standing of the regional ground-water hy-drology cannot be attained withoutknowledge of the local hydrology, here rep-resented mainly by the Sierra de El Abrastudies, and vice versa. It was deemed bestto begin with the hydrochemical modelingand its hydrologic implications, followedsuccessively by analyses and discussionsof the regional ground-water system, theSierra de El Abra hydrology, and the caves.This sequence, therefore, proceeds fromregional to local considerations.

7.1. Hydrochemical Analysis

In this section the hydrochemistry of thesprings will be analyzed to develop ahydrochemical model of the karst flow sys-tems in the region.

7.1.1. Review of hydrochemical data andinterpretations

Excluding the thermal springs, thechemical analyses presented in Chapter 5reveal that there are two basic types of ef-fluent karst ground waters in the region.The first is a “normal” karst water, i.e., acalcium bicarbonate type, with a calciumcarbonate hardness of 120 to 220 mg/l (2.4to 4.4 meq/l of Ca). The Mg content is smalland variable, probably depending upon howmuch dolomite or magnesian calcite thewater has been in contact with and the resi-dence time in the aquifer. The second typehas much higher concentrations of Ca, Mg,and SO4 than the first and about the sameor slightly higher alkalinity. It is really acalcium-sulfate water, but with a lowerCa/Mg ratio than the first type.

These springs represent the integratedoutput of complex karst ground-water flowsystems. Although the precise source areasof the springs are not known, it is oftenassumed that the water comes from themountain range in which each spring oc-curs. However, in this region the possiblesource rocks for the dissolved constituentsindicate a more complicated hydrogeologicsystem. The source of the aqueous SO4 andhigh Ca concentrations is believed to bethe gypsum and anhydrite of the GuaxcamáFormation, which is thought by Pemex ge-ologists (Carrillo Bravo, 1971) to underliethe El Abra Formation over a large portionof the Platform, west of a Lower Cretaceousreef. This reef is thought to be approxi-mately coincident with the Sierra laColmena in the southern part of the Plat-form, but to shift somewhat eastward far-ther north (see Figures 2.2 and 2.4). Someof the SO4 might be derived from the LaBorreguita Formation, which occurs in thebroad Río Verde valley in the western por-tion of the Platform; this possiblity will bediscussed later. In Figure 7.1, the Mgtot andSO4tot contents of thirteen springs are plot-ted, using the best base flow analysis ofeach. It clearly distinguishes the two watertypes discussed above and shows that theMgtot and SO4tot concentrations are broadlycorrelated. Some of the Mg may be derivedfrom a solution of dolomite-bearinq lime-stone and leaching of magnesian calcite;however, a more likely source rock is thebasal dolomite of the El Abra Formation.

The distribution of minerals in the re-gion requires that a mixing model of twoor more sources be developed to explainthe chemistry of the calcium-sulfate typesprings. The correlation between Mg andSO4 suggests that a simple two-sourcemodel may be adequate. Such a model forthe Coy and Choy springs is shown in Fig-ure 7.2, where a local source supplies acalcium-bicarbonate type water derivedfrom circulation strictly within the El Abraor Tamasopo limestones and a deep re-gional source from recharge areas fartherwest brings a concentrated solution of Ca,

Mg, and SO4 plus carbonate species to thesprings. For the Choy, the local source isof course the Sierra de El Abra. For the Coy,the nearest possible local source is the areawest of Aquismón; thus both its local andregional flows are artesian in the valleysurrounding the Coy dome, and water fromboth sources must flow a long way to reachthe spring. The flood paths of the Coy andChoy springs from Figure 5.8 are repro-duced in Figure 7.1. The approximately lin-ear paths strongly support a two-sourcemodel, meaning the mixing of two grosschemical types of water. (Of course, in re-ality there are a multitude of physicalsources for the water that is discharged fromeach spring, but only two chemical types.)Whatever the composition (somewhat vari-able) of the regional source, the local sourceis a nearly perfect dilutant, because it hasvery little dissolved SO4 and Mg. For eachspring, the dry season discharge would bea mixture of differing amounts of the twosources of water. During flooding, the out-put concentrations would greatly decrease,as the flux from the local source dominatesthe mixing volumes (Figure 7.2). The Coyand Choy lines appear to represent the ex-treme values of the Mg/SO4 ratio of theregional water; the other springs sampledlie somewhere between them.

It was shown in Section 5.3 that there isalso a correlation between temperature andSO4 concentration for individual springsduring flooding. The interpretation is thatthe local sources for each spring will havea temperature dependent largely on the cli-matic conditions of its recharge area; someof the local water mixes with the warmerregional water to produce the family ofparallel flood mixing curves shown in Fig-ure 5.9. Thus, the Choy and the Mante arevery similar, while the Coy dilution curveis nearly parallel, but offset because its lo-cal source water comes from a higher,cooler area. The temperature of the regionalwater also undoubtedly differs (slightly?)from spring to spring, but it is likely con-stant for a given flow system.

135


gypsum-dolomite-water-carbon dioxidesystem will be examined.

The chemical model is shown in Figure7.3. Initially, water infiltrates the massiveEl Abra limestone, becoming highlycharged with CO2 from the soil and fissure(or cave) air. Ground-water type A is thegeneral chemical composition of the vadoseand phreatic waters circulating within thelimestone, where calcite (and a little dolo-mite) solution in a relatively open systemis the principal process. The residence timefor most of the type A water in the aquiferis probably short, and most of it is dis-charged back to the surface as type 1 springwater. This couplet is represented by thecave drip and cave lake samples taken inthe El Abra range (type A) and by the smallsprings (Canoas, Santa Clara, Arroyo Seco)and some of the large springs such as theHuichihuayán, the San Juanito, and theSabinas (type 1).

Some fraction of the total of the type Aground water circulates more deeply. Itpasses through the limestone to reach thebasal dolomite of the El Abra Formationand the gypsum below that. By then, thewater is well below the water table, thuseffectively closed to additional CO2, and

Figure 7.1. Plot of Mg against SO4 for best base-flow analysis of thirteen springs. A high Ca, Ma, SO4 water mixes invarying amounts with a Ca, HCO3 water.

Figure 7.2. Two-source mixing model for the Choy and Coy springs.

7.1.2. The geochemical model

In this section, a geochemical model ofthe hydrogeology will be developed. Suchmodels are helpful in understanding theflow system and placing limits on the con-centrations of dissolved species and may

be useful for quantitative hydrology. Wigley(1973) has given a theoretical discussionof the calcite-gypsum-water system, whichshows various evolutionary chemical pathsof water as it comes in contact alternatelywith each of the minerals. Here, some as-pects of the more complicated calcite-

136


will probably have reached equilibriumwith respect to calcite. Upon contact withthe new minerals, the water begins to evolvetoward type B ground water. As gypsum(or anhydrite) and dolomite dissolve, theCatot content greatly increases, causing pre-cipitation of calcite. Note that the dolomitesolution reaction (Figure 7.3) is written inthe incongruent form to show that, to a largeextent, what occurs is the exchange of Cafor Mg in the solution. The result would becalcitization of the dolomite beds, with theadded possibilities of the formation ofdedolomite and calcite after gypsum (oranhydrite). A similar scheme has been re-cently proposed by Abbott (1974) for theEdwards limestone (equivalent to the ElAbra Formation) of central Texas, wherethere is much more petrographic, geo-chemical, hydrologic, and geologicalinformation available. An instance and apetrographic description of a dedolomitehas been given by Lucia (1961).

Type B ground water eventually losescontact with the gypsum and the dolomite.It passes laterally and upwards throughlimestone beds, but, theoretically, thechemical composition should not changebecause the solution is saturated with re-spect to calcite and it has left the gypsum-dolomite region. Ultimately, it mixes withsome local Ca, HCO3 (type A) ground waterto form type C ground water. This water isthen discharged from the aquifer, provid-ing the second type of spring water chem-istry found in the region—the high Ca, Mg,and SO4 waters of springs such as the Coyand the Choy (type 2). The undersaturationwith respect to dolomite and gypsum in allof the type 2 springs and the nearlymonomineralic solid phase of calcite in themixing zone of type C water mean that SO4and Mg are conserved after the mixing oc-curs. As shown in section 5.3, the base flowof these springs is saturated to slightly su-persaturated with respect to calcite exceptfor periods of a few weeks (for the Coy) toa few months (for the Choy) after moder-ately large floods. This means that Ca andthe CO3 (species) may not be conserved—calcite could be precipitated or dissolvedafter mixing. If the volume of the mixingzone is approximately the same as the vol-ume that was flushed out by the local sourceflood water, then type C ground waterwould have a relatively short (up to a fewmonths) residence time. This suggests thatnot much additional calcite could be dis-solved by type C ground water. Hence, inthis chemical model, there are three kindsof ground water and two kinds of spring Figure 7.3. The geochemical model.

Table 7.1Equilibrium-constant functions

Equilibrium constant, temperature t dependent Source

log(Kcalcite) = –8.3389 –0.001236t – 0.00005t2 1

log(Kdolomite) = –16.5614 –0.012486t – 0.0001999t2 1

log(Kgypsum) = –4.6407 + 0.002488t – 0.0000592t2 1

log(KCaSO40) = –2.2000 – 0.002517t – 0.0000433t2 1

log(KMgSO40) = –2.03 – 0.0132t 2

log(KMgHCO3+) = –0.95 (at 25°C; assumed for all temperatures) 3

log(KH2CO3) = –6.577 + 0.01264t – 0.000144t2 1

log(KHCO3–) = –10.25 (35°C); –10.29 (30°C) 4

log(KCO2) = –1.1585 (35°C); –1.53 (30°C) 4

1. Wigley (1972)

2. Langmuir (1971)

3. Hostetler (1963)

4. Garrels and Christ (1965), interpolated from their values

Individual activity coefficients were calculated with the Debye-Huckel equation–log γi = (A zi

2 √I) / (1 + åi B √I), with constants taken from Garrels and Christ (1965).

137


water.The ionic concentrations and the degree

of calcite saturation varies spatially withinthe limestone aquifer (they are distributedparameters of the system), depending oneach hydrochemical environment (see sec-tion 5.2) and on the residence time. Sometheoretical evolutionary models for the so-lution of pure calcite in water with CO2have been given by Langmuir (1971). Thesehave been recalculated for 25°C and areshown in Figure 7.4. Ground water opento a reservoir of constant PCO2 and dissolv-ing calcite would pass diagonally upwardsalong a line parallel to the lines of constantPCO2 shown. If the water at some time losescontact with the CO2 reservoir, then theCO2 is gradually used up, and the chemis-try will move along a curved path such asthe one marked “no more CO2 added.” Awater that has become saturated, but that isexposed to varying PCO2 along its flowpaths, will adjust to each new condition bymoving along the calcite saturation line.Based on these models and the field data(Table 5.1), the deep type A ground waterwill probably be saturated with respect tocalcite, and have 3.5 to 5.25 meq/l of Ca++

and HCO3– at pPCO2 ranging from 2.0 to

1.5.Unfortunately, there are no samples of

type B ground water for analysis. Becauseit would be useful to know what its com-position might be, a theoretical investigationwas made. The pertinent chemical reactionsare the solution (or precipitation) of cal-cite, dolomite, and gypsum, solution of CO2and the various carbonate species’ disso-ciation equilibria, and the ion pair reactionsto form CaSO4, MgSO4, and MgHCO3. Theequilibrium constant functions used aregiven in Table 7.1. A computer program waswritten which calculated the concentrationsof the dissolved species and the pPCO2 fora specified temperature, pH, and degree ofsaturation with respect to calcite (SIc), do-lomite (SId), and gypsum (SIg). Selected re-sults are given in Table 7.2.

An equilibrium model (SIc = SId = SIg = 0)was tested over the pH range 6.0 to 8.5 at30° and 35°C (Table 7.2a). The followingobservations are made:

1. The CO3tot and the PCO2 decreasegreatly with an increase in pH. The extremevalues of pH give unreasonable values ofCO3tot and PCO2.

2. For the pH range of 6.5 to 8.5, thereis very little change in Catot, Mgtot, andSO4tot. For increasing pH, the Catot andMgtot decrease slightly, while the SO4tot

increases slightly. Thus within the range ofits possible values, the PCO2 is not an im-portant control of the concentrations of thethree principal constituents.

3. The CaSO4 and MgSO4 ion pairs areimportant components of the solution.About 45% of the Mgtot and 40% of theSO4tot are tied up in these ion pairs.

4. The equilibrium calcite-dolomite-gypsum-water system has one third moreSO4tot than the simple gypsum-water sys-tem at 35°C (Column 1), because the ionicstrength is larger and because the MgSO4ion pair is quantitatively important. Thesefactors overwhelm the suppressing effectof the common ion Ca++. Note that the equi-librium models do not contain more Catotthan the pure gypsum water.

5. The ionic strength decreases slightlywith both increasing pH and increasingtemperature. The greatest change is in Mgtotwith temperature.

6. In the evolution of type A water to a

saturated type B water, the aCa++ will in-

crease about 2.5 times, causing a recipro-cal change in the aCO3

– – . Then, from thedissolved carbonate species equilibria, theexpected change in the pH would be a de-crease of a few tenths (less than log 2.5)and the CO3tot would remain about thesame. Thus the favored composition wouldhave a pH in the range 6.5 to 6.9 for tem-peratures between 30 and 35°C; at 35°C, apH of 6.84 is neutral.

Some authors (e.g., Langmuir, 1971)believe that the theoretical solubilities ofcalcite and dolomite may effectively con-trol the chemistry of ground water in car-bonate rocks. Others (e.g., Back andHanshaw, 1970 and 1974) have found thatthe concentrations of dissolved speciescommonly increase with the length of the flowpath. Frequently the water becomes super-saturated with respect to calcite and dolo-mite, but often the specific concentration

Figure 7.4. Evolutionary calcite solution diagram, HCO3 against pH.

138


of sulfate and the saturation condition withrespect to sulfate minerals in particular isnot given. It appears that the waters aregenerally undersaturated with respect togypsum. Possible explanations of these dis-crepancies include an uncertain value ofKdolomite, precipitation kinetics and the de-gree of supersaturation necessary to over-come the nucleation potential (Plummer,1972, suggests that an SIc of about +0.3 isnecessary to precipitate calcite), and greatersolubility of impure calcite. Supersatura-tion with respect to calcite or dolomite mayespecially occur where gypsum is beingdissolved faster than calcite (or dolomite)can be precipitated to maintain equilibrium(Back and Hanshaw, 1970; Mercado andBillings, 1975). Considering the sequen-tial exposure of source minerals along theflow path in this model and the increase of

ionic strength in type B water, it is alsopossible that the regional flux may be un-dersaturated with respect to either or bothdolomite and gypsum. Therefore, using thecomputer program, the chemical composi-tion for a variety of saturation states hasbeen determined at pH 6.8 and 7.0. Selectedresults are listed in Table 7.2b.

The only checks available on the pos-sible composition of type B ground waterare the analytical data for the springs, aque-ous chemistry theory, and field results ofother studies. Figure 7.5 is a plot of Catotand Catot+Mgtot against SO4tot for selectedanalyses from this study; most of thesamples are from the dry season base flow,but a few intermediate to high flood flowsamples are included for comparison. Pre-sumably, the higher the concentrations ofthese constituents, the greater the percentage

of type B water in the mixed output. TheCatot+Mgtot line is approximately parallelto the Catot = SO4tot line (gypsum solution),but the Catot line has a lesser slope. Therelatively linear fit of the data shows that amixing model of two waters having greatlydifferent and constant composition is agood first approximation of the system. Thedivergence of the two lines supports thechemical model of type B water in which,as gypsum is dissolved, the common ionCa++ is suppressed by calcite precipitationand the exchange of Mg for Ca by incon-gruent solution of dolomite. Note that theCatot/SO4tot of the Media Luna and of theCoy (sample no. 65, Table 5.1; not plotted)have just crossed into the realm where theCatot/SO4tot is less than 1. Also, the Coyvalues of Catot+Mgtot fall slightly below theline indicated by the others.

Figure 7.5. Ca against SO4 and Ca + Mg against SO4 for springs and theoretical models of type Bground water.

139

AMC

S Bulletin 14 — C

hapter 7

Table 7.2aPossible theoretical compositions of type B ground water

Equilibrium modelsConcentrations in millimoles per liter

Column 1 2 3 4 5 6 7 8 9 10 11 12 13 14

T (ºC) 35 30 35

SIc –∞

SId –∞ SIc = SId = SIg = 0 SIc = SId = SIg = 0

SIg 0

pH —— 6.5 6.8 7.0 6.0 6.5 6.6 6.7 6.8 6.9 7.0 7.5 8.0 8.5

Catot 15.07 15.12 14.48 14.23 17.66 15.36 15.13 14.93 14.78 14.65 14.55 14.29 14.20 14.18

Mgtot —— 8.57 8.19 8.05 8.61 7.49 7.38 7.29 7.21 7.15 7.10 6.98 6.94 6.92

SO4tot 15.07 20.47 20.97 21.19 18.80 20.11 20.29 20.44 20.55 20.65 20.74 20.97 21.05 21.07

CO3tot —— 9.65 4.21 2.50 38.03 8.15 6.13 4.65 3.55 2.73 2.11 0.619 0.192 0.061

pPCO2 —— .946 1.522 1.91 0.046 0.974 1.17 1.36 1.55 1.75 1.94 2.93 3.93 4.93

Ca++ 9.88 10.16 9.52 9.27 12.47 10.17 9.94 9.75 9.59 9.46 9.37 9.10 9.02 8.99

Mg++ —— 5.02 4.71 4.59 5.07 4.16 4.07 3.99 3.93 3.88 3.84 3.74 3.70 3.69

SO4– – 9.88 12.10 12.61 12.82 10.38 11.69 11.87 12.02 12.13 12.23 12.32 12.55 12.63 12.65

HCO3– —— 6.17 3.25 2.10 14.32 5.28 4.28 3.45 2.77 2.23 1.78 0.577 0.184 0.058

CaSO40 5.19 4.96 4.96 4.96 5.19 5.19 5.19 5.19 5.19 5.19 5.19 5.19 5.19 5.19

MgSO40 —— 3.41 3.41 3.41 3.23 3.23 3.23 3.23 3.23 3.23 3.23 3.23 3.23 3.23

MgHCO3+ —— 0.138 0.069 0.043 0.315 0.098 0.078 0.062 0.049 0.039 0.031 0.010 0.003 0.001

CO3– – —— .0019 .0019 .0020 .0015 .0017 .0018 .0018 .0018 .0018 .0018 .0019 .0019 .0019

H2CO3 —— 3.34 0.886 0.361 23.40 2.76 1.78 1.14 0.729 0.465 0.296 0.030 0.003 0.003

% CO2 —— 11.32 3.00 1.22 89.98 10.63 6.84 4.38 2.80 1.79 1.14 0.117 0.012 .0012

Catot/Mgtot —— 1.764 1.768 1.768 2.051 2.051 2.050 2.048 2.050 2.049 2.049 2.047 2.046 2.049

Catot/SO4tot 1.00 0.739 0.691 0.672 0.939 0.764 0.746 0.730 0.719 0.709 0.702 0.681 0.675 0.673

Mgtot/SO4tot —— 0.419 0.391 0.380 0.458 0.372 0.364 0.357 0.351 0.346 0.342 0.333 0.330 0.328

140

AMC

S Bulletin 14 — C

hapter 7

Table 7.2bPossible theoretical compositions of type B ground water

Non-equilibrium modelsConcentrations in millimoles per liter

Column 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

T (ºC) all 35

SIc 0 0 0 0.18 0.30 0.18 0.30 0 0 0 0.10 0.12 0.12 0.12 0.30 0.30 0 0.18

SId 0 0 0 0.18 0.30 0.18 0.30 –0.12 –0.12 –0.15 0 0 0 0 0.18 0 0.18 –0.18

SIg 0 –0.18 –0.30 0 0 –0.30 –0.30 0 –0.30 0 0 0 –0.10 –0.30 0 0 0 –0.18

pH all 6.8

Catot 14.78 10.95 8.98 15.08 15.36 9.35 9.71 15.17 9.28 15.25 15.28 15.37 13.00 9.53 15.80 16.18 13.93 12.02

Mgtot 7.21 5.31 4.33 7.36 7.49 4.50 4.67 4.25 2.57 3.72 4.70 4.31 3.63 2.64 4.33 1.97 15.37 1.15

SO4tot 20.55 14.58 11.44 20.33 20.11 11.18 10.96 18.10 10.13 17.67 18.31 17.96 14.77 9.97 17.62 15.81 27.55 11.11

CO3tot 3.55 4.19 4.67 5.21 6.79 6.69 8.57 3.30 4.33 3.25 4.14 4.28 4.68 5.54 6.27 5.89 4.23 5.20

pPCO2 1.55 1.47 1.42 1.39 1.27 1.27 1.16 1.58 1.45 1.59 1.48 1.47 1.42 1.34 1.30 1.33 1.49 1.37

Ca++ 9.59 7.49 6.38 9.89 10.18 6.75 7.11 9.98 6.67 10.06 10.09 10.18 8.89 6.93 10.61 10.99 8.75 8.56

Mg++ 3.93 3.10 2.66 4.05 4.16 2.81 2.96 2.36 1.61 2.08 2.62 2.41 2.12 1.67 2.45 1.13 7.98 .706

SO4– – 12.13 8.96 7.22 11.91 11.69 6.96 6.73 11.06 6.59 10.86 11.08 10.92 9.17 6.43 10.61 9.81 15.09 7.23

HCO3– 2.77 3.26 3.63 4.06 5.30 5.21 6.66 2.58 3.38 2.55 3.24 3.35 3.66 4.32 4.92 4.63 3.27 4.08

CaSO40 5.19 3.46 2.60 5.19 5.19 2.60 2.60 5.19 2.60 5.19 5.19 5.19 4.12 2.60 5.19 5.19 5.19 3.46

MgSO40 3.23 2.15 1.62 3.23 3.23 1.62 1.62 1.86 .932 1.62 2.04 1.86 1.48 .932 1.82 .808 7.27 .426

MgHCO3+ .049 .048 .048 .074 .099 .072 .096 .028 .027 .024 .039 .037 .037 .036 .055 .024 .113 .014

CO3– – .0018 .0020 .0022 .0027 .0035 .0031 .0040 .0017 .0020 .0016 .0021 .0021 .0023 .0025 .0032 .0029 .0022 .0024

H2CO3 .729 .875 .986 1.066 1.388 1.410 1.800 .684 .923 .676 .857 .887 .980 1.179 1.298 1.230 .845 1.107

% CO2 2.80 3.36 3.79 4.10 5.34 5.42 6.92 2.63 3.55 2.60 3.30 3.411 3.77 4.53 4.99 4.73 3.25 4.26

Catot/Mgtot 2.050 2.062 2.074 2.049 2.051 2.078 2.079 3.569 3.611 4.099 3.251 3.566 3.581 3.610 3.649 8.213 .906 10.452

Catot/SO4tot .719 .715 .785 .742 .764 .836 .886 .838 .916 .863 .835 .856 .880 .956 .897 1.023 .506 1.082

Mgtot/SO4tot .351 .364 .372 .362 .372 .403 .426 .235 .254 .211 .257 .240 .246 .265 .246 .125 .558 .104

141


The Catot and Catot+Mgtot lines of Fig-ure 7.5 have been extrapolated so that thevarious hypothetical chemical compositionsin Table 7.2 may be tested for agreement,within the reasonably expected deviationsfrom the lines. First note that the extrapo-lated total hardness line shows that the typeB water would have an alkalinity of about4 meq/l if, as previously discussed, there isno great change in SIc of the type C water.The following general observations aremade for 35°C:

1. It is assumed that 0 ≤ SIc ≤ + 0.3, andthat SIg ≤ 0.

2. SIc – 0.15 ≤ SId ≤ – SIc defines ap-proximately the relationship of SId to SIcin the potentially acceptable compositions.Therefore –0.15 ≤ SId ≤ +0.3. SId < SIc –0.15 suppresses the Mgtot too much, andSId > SIc gives too high a Mgtot.

3. Statement 2 above holds for the en-tire tested range of –0.3 ≤ SIg ≤ 0. As SIc – SIdincreases, the Mgtot/SO4tot decreases.

4. Slight adjustments of the pH (andhence PCO2) will cause very small changesin CO3tot. This adjustment may be used tobring a solution of particular Ca/SO4,Mg/SO4, and Ca/Mg ratios into a better fit.

5. Solutions which have SIc = SId = vari-able and SIg = constant are almost identi-cal in Catot, Mgtot, and SO4tot concentra-tions. Allowing SIc and SId to vary (whileremaining equal) simply changes the CO3totand the PCO2.

6. The equilibrium model SIc = SId = SIg =0 for a specified pH has the highest SO4totof any model composition permitted by theassumptions above in 1. SIc = SId > 0 withSIg = 0 give nearly equivalent results. ForSIc and SId constant, the Mgtot varies di-rectly with the SO4tot (and SIg).

The analysis so far has placed an upperlimit on the concentrations of Catot, Mgtot,and SO4tot in type B water and indicatedthat there is a range of acceptable compo-sitions if a nonequilibrium model is allowed.Certainly it is possible that equilibrium withrespect to dolomite and, more particularly,gypsum may not be reached in the type Bwaters.The Mgtot/SO4tot, and to a lesser ex-tent the Catot/SO4tot and the Catot/Mgtot,serve to limit the possible compositions oftype B ground water and thus the range ofpossible saturation conditions as well. Thiswill be examined more closely in the nextsection.

7.1.3. Mixing-model calculations

The two-source model hypothesized inprevious sections may be analyzed quanti-tatively by the dissolved load equation

CAQA+CBQB = C2Q2 (7.1)

and QA+QB = Q2, (7.2)

whereQA is discharge of the local source,CA is concentration of some ion in the local

source,QB is discharge of the regional source,CB is concentration of the same ion in the

regional source,Q2 discharge of the spring, andC2 is concentration of the same ion in the

spring water.

Hem (1970), Hall (1970), and Pinderand Jones (1969) have used mixing modelequations in other works. The relationshipsassume that there is no change in the vol-ume of the mixing reservoirs (i.e., nochange in storage) and that the parametersor dissolved species used are conservedafter mixing. Based on the geology of theregion and the undersaturation of thesprings with respect to dolomite and gyp-sum, it is believed that both SO4tot and Mgtotqualify as conservative parameters. Equa-tion 7.1 is often modified by dividingthrough by Q2, so that QA and QB becomefractional discharges:

CAQA′+CBQB′ = C2, (7.3)

where QA′+QB′ = 1. (7.4)

During the dry season, the chemistry andthe discharge of the flow systems are ap-proximately stable; this is the optimumtime for model calculations. There are fourindependent variables, two dependent vari-ables, and two equations relating the vari-ables. C2 and Q2 can be measured, CA maybe estimated from nearby samples of typeA water, and a value of CB may be chosenfrom Table 7.2; this leaves only two un-knowns and two equations. Using the SO4totconcentration in the Choy spring in May1973 and selecting the equilibrium modelat pH 6.8 and 35°C (column 15), for C2 =4.40 mmol/l, CA = 0.1 mmol/l based on cavelakes and Santa Clara and Arroyo Secosprings, and CB = 20.55 mmol/l, we have0.1(1–QB)+20.55QB = 4.40, QB′ = 4.30 / 20.45= .2103 or 21%, and QA′ = 0.7898 or 79%.Thus, remembering that the equilibrium

model had the highest theoretically possibleSO4tot concentration, at least 21% of thedischarge of the Choy was supplied by theregional flow system on that date. This rep-resents a QB of 0.60 m3/sec.

Because the Mgtot and the SO4tot are cor-related and are conserved, the Mgtot of theequilibrium model type B solution may betested to see if it predicts a reasonablevalue of Mgtot in the type A water:0.7897CA+0.2103×7.211 = 1.82, and CA =0.385 mmol/l = 9.4 mg/l of Mgtot. The cal-culated CA is about twice the value thatwould be expected from the average of thecave lake and type 1 spring water samples;however, the local discharge to the Choymay have sufficient average residence timeto dissolve more dolomite or leach magne-sian calcites than the small local springs.Also, because the expected value is small,any deviation appears to be a large error,but the predicted value is actually quiteclose. A slightly greater Mgtot in the type Bwater would give a better fit.

Although the Catot is not necessarilyconserved, no large change is anticipated:0.7897CA+0.2103×14.775 = 5.08, andCA = 2.498 mmol/l = 100 mg/l of Catot. Thisresult is only 12 to14 mg/l (15%) larger thanthat of the Santa Clara and Arroyo Secosprings, which were measured early in thedry season, when they were slightly under-saturated with respect to calcite.

The calculated composition of the equi-librium model at 35°C (Table 7.2b, column15) fits very closely the requirements oftype B water for the two-source model ofthe Choy spring. Similar or even improvedresults are obtained for SIc = SId > 0 andSIg = 0 (columns 18 and 19) and for theequilibrium model at 30°C (column 3); thecalculated results are shown in Table 7.3a.Solutions with SIc = SId ≥ 0 and SIg = –0.3also meet the compositional requirementsof the model (see Table 7.3a). Hence, thetype B ground water that mixes with typeA El Abra water in the Choy spring mayhave an upper limit of the SO4 concentra-tion of 20.55 mmol/l (1974 mg/l, variesslightly with pH and temperature choices)and have proportions specified by Catot/Mgtot ≅ 2.07 and Mgtot/SO4tot ≅ 0.40. Thereis little change in the results if, say, 0.05 or0.2 mmol/l is used as the SO4tot of the Si-erra de El Abra type A water; in fact, 0.1mmol/l appears to be a reasonable figurefor at least the eastern part of the region.

Unfortunately none of the Coy analysesmay be taken as completely untaintedsamples of the stable system conditions atthe end of the dry season, because small

142


floods occurred before the 1971 and the1973 samples were taken (Figure 4.7). Nev-ertheless, three samples probably closelyapproach the desired condition. Calcula-tions similar to those above have been madefor the Coy spring using the May 17, 1973,sample and a large number of possible typeB compositions; selected results are givenin Table 7.3b. The Huichihuayán, SanJuanito, Sabinas, and Canoas spring watersamples and the minimum concentrationsobserved during flooding at the Coy spring(see Table 5.1) are taken to be representa-tive of type A ground water farther awayfrom the reef than the local source area forEl Choy. The equilibrium model does notmake close predictions of the Mgtot andCatot of this type A water. The best resultsat 35°C were obtained for SIc–SId of +0.10to +0.12; a slightly larger difference wouldbe necessary at 30°C. Thus, the regionalflux to the Coy spring has a maximumSO4tot content of about 18.5 mmol/l (1780mg/l) and has chemical proportions speci-fied by Mgtot/SO4tot ≅ 0.25 and Catot/Mgtot≅ 3.4. A few of the “best fit” type B solu-tions were also tested with the June 4, 1973,and June 8, 1971, Coy samples; the resultswere about the same. At least 40% of theCoy base flow discharge is derived fromregional sources.

The studies discussed above have shownthe character and possible range of ion

concentrations that the regional water musthave to satisfy the hypothesized two-sourcemodels of the Choy and Coy springs. Asshown in Figure 7.1, these two springsapprear to form the “boundaries” of thechemical character of the high sulfatesprings. All the other springs sampled havean intermediate Mgtot/SO4tot; hence theymust have regional sources whose chemis-tries fall within the rather narrow differencebetween the chemistry of the regionalsources of the Coy and the Choy. This mayalso be seen in Figure 7.5, where the Choyand Mante samples plot above the trendline, while the Coy samples consistently arebelow the line.

Now that the upper limits of the type Bwater have been established, a graphicaltechnique may be used to determine themixing proportions for any spring. Mixingequation 7.3 is used, with CA equal to 10mg/l of SO4tot. In Figure 7.6, QA and QBmay be determined in percent (%) from theSO4tot of the mixed output (type 2 spring)and the chosen curve. Curves numbered 1and 2 represent the dilution curves for theChoy and Coy respectively, using theirmaximum permitted type B SO4tot. Thusthese curves give the minimum possible fluxof the regional system. Because the linesare close together for SO4tot less than 1000mg/l, the mixing proportions of othersprings may reasonably be determined by

choosing either line or by estimating theircorrect position. For type B water super-saturated with respect to calcite and dolomiteaccording to the specifications previouslygiven, and at equilibrium with respect togypsum, the respective curves would belittle different. The exact SO4tot content ofthe type B ground water is not known, andso far no minimum SIg or SO4tot has beenestablished. It is assumed that the type Bwater flowing to each of these type 2springs will have reached about the sameSIg. On this basis, a lower limit may betaken as the 836 mg/l of the June 8, 1971,Coy sample. This would of course meanthat all of the Coy discharge on that datewas from the regional source. The lowerlimit of 836 mg/l was used to computecurve 3, which yields what is assumed tobe the maximum possible proportion oftype B water in the mixed output of thesprings.

An estimate of the regional base flowhas been determined using Figure 7.6 andthe discharge of the principal eastern sul-fate (type 2) springs. For these calculations,it is very important that the flow system bestable. In terms of the model as shown inFigures 7.3 and 7.2 and the mixing equa-tion 7.3, it is necessary for QA, CA, QB, andCB to be constant for a sufficiently long timeso that CC, the concentration of the ion inthe mixing reservoir, becomes constant.

Table 7.3Mixing-model calculations for the Choy and Coy springs

Composition of Calculated Mgtot Calculated Catottype B water % % for type A for type A

(Table 7.2 column #) SIc SId SIg type A type B mg/l mg/l

a: Choy spring, May 21, 1973, base-flow sample

15 0 0 0 79.0 21.0 9.4 100

18 +0.18 +0.18 0 78.7 21.3 8.2 95.2

19 +0.30 +0.30 0 78.5 21.5 6.5 90.8

3 0 0 0 79.4 20.6 4.05 106

17 0 0 –0.30 62.1 37.9 7.0 108

20 +0.18 +0.18 –0.30 61.2 38.8 2.9 95.1

b: Coy spring, May 17, 1973, base-flow sample

15 0 0 0 65.2 34.8 –21.9 130

22 0 –0.12 0 60.5 39.5 9.6 83.8

23 0 –0.12 –0.30 29.0 71.0 8.2 93.7

25 +0.10 0 0 60.9 39.1 3.4 85.0

26 +0.12 0 0 60.2 39.8 8.3 75.9

28 +0.12 0 –0.30 27.9 72.1 1.7 56.1

29 +0.30 +0.18 0 59.4 40.6 6.7 56.8

143


This will be most closely achieved late inthe dry season. The effects of flooding areshown in Figure 5.6, where a flood pulsefrom the local source area flushed out themixing reservoir; the chemical concentra-tions could not be strictly correlated withdischarge because of the aquifer hydraulicsand the volume of the mixing reservoir.Even slight flooding will give an aberrantdischarge for the SO4tot during the flood,and vice versa when the discharge is backto normal but the SO4tot is somewhat di-luted.

From the results (Table 7.4), the regionalflux of type B water is predicted to be atleast 12 m3/sec to the springs at or near theeastern margin of the region. The calculatedmaximum possible flux to these springs is

27 m3/sec. It is focused primarily on twovery large springs, the Mante in the north-ern and the Coy in the southern part of thearea. Coincidentally (?) these two springsalso have the largest base flows of all thesprings in the region. The calculated re-gional flux may be increased from by 0 to10 m3/ sec when the springs above the AguaClara in the Lower Tampaón basin (see sec-tion 4.3) are located and sampled. Theamount of increase will depend on theirSO4tot and the dilution curve selected. Thevalue given for the regional flux is only forthe high sulfate water; any regional flowof type A water would not be detected bythe model.

For the springs with more than one cal-culated QB, there is some variation in the

result. This may be caused by the factorsdiscussed previously, i.e., those that maycause slight instability of mixing condi-tions. Alternatively, some of the variationmight be real. This would be of significancein understanding the hydraulics of the re-gional aquifer. From the results shown inTable 7.4, it is suggested that there mightbe a small gradual decrease in QB duringthe dry season, and that QB might be greaterfollowing a wet year(s) than after a dryyear(s). These possibilities should bechecked by carefully selected sample timesin the future. The samples for the Choy andespecially for the Mante are from yearshaving relatively high base flows. It wouldbe desirable to learn what their SO4tot andcalculated QB values are during seasons oflower base flow.

Additional knowledge of the aquifermight be gained by studying the dilutioneffects of floods. During major floods atthe springs, it has previously been shown(section 5.3) that the local source is stronglydominant. If it can be assumed that eachinstantaneous great slug of local watercauses a dilution in the mixing reservoirproportional to its flow and that each mixedslug will eventually appear at the spring,then the maximum spring discharge can becorrelated with the SO4tot minimum in thechemical curves. Using the QB given inTable 7.4 and the peak spring discharge ofthe series of floods (Figure 4.7), the mix-ing proportions are calculated, and then anexpected value is read from Figure 7.6. Thepredictions are very close to the analyticaldeterminations (Table 7.5) for the two ma-jor flood events monitored at El Choy. Thisindicates that the dilution model of theChoy works over a wide range of discharge;specifically, QB remains constant while QAvaries due to local precipitation and dryseason drawdown. However, the two floodsare of comparable magnitude. Similar dataand calculations need to be obtained dur-ing floods of, say, 15 and 100 m3/sec; dilu-tion measurements for 10 to 30 m3/secfloods will be especially critical.

Some of the Mante and the Frío springdata may also be used to check the model.For these springs, there is not enough in-formation to compare the flood peak withthe SO4tot minimum, as was done for theChoy. The data for which the calculationsare made are believed to represent approxi-mately stable mixing reservoir conditions,in which QB is constant, but QA is changingslowly and is about 2 to 5 times larger thanduring the average late dry season. This isa test of the dilution model over a rangeFigure 7.6. Mixing curves of type A with type B ground water.

144


Table 7.4Dry-season regional flow (QB)

Maximum QBQtotal Minimum QB (curve 3)

Spring Date m3/sec m3/sec m3/sec Comments

Choy 05/21/73 2.85 0.60 1.42 Stable conditions at end of long dry season;showers one week earlier

06/07/73 2.85 0.60 1.42 Stable conditions at end of long dry season

12/23/71 4.04 0.80 1.93 Early in dry season after September and Octoberfloods; 40 mm rain on Dec. 21 caused slightincrease in flow

01/15/72 3.64 0.73 0.65 1.76 1.6 More stable than on 12/23/71

05/20/72 3.07 0.67 1.60 At end of long dry season; preceded only byshowers

06/05/71 3.00 0.65 1.53 End of dry season; minor flood of 3.6 m3/sec inlast week of May

Mante1 06/06/71 11.85 3.26 7.56 End of severe dry season; preceded by someMay showers and wet year

05/25/73 11.32 2.75 3.0 6.54 7.0 End of dry season; discharge seems stable; 50 to100 mm rainfall during April–May

01/17/72 14.514 3.09 7.16 Middle of dry season; discharge slowly declining

Coy2 06/08/71 (15) (7.0) (15) Follows one large storm of 140 mm on May 24,(7.0) (15) 1971, but should be about stable

12/23/71 22.2 7.41– 16.1– Early dry season; small flood on Dec. 13

Tanchachin 01/01/72 2.21 0.52 0.5 1.17 1.2 Early dry season

Agua Clara 01/01/72 2 (est) (0.53) (0.5) (1.2) (1.2) Early dry season

Frío springs3 05/25/73 5.40+ 0.46+ 1.03+ Discharge from Poza Azul unknown, butto to probably <0.5 m3/sec; end of dry season5.90+ 0.50+ 0.55+ 1.2+

06/15/71 8+ 0.72+ 1.64+ Follows wet year and severe dry season; Mayand early June showers

Total min. 12.2 max. 27.2

1. P and Q data indicate stability; the available Q data were once-weekly measurements for thecanals; for the stream channel, average daily flows through 1972, and once-weekly measure-ments in 1973.

2. Gauging station Q includes surface drainage and temporary springs; on June 8, 1971, total Qwas 16.2 m3/sec, and an estimated 1.2 m3/sec was subtracted to obtain Coy spring Q; onDec. 23, 1971, a Coy Q of 20 m3/sec gives QB = 6.7 m3/sec.

3. Q of Frío springs used here is Canal Alto + Poza Azul + Río Nacimiento; the + indicatesprobable ungauged springs.

4. Excluded springs are: Pimienta, a seasonal or overflow spring; Puente de Díos and MediaLuna, western springs.

where the predicted value is very sensitiveto small changes in discharge. The calcu-lated values are in error by not more than22%, which is a good result consideringthe limited knowledge of QB and the quasi-stability of the mixing reservoir. Thesecalculations test the proposed two-sourcemixing model for the Choy, Mante, and Fríoover a range of discharge conditions; they

have shown that QB may be considered ap-proximately constant and QA variable.These results are independent of whichmixing curve in Figure 7.6 is selected; i.e.,the calculations are a test of the two-sourcemodel, and not of the specific compositionof the type B water.

The same assumptions and computa-tions as above do not work when applied

to the Coy spring (Table 7.5). It would takea spring discharge of 1000 m3/sec to dilutea constant QB of 7 m3/sec to the measuredSO4tot concentration. This implies that QBis not a constant for the Coy spring duringflooding; and, for the example, it must havebeen reduced to something less than 0.83m3/sec. The contrast in the mixing behav-ior of the springs is believed to result from

145


the fact that the Choy, Mante, and Fríosprings drain directly from large unconfinedlocal aquifers, whereas all of the Coy dis-charge is artesian water (except for the verysmall amount that is derived from infiltra-tion into the small dome itself; see Figure7.2). A large increase in QA seems to havelittle or no effect on QB at the Choy, Mante,and Frío springs, but it hydraulically dampsthe regional flux to the Coy spring.

The above result may explain a puzzleregarding the Pimienta, a spring very nearthe base of the limestone scarp south ofAquismón, at about 100 m elevation. Thetwo water samples collected show that it isa high sulfate spring, yet it is also a sea-sonal spring (Section 4.2). Its aquifer isunconfined. Hence, when ground-waterlevels in the area west of Aquismón arehigh, part of the regional flux as well assome of the local flux is apparently divertedfrom the Coy to the Pimienta, and perhapsto other springs.

7.1.4. Summary and comparison of thehydrochemistry with other regions

The studies reported in this thesis haveshown that there are two major types ofkarst ground water in the region. One is anormal karst calcium-bicarbonate type hav-ing a calcium carbonate hardness of 120 to220 mg/l (type A water); this type resultsfrom flow within the El Abra limestone. The

second (type B) is believed to be a deeplycirculating regional flow system water thathas much higher Ca, Mg, and SO4 concen-trations. Wherever the deep regional waterresurges, these two combine, forming amixed effluent. Hydrochemical techniquesused in conjunction with hydrogeologicinformation often provide powerful meth-ods of studying ground-water flow systems.In this study the chemical and physicalchanges that occur in spring waters duringflooding have been particularly important,because flooding changes the mixing ra-tio. Many of the large eastern springs havebeen found to discharge mixed-source wa-ter (type 2). Based on limited data, theyappear to behave as El Choy, where duringfloods the local source greatly predomi-nates, and the chemistry is reduced to nearlythat of the local source springs (type 1).

Because no samples of type B waterwere available for analysis, it was neces-sary to investigate its possible compositionby other means. A geochemical model wasdeveloped based on sequential exposure tominerals (calcite, dolomite, and gypsum),aquatic chemistry theory, and field resultsof other works. The model took into ac-count the very important ion pairs CaSO4

0

and MgSO40, as well as the much less

significant MgHCO3+. Using the spring

water samples as controls, it was found thatthe type B water had to have specific pro-portions of Ca, Mg, and SO4. This in turn

restricted the appropriate saturation condi-tions with respect to calcite, dolomite, andgypsum; the difference between SIc and SIdcould range from 0.0 to about +0.12, de-pending on the spring, and SIg could varyfreely, except that it was not allowed tosupersaturate. The equilibrium model (at35°C) proved satisfactory for the type Bsource to the Choy spring and had a com-position of about 29.5 meq/l of Catot, 14.5meq/l of Mgtot, and 41 meq/l of SO4tot. Theequilibrium model allowed the highestSO4tot for the regional source. Models suit-able for other springs had up to 10% lessSO4tot than the equilibrium model.

Using a two-source dissolved load equa-tion, which has four independent variablesand two dependent variables, the amountof local and regional flow to the easternmargin type 2 springs was determined forstable base flow conditions. The total re-gional flow of type B ground water wascalculated to be between 12 and 27 m3/sec;the amount as determined by the modeldepends on the degree of saturation withrespect to gypsum in the type B water. Itwas also shown that the regional flux isapproximately constant, possibly exhibit-ing small variations in response to seasonalor longer term changes in hydraulic head.For most of the springs, it appears the re-gional flux is constant even during floods.

There have been numerous studies ofsurface streams that have related dissolved

Table 7.5Flood dilution mixing model calculations

QB from Table 7.4 Qmax of spring QB/Qmax Predicted* MeasuredSpring Sample date m3/sec m3/sec ×100 SO4tot mg/l SO4tot mg/l Note

Choy 7/07/71 0.65 56 1.16 32 30 1

6/15/72 0.65 48 1.35 37 35 1

Mante 8/17/71 3.0 ≈28 10.7 220 180 2

7/11/72 3.0 ≈23 13.0 255 240 3

Frío 12/27/71 0.55 13.2+ 4.17 90 92 4

Coy 6/20/72 7.0 <118 >5.93 >112 25 5

* Using curves 1 or 2 in Figure 7.6.

1. First major flood of the year.

2. Flood of 80 m3/sec (plus Q of canals unknown) on June 25, declining to 19.5 m3/sec on July 29; small,broad flood with approximately constant Q of 28 m3/sec (including canals) from August 5 to 12, falling to24.0 m3/sec on August 19.

3. Flood of 27.6 m3/sec (plus canal flow, if any) in mid-June; constant Q for week preceding sample.

4. Strong wet-season flooding ending in mid-October; Q fluctuates slightly in December.

5. First major flood of year; follows two smaller ones; the peak Q at the gauging station was 118 m3/sec;thus when the Q supplied by surface runoff from 194 km2 of basin and small wet-weather springs issubtracted, the Coy spring Q is less, which affects the calculations as shown.

146


load to discharge, some studies that haveshown annual trends or cycles in waterquality (e.g., Hendrickson and Krieger,1964; Gunnerson, 1967), and, recently, afew that have used a two or three sourcedissolved load equation to separate thehydrograph of surface streams into surfacerunoff and ground-water components(Hem, 1970; Pinder and Jones, 1969;Drake, 1974). Bassett (1976) presentedextensive correlations between chemicalspecies, calculated parameters, and dis-charge for the Orangeville Rise, a largekarst spring in southern Indiana, and someresearchers (initiated by Ashton, 1966) havediscussed some of the chemical and physi-cal variations that occur during floods ofkarst springs in limestone and dolomite ter-rain. This thesis is believed to present thefirst development of a quantitative mixingmodel for ground-water systems. The datagiven by Bassett (1976) and the local geol-ogy of southern Indiana suggest that sucha model could be applied there, and that itmay be useful elsewhere.

Curiously, the thermal, sulfurous TaninulSulfur Pool also responds to local ground-water conditions, as shown in Figure 5.11.Thus it too can be modeled as having two(or possibly more) sources, one of whichis a local recharge area in the Sierra de ElAbra and the other is some unknownsource(s) of hot, sulfurous, and sodiumchloride bearing water. Each of the threethermal springs examined has a differentchemical character and thermal regime. Itis not known if the other two undergo suchlarge changes in discharge, chemistry, andtemperature as the Sulfur Pool. The ther-mal springs discharge a rather ill-definedthird type of ground water that probablydoes not exceed a total flow of 0.1 m3/secin the region.

The flood behavior of small springs wasnot studied, because it is very difficult toreach them during the rainy season. Notvery many of them have been sampled, butit is believed they will usually be localsource springs. During large floods, prob-ably most of the water discharged at thesmall springs and perhaps the major partof that discharged at El Choy are believedto have been derived from the causal pre-cipitation event. The short-resident localsource waters undoubtedly have lower cal-cium bicarbonate concentrations and lowerSIc values than the longer resident localsource waters that are discharged sometimeafter the flood or during the dry season; thedecrease might be about 20 to 40% and 0.2to 0.5 log units, respectively, depending on

the spring, magnitude of the flood, andantecedent conditions. This and certain lageffects associated with the mixing reservoirare of secondary importance in the mixingmodel of the type 2 springs. More data arerequired to make a detailed analysis of themore subtle aspects of the mixing model.The more obvious effects of volume of themixing reservoir and of the phreatic hydrau-lic conditions are the lag of the chemicalresponse behind the discharge increase atthe onset of flooding and the much greaterperiod of time required for the chemistryto return to its normal base flow characterthan for the discharge. This effect rules outthe straight-forward comparison of chemi-cal concentration with discharge.

Besides the need for more data from theregion of interest, there are some questionsor problems with the geochemical model.They are really a group of problems con-cerning some disparities between aqueouschemistry theory and results reported in theliterature, and the need for more laboratorymeasurements, such as the temperature de-pendence of KMgHCO3+ reaction rate data.The only problems that might be thoughtof serious consequence to the model are thestill uncertain value of Kdolomite and the pre-cipitation kinetics and nucleation potentialof dolomite, calcite, magnesian calcites,and possibly aragonite. In Florida, the highMg concentrations relative to Ca and thedolomite supersaturations reported by Backand Hanshaw (1970) clearly pose a geo-chemical problem, but it does not seem tobe a concern for the Valles–SLP Platformarea. The compositional requirements oftype B water for the mixing model as speci-fied by Mg/SO4 and Ca/Mg ratios andtested by extrapolation of chemical data inFigure 7.5 are such that the dolomite satu-ration state relative to the calcite saturationstate is controlled and the Florida problemdoes not occur. This study used a tempera-ture dependent Kd that has a value of 10–17.0

at 25°C (after Langmuir, 1971). Anothercommonly used value for this temperatureis 10–16.7. On the assumption that the dif-ference in the Kds is also 0.3 log units at35°C, the geochemical model was testedto determine what effect the increased solu-bility of dolomite would have; the newequilibrium model is given in Table 7.6.

This model contains 27% greater SO4tot,91% greater Mgtot, and 5% less Catot thanthe other equilibrium model (column 15,Table 7.2). From the Mgtot/SO4tot andCatot/Mgtot ratios and comparison withFigure 7.5, it is clear that this new equilib-rium model is not a satisfactory choice fortype B water. If 10–16.7 (at 25°C) is trulythe equilibrium constant for dolomite,nonequilibrium models similar to those inTable 7.2b must be generated. It was foundthat SIc–SId must be increased by about+0.15 (which is 1/2 log[10–16.7/10–17]) com-pared with the differences given earlier inthis chapter to obtain suitable compositionsfor type B water. In fact, this reduces thechemical concentrations to almost exactlythose given earlier (Section 7.1.2), and nochanges are necessary in the mixing modelcalculations. Changes in SIg alone cannotcompensate for the increased Kd. Thus so-lutions having the required Mgtot/SO4tot andCatot/Mgtot ratios have saturation conditionslimited to specific relationships: SIc–SId isconstant and SIg is variable, but less thanor equal to zero. The important aspects ofa greater Kd, if correct, are the implicationsthat the solution may be undersaturatedwith respect to dolomite, thus suggesting ayounger age than the Florida ground wa-ters, where radiocarbon dates indicate thatdolomite saturation is reached in about10,000 years (based on Figures 1 and 11 inBack and Hanshaw, 1970) and that al-though acceptable type B waters for the areahave a maximum SO4tot of about 41 meq/land reach equilibrium with respect to gyp-sum, if the waters were exposed to dolo-mite and gypsum for a longer period oftime, the SO4tot would increase with theMgtot as dolomite continues to dissolve, andthe molar ratios of the constituents wouldbe changed.

One of the objectives of the water chem-istry studies was to compare the concen-trations of dissolved species, the calculatedequilibrium PCO2, and the estimated denu-dation rate with published results, particu-larly for other climatic regions. However,the chemistry of the Valles–San Luis Potosíregion karst waters has been shown to be morecomplex than normal calcium-bicarbonate orcalcium-magnesium-bicarbonate type wa-ters most commonly reported in karst areas.

Table 7.6Equilibrium model with greater Kd

pH Catot Mgtot SO4tot CO3tot pPCO2 CaSO40 MgSO4

0 Mg/SO4 Ca/Mg

6.8 14.08 13.76 26.15 4.10 1.50 5.19 6.45 0.53 1.02

147


This is partly because many workers haveonly measured the calcium and total hardness(especially for comparing the hardnessesof different regions), and partly because thehydrochemistry of the area is indeed morecomplex. The chemistry of the ground wa-ter in the region of interest here is consid-ered to be a property of the hydrogeologicflow systems, which will be discussed inSection 7.2. Some support for the geo-chemical model is obtained from studiesin eastern Nevada and in Yucatan. Maxeyand Mifflin (1966) list detailed analyses offorty limestone spring waters in Nevada.Here there also seem to be two chemicaltypes The waters have a nearly constanttotal hardness, except two that have a veryhigh sulfate content (see Table 7.7; alsoplotted in Figure 7.5). The concentrationsof some species appeared to increase withlength of flow path. The two samples plot-ted contained considerable Na+K and Cl.It is suggested that they have an excess ofNa over Cl caused by cation exchange andthat this is the reason they fall somewhatbelow the extrapolated lines in Figure 7.5.These samples are from a still more com-plicated hydrochemical system. In Yucatan,Back and Hanshaw (1974) have shown thattwo principal types of chemical trends oc-cur. One is the mixing of normal calcium-bicarbonate type water with varyingamounts of sea water, and the other is agypsum or anhydrite solution trend coupledwith solution of carbonate minerals (as inthe geochemical model developed here) andpossibly with minor mixing with sea wa-ter. The Polyuc sample (from the south-central Yucatan peninsula) had by far thehighest SO4 concentration and is also plot-ted in Figure 7.5. It too has some Na andCl (Table 7.7), but fits the model extremelywell. The Florida data reported by Backand Hanshaw (1970) do not fit the model,

because the Ca/Mg is much lower (theirsample having the highest SO4 is given inTable 7.7). This indicates that dolomitesolution is more important there, that wa-ter might be exposed to the minerals in a dif-ferent sequence than in this model (possiblygreater dolomite solution before encoun-tering gypsum), and that the solution mayevolve along Ca versus SO4 and Ca+Mgversus SO4 lines with different slopes, es-pecially if a greater Kd is allowed.

One of the major controversies in thekarst literature concerns the intensity ofkarst development as a function of climate.The debate has significance for permeabil-ity development as well as surface karstforms. Many authors believed that themaximum intensity (or rate) occurres in thetropics, putting forward the spectacularcorrosional landforms as evidence; the op-position camp maintained that the intensitywould be higher in northern latitudes be-cause carbon dioxide is more soluble incold water than in warm water. Quantita-tive support eventually was sought from thechemistry of karst waters. Because so manyof the El Abra region spring waters haveproven to be more complex than normalkarst waters, not enough appropriate datahave been gathered to offer either a statis-tically based or scarcely even a rationallybased measure of the limestone denudationrate in the region. Hence, an extensive sur-vey of the literature and comparison withpublished results will not be given here.Instead, a general discussion of some of therequirements for such chemical compari-sons will be given (especially with refer-ence to some recently published papers),and the possible implication of the El Abrawaters to the debate will be suggested.

The chemistry of ground water, evenkarst ground water, is most complex. Someimportant factors are mineral solubilities

and degrees of saturation, solution and pre-cipitation kinetics, ion exchange, ionicstrength and common ion effects, aquifermineralogy, effects of trace metals, gaseousphases, diffusion of gases and ions, hydro-dynamic effects including type of flow(granular, diffuse, or conduit), seasonalvariations, and whether the system is openor closed. Thus, it is very important thatonly karst waters that are from appropriateor similar situations be used, if it is desiredto critically and objectively compare theirchemistry and their calculated denudationrates. William B. White and students (Wil-liam B. White, personal communication)conceived the idea of using water tempera-ture as a measure of latitude (and hence atleast a partial indicator of climate) and test-ing spring waters for correlations betweentemperature and various chemical speciesconcentrations or calculated parameters.Russell S. Harmon took up the concept,collected data from a number of coworkers(including this author) and the literature,attempted to prove the idea, and publisheda series of papers comparing averaged val-ues from geographic areas (Harmon et al.,1972, 1973, 1975). Unfortunately, the data,particularly the Mexican data, have notbeen properly handled; therefore, althoughthe published results “proved” the theory,it is here considered to be an untestedtheory.

Some of the principal points of conten-tion are:

1. The data sets used for all the paperscontain analytical data by Harmon from apreliminary report on the Sierra de El Abrakarst waters (Harmon, 1971) which havesome very large analytical errors for allconstituents, sometimes as much as severalhundred percent. The alkalinity and sulfatevalues are especially in error.

2. Figure 7.7, reproduced from the 1973paper, shows the “correlation” betweenCa++ and water temperature. The Mexicopoint is simply the average of the Ca++ val-ues of all samples of all the El Abra regionsprings that were sampled, including thevery high calcium-sulfate type springs.Clearly, the latter waters are not appropri-ate! Of the springs sampled in this thesisstudy, only six (Table 5.1) may be suitablefor such a plot (flow conditions prior tosampling are uncertain for three of them).They have an average Ca++ of 69.5 ppm(the range was 45.5 ppm to 88.1 ppm),which would greatly change the correlationline. Note that the original line has a nega-tive intercept on the Ca++ axis. A value of

Table 7.7Some published data for comparison with the geochemical model

Sample °C pH Ca Mg Na K HCO3 SO4 Cl

Blue Point Spring1 — — 482 169 428* 122 1910 368

Rogers Spring1 — — 451 149 395* 185 1670 343

Polyuc2 26.7 6.71 384 85 98 6.5 354 1020 106

Ellenton3 26.2 7.58 140 68 19 2.7 178 458 25

* value quoted is ppm Na+K

1. Nevada, concentrations in ppm, Maxey and Mifflin (1966)

2. Yucatan, concentrations in mg/l, Back and Hanshaw (1974)

3. Florida, concentrations in mg/l, Back and Hanshaw (1970)

148


1.99 was given as the ratio of the slope ofthe line in their HCO3

– versus T plot di-vided by the slope of their Ca++ versus Tplot; this was taken to show that the waterswere of calcium-bicarbonate type, when inreality, it was a statistical accident whichmisrepresents the true water chemistries.Aside from the calcium-sulfate waters inMexico, many of the waters contain sig-nificant amounts of Mg, and a comparisonbetween total hardness and alkalinity mightbe more appropriate, or better still, use onlyof waters from nearly pure limestone ter-rains.

3. The 1975 paper gives regression re-lationships between a number of chemicalvariables and temperatures, but the use ofsome inappropriate and some unreliabledata negates the effort. Greater emphasisis placed on PCO2 versus T, using |SIc| < 0.1as the criterion for selecting only diffuseflow springs. Although conduit and diffuseflow springs could be distinguished inPennsylvania by the coefficient of varia-tion of SIc or other chemical variables(Shuster and White, 1971), it is clear from

this study that the chemistries and satura-tion values of many of the Mexican springsare strongly influenced by gypsum solution,and that the local flow systems are of theconduit rather than the diffuse type. Itseems very likely that many of the otherchemical analyses taken from the literatureare actually from conduit springs. Thetwenty-nine Mexican samples (all from theCd. Valles region) are from only elevensprings, some having been sampled manymore times than others. Twelve analyseswere assumed to be from diffuse flowsprings based on their SIc = 0.0±0.1. If allthose that have been significantly influ-enced by gypsum solution are deleted, theirMexican data collapses to just one sample(from the Harmon (1971) paper). Some ofthe Texas data may also be influenced bygypsum solution. There is also the prob-lem that probably many of the pHs reportedin the literature were measured in the labo-ratory rather than on location, which makesthe calculated SIcs and PCO2s unreliable.Furthermore, despite the statement thatthe computer program used makes “full

allowance for ionic strength and ion pair-ing,” because the sulfate concentrationswere never included with the input data,the important ion pairs CaSO4 and MgSO4could not have been taken into account, northe effect of high sulfate concentrations onthe ionic strength.

There are additional problems with thedata treatment and analysis presented in theseries of papers, but they will not be dis-cussed here. Using the corrected Mexicopoint in Figure 7.7 and presuming the otheraverages are valid (?), it appears that theremight be a trend, although the slope of aregression line would be much less thanbefore; the Mexico and Canada points arebased on very few samples. Brook (1976)and Cowell (1976) have added a fewsamples to the Canadian data from theirstudies, and they have reanalyzed the data.They found that there may be a slow in-crease in Ca++ with increasing temperature,but that the HCO3

– versus temperature cor-relation was inconclusive. A recent studyhaving data from many more areas in east-ern North America showed that there wasno relationship between bicarbonate con-tent and latitude in the region (Trainer andHeath, 1976). They attribute the variationsof HCO3 concentration primarily to factorswhich control CO2 production and storagein soil, to the type and timing of aquiferrecharge (i.e., how much CO2 is carrieddown with the recharge), and to the type ofaquifer flow system on residence times ofthe water (after Shuster and White, 1971).They did not give a regional comparisonof data restricted to one type of flow sys-tem or aquifer mineralogy.

Figure 7.8 shows a plot of SIc againstpPCO2 for springs in the El Abra region,using samples from dry season conditionswhere possible. The large calcium-sulfatetype springs are approximately at equilib-rium with respect to calcite and have PCO2ranging from 1.72 (1.9%) to 1.26 (5.5%).The Sabinas is the only large non-sulfatespring for which a sample at or near equi-librium with respect to calcite was obtained,and it had the lowest pPCO2, 2.0 (equiva-lent to 1%) of all the springs sampled. Notenough samples were obtained from thesmaller El Abra springs to determine whatvariations occur in their saturation condi-tions or chemical concentrations. Based onearly dry season samples from the SantaClara and Arroyo Seco springs, it appearsthat their effluents may approach near equi-librium with respect to calcite later in thedry season. The effect of flooding is shown

Figure 7.7. Plot of Ca++ against T (°C) for various geograpical regionsfrom Harmon et al. (1973), with a new Mexico point from this study.

149


for the Coy and the Choy springs, wheremoderate floods reduce their SIcs to about–0.4. The interesting aspect of the data forthe Sierra de El Abra springs is that theyhave equilibrated with a high PCO2 reser-voir, from 2.5 to 5%, whereas the surfaceof the range has only a thin, spotty coverof soil. It is suggested that a significant partof the CO2 is generated by decay of organicmaterial washed down into fissures andcaves both above and below the water table.From the calculated PCO2s for spring wa-ters in the region, it is concluded that thekarst ground waters have a large potentialfor limestone solution. Based on the calcu-lated SIcs, (remembering that only six “nor-mal” springs have been sampled), most ofthis potential is used at base flow, but floodwaters flow too quickly through the cavesto use all the potential, as shown by theChoy and Coy, where flood waters are con-siderably undersaturated. The average val-ues for twenty-nine Pennsylvania springs(includes both diffuse and conduit types,from Langmuir, 1971) and eight Floridawell waters (those unaffected by gypsumsolution or ocean water mixing, from Backand Hanshaw, 1970) are shown in Figure

7.8 for comparison. Because the Floridawaters are old, often several thousands ofyears, and evidently are from relatively dif-fuse flow systems, they perhaps are notcomparable. Four normal karst waters fromYucatan (Back and Hanshaw, 1974), whichhas a conduit type aquifer, have averagesof +0.015 and 1.61 for SIc and pPCO2, re-spectively. Hence, the Yucatan karst waters,before they mix with sea water, appear tohave about the same potential for solutionas El Abra region waters, but utilize thatpotential more fully, as shown by their SIcvalues and their average of 92.5 mg/l ofcalcium. The Yucatan waters must have alonger residence time than the El Abra wa-ters, and the Yucatan aquifer is likely lessdynamic than the El Abra aquifer, althoughboth areas receive similar amounts of pre-cipitation and are affected by large stormsand hurricanes. It is concluded that thepotential for karst solution, as indicated byPCO2, may be generally higher in the trop-ics than in more northerly climates, but theextent to which the potential is utilizeddepends on the type of aquifer and otherenvironmental factors. More high qualitydata from selected environments are needed

before this conclusion can be consideredproven.

7.2. The Ground-Water Hydrologyof the Region

“A ground-water flow system is a regionwithin saturated earth materials where thereis dynamic movement of ground water froma source [recharge zone] to a sink [dis-charge zone]” (Mifflin, 1968, p. 3). Waterat each point within a flow system possessesa specific velocity, fluid potential, chemis-try, temperature, viscosity, turbidity, anddensity. A system is perfectly defined or“known” when the value of all the param-eters are known at all points, i.e., as a func-tion of their spatial distribution (they arereferred to as distributed parameters). Themeasurement of these parameters at even amodest number of points is exceedinglyexpensive and time consuming. Hence it isusually only possible to hypothesize thegeneral characteristics of the ground-waterbody based on geologic and geomorphicrelationships, output characteristics of theflow systems, inference of likely rechargezones and recharge capacity, measured orestimated permeabilities, and limited welldata. Karst springs represent the integratedoutput of karst groundwater flow systems,and the same parameters as above, but alsoincluding the discharge, may be studied.The values of the parameters measured at aspring are the average for all of the flowpaths that reach the spring; their magni-tudes and temporal variations, as shown inthis study, provide a great deal of informa-tion about the gross characteristics of anaquifer.

One of the principal contentions of thisstudy is that the physical and chemical char-acteristics of the spring waters in theValles–SLP region and the elevation anddischarge of the springs together have aconsistent and reasonable pattern. They arefunctions of ground-water flow systems,and the variables of interest are geologicsetting, aquifer hydraulics, fluid potential,climatic events, chemical reactions, heatflow, heats of reactions, and time. In thepreceding section, a schematic hydro-chemical model of the flow systems wasdeveloped that was based on spring waterchemistry (plus a few chemical measure-ments within the aquifer), chemical theory,and the schematic distribution of minerals.It successfully modeled the mixing phenom-ena of the high sulfate springs, estimatedthe amounts and locations of regionalground-water discharge, and showed that

Figure 7.8. Calcite saturation index (SIc) versus negative logarithm of partialpressure of carbon dioxide (pPCO2) for region springs.

150


the extremely variable output of the springscould be separated in most instances intodynamic local sources and approximatelyconstant flux (of water and chemistry) re-gional sources. In this section, the infor-mation and interpretations of previouschapters and the hydrochemical model aredrawn together to provide a summary andspeculative discussion of the real ground-water hydrology. Although there are no dataon ground-water as a distributed parameterexcept for a few in the Sierra de El Abra, itis possible to gain some knowledge of vari-ous portions of the Platform by analyzingthe discharge characteristics of each of thebasins given in Chapter 4.

7.2.1. Geology and hydrogeology

Figure 7.9 is a regional cross sectionbased on the topography along 22°N lati-tude (through Cd. Valles) of the new1:50,000 maps by CETENAL, the geologyand well records as given by Carrillo Bravo(1971), Heim (1940), and Muir (1936), anda geologic map of the state of San LuisPotosí provided by the Instituto de Geo-logía (1965, compiled by Ing. LopezRamos). Names enclosed in parenthesesgive the relative location of features thathave been projected into the section fromsome distance away. No structural crosssection of the region has been previouslypublished, probably because subsurfaceinformation is still scanty. This simplifiedsection emphasizes hydrogeologic aspects.Shown are the broad coastal plain of im-permeable rocks, the Sierra de El Abra atthe eastern edge of the folded Platform, thehigh Sierra Madre Oriental fold belt, andthe broad Río Verde valley. This sectionextends about 60% of the way across thePlatform, just north of the structurally low-est part of the Sierra Madre. Nearly all ofthe ridges are anticlines. The thick sectionof Upper Cretaceous rocks has been re-moved from most of the region, and the ElAbra Formation is either exposed or cov-ered by only a thin veneer of sediments overmost of the Platform. The Guaxcamá For-mation is shown here to have a thicknessof about 2000 m except where thickenedby folding, because according to CarrilloBravo (1971), it extends downward to 2110m below sea level in Pozo Agua Nueva(penetrated 120 m of basal Cretaceous andUpper Jurassic before reaching redbeds)and to –1834 m in Poza Guaxcamá (westof the section), where 3009 m of the for-mation was cut (underlain by 800+ m ofredbeds). Thus, the base occurred at

roughly the same elevation. The easternlimit of the Guaxcamá Formation is un-known.

The principal aquifer of the region is thereef and platform limestone facies of theEl Abra Formation, where the flow isthrough large conduits (caves) and solu-tionally widened bedding planes and joints.Because the large tectonic structures arefolds rather than faults, the aquifer is con-tinuous over most of the region. One knowndiscontinuity in the aquifer is the Guaxcamáanticline, where erosion has removed theEl Abra, exposing gypsum in its interior.Possible discontinuities in the ground-wa-ter body may occur in the Xilitla area,where faulting is important, the Miqui-huana arch, where Precambrian (?) meta-morphic and other pre-Cretaceous rocks areexposed in the interior of the fold, and inthe interior of the Sierra de Guatemala,which has great structural relief (Figure2.5). The Tamasopo Formation is part ofthe same karstic aquifer, but it is volumetri-cally much smaller. The Tamasopo and ElAbra Formations plus similar Lower Cre-taceous platform facies limestones consti-tute a hydrostratigraphic unit (Maxey,1964). The basal and forereef dolomite ofthe El Abra Formation has been found inwells to be porous and permeable. In PozoTamalihuale No. 1, Carrillo Bravo (1971)reported that the dolomite had “frequentand small caverns, irregularly distributed”;hence the flow may be more restricted thanin the limestone caves, i.e., it is probably afractured dolomite aquifer. The GuaxcamáFormation has some limestone and dolo-mite beds within the gypsum and anhydrite.It does not seem likely that there would bea significant flow in this unit, but CarrilloBravo (1971) reported a loss of circulationduring drilling. The only other aquifers arethe late Tertiary or Quaternary alluvial fan,fluvial, and lake deposits. These sediments,which are relatively thin, occur mainly inthe broad flat valleys and bolsons west ofthe Sierra Madre Oriental folded ranges(Figures 7.9 and 4.1). They are quite vari-able in type and size of sediments, andhence in permeability. Based on a reportby J. Lesser (unpublished, circa 1973), theyare of limited importance as aquifers in thetotal regional picture developed here.

7.2.2. Spatial variation of runoff

In Chapter 4, some aspects of the runofffrom the region were discussed. Table 4.9lists the yearly discharge for the study pe-riod 1961–68 of the S.R.H. gauging stations

in the area. The average unit runoff wascalculated for the supposed surface basinof each station, and this result was com-pared with the average value for the wholeregion. A specific relationship between an-nual precipitation and annual runoff has notbeen possible, because the true boundariesof the basins, surface and subsurface, arenot known. Therefore, based on the discus-sion of Section 4.3, Figure 7.10 distin-guishes the areas that are believed to haveclearly excessive or deficient average unitrunoff from those that have values deemedreasonable for the given rainfall (see Fig-ure 4.3).

The basins are the same as those givenby the S.R.H. in Bulletin 32 (and in Figure4.2), except that the Sierra de El Abra, in-cluding the drainages in the adjacent west-ern valley that are pirated underground, isdelineated as a karst area. More detailedmapping of the region will reveal other ar-eas that have totally or partially internaldrainage. The hydrogeologic map (Figure4.1) may be used to infer areas of infiltra-tion or surface runoff within the basins ofFigure 7.10.

It is striking what a large proportion ofthe Platform has a deficient yield comparedwith the proportion that has excessive yield.Excluding that part of the Río Santa Maríabasin west of the Platform, but includingthe small areas of surface basins beyondthe eastern reef trend, 58% of the Platformhas below the expected runoff, only 7.3%has considerably more runoff than ex-pected, and 34% can not be clearly assignedto either category. This distribution is duemostly to karstic effects, even in those ar-eas not marked K (K indicates areas of com-plete internal drainage). For example, theRío Comandante (Estación La Servilleta),which has roughly one half of its area ashigh relief El Abra limestone and the otherhalf as low elevation impermeable cover,produces sharp flood pulses, deficient av-erage flow, and virtually nil base flow.During the study period 1961–68, theComandante basin of 2532 km2 had lessrunoff (4.95 m3/sec) than the Choy spring(5.07 m3/ sec). The lower Micos basin ap-pears to be a zone of karst losses. Thenorthwestern portion, about one third ofthe total Platform area, is not yet integratedinto the surface drainage network of theregion. However, despite its low annualrainfall, there must be a significant amountof infiltration to the ground-water zone. TheEl Abra limestone, which forms the moun-tain ranges and underlies the sediments in thevalleys, has undergone sufficient solution to

151

AMC

S Bulletin 14 — C

hapter 7

Figure 7.9. Regional E-W hydrogeologic cross-section showing hypothesized local and regional ground-water flow systems.

152


Figure 7.10. Map summarizing spatial distribution of runoff in the region, possible directions of ground-water flow, andbase flow chemistry, discharge, and elevations of the major springs.

153


allow some infiltration. Carrillo Bravo(1971) reported frequent loss of circulationin the El Abra and Guaxcamá formationsin the Agua Nueva No. 1 well west of Cd.del Maiz, a “dolina” (in this case a cavity)in the middle of the El Abra limestone inthe same well that has a vertical dimensionof 251 m and is filled with an upper Ter-tiary deposit of clastic material, and gen-eral lack of success (repeatedly empha-sized) with their seismic work in the entirenorthwestern area, which he attributed tothe “cavernous” limestones (and in someinstances to gypsum). Personal observa-tions of gypsum caves near Matehuala andin the same valley farther south, one ofwhich had a small flowing stream 25 mbelow the surface, show that the valley sedi-ments permit some water to escape the zoneof evapotranspiration. J. Lesser (unpub-lished report, c. 1973) has estimated theaverage ground-water discharge into theRío Verde of the sedimentary aquifer of thelarge Río Verde valley to be 0.5 m3/sec forthe period 1962–70. It is believed that mostof the northwestern region acts as a re-charge zone to deeply circulating ground-water flow systems, although the rainfallis not large and there is no surface runoff.This ground water is believed to flow tosprings along the surface drainage network;therefore the surface basins shown in Fig-ure 4.2 have a larger catchment than shown,and the actual runoff derived from them isslightly smaller than shown in Section 4.3.

The basins having a “normal” amountof runoff are principally of two types, thelow to moderate rainfall western areas thathave been dissected by a major river re-ceiving both surface runoff and ground-water discharge (water budget calculationsare subject to great errors in such areas),and the eastern basins that have a large per-centage of impermeable rocks in their val-leys. The one exception is the upper Micosbasin (above Estación El Salto), which isdominantly karstic and has neighboringkarstic ranges to the north and south. Thishigh rainfall basin is tapped by the Río delos Naranjos and has a yield that verges onthe excessive (0.022 m3/sec/km2 for 1949–1964, S.R.H. Boletín 32).

The basins having excessive averageyield occur along the eastern margin of thePlatform or in the lower reaches of the RíoTampaón. All of the known major springsin the region except the Media Luna liewithin these basins; however, it is likelythat some additional second or possiblysome first magnitude springs will be foundalong the Río Santa María above Estación

Tansabaca up to and including the Ayutlaarea, in the upper Micos basin, and in theGallinas basin. From the basin relationshipsshown in Figure 7.10, the areas having sur-plus flow undoubtedly derive most of theirextra water from adjacent, higher karst ar-eas. The lower Tampaón basin, which over-all has excessive yield, was divided on themap into two parts, excessive (karstic) andnormal (impermeable Valles valley). It isestimated that the total regional runoff maybe divided as follows: 33% from the majorsprings shown in Figure 7.10 (excluding

the Puente de Díos), 15% from the imper-meable valleys, and 52% from the combi-nation of smaller springs, wet seasonsprings, and direct surface runoff from karstareas.

7.2.3. Base flow and flood flow

Insights into the hydraulics and storageof the aquifer and into the types of flowsystems that may be present may be gainedby studying the low and high discharges ofthe basins. Section 4.3 discussed the low

Table 7.8Ratio of minimum to average flows for basins

Ratio*Station Elevation (m) 1 May 1963 28 March 1965

Media Luna 1000 (80 to 95)

Coy 31 (at spring) 53.0 53.9

Mante 80 (at spring) 53.5 —

Vigas 897 40.8 35.5

Choy 35 (at spring) 37.6 33.3

El Salto† 399 (29.3) (39.6)(upper Micos basin)

La Encantada (310)‡ 28.9 27.7

Santa María 174 29.9 22.5(Tansabaca – Tanlacut)

Tampaón lower 28.2 23.0 26.7

Frío springs (90) (at springs) 19.4 25.8

Tansabaca 174 22.6 19.6

El Pujal 28.2 15.8 17.9(R. Tampaón)

Micos 210 14.8 13.5

Tampaón upper — 12.6 14.1

Santa Rosa 72.8 9.96 13.5

Gallinas 280 5.17 9.06

Tancuilín (100?) 5.81 7.32

Valles valley 72.8 negative 13.3

Tanlacut 257 0 10.8

Requetemú 71.1 4.50 5.45

Río Huichihuayán — (~100 at springs) 3.86 4.54

Sabinas 89 (160 at spring) 1.95 1.74

La Servilleta 75 0.04 0.04

Rayón basin 257 negative negative(Tanlacut – Vigas)

* (Qbase on date/Q8-year avg)×100

† 16 yr. average of 20.499 m3/sec for 1949–64 (S.R.H. Boletín 32)

‡ ( ) = estimated or approximate

154


and average flows. The eleven great springsshown in Figure 4.10 account for 53% ofthe estimated average annual minimumregional discharge (taken to be about 75m3/sec). The Puente de Díos, west ofTampasopo, is not included because it isnot known whether it is a piracy across ariver bend or a genuine point of ground-water discharge; it is probably the latter.Including the lower Tampaón basin and theunguaged springs of the front ranges (Table4.6), about 70% of the annual minimumflow is discharged at less than 100 m el-evation. In the southern zone, the Coy isby far the largest spring (about 15 m3/sec),but many others are also important (Figure4.10). In the northern zone, only the Mante(8 m3/ sec) and the Frío springs (6 m3/sec)maintain large base flows. The sources ofthe base flow (2 to 5 m3/sec) of the RíoGuayalejo above Estación La Encantada arenot known, but are very likely springs.Excepting the upper Micos basin, the karstbasins at intermediate to high elevations,such as Gallinas and Sabinas, have rela-tively small base flows.

For each basin, the relative importanceof base flow versus wet season flooding inthe total surface runoff (that which leavesthe basin on the surface as measured at thegauging station) may be determined by theratio Qminimum/Q8-year average (Table 7.8). Val-ues from Table 4.6 are used as representa-tive minima, because the lowest recordeddischarge for each station given by theS.R.H. (Boletín 32) is often affected. Thesmaller the percentage, the greater the de-pendence on flooding, or conversely thelarger the percentage, the greater its depen-dence on karstic base flow.

No single factor is apparent that willexplain the variable ratios calculated inTable 7.8. Rather, it is believed that theratio is dependent on the combination ofelevation, geographic location, geomor-phology, type of rock exposed, and aquiferhydraulics. The first two of the above fac-tors relate the ground surface to localground-water levels. Generally, among agroup of springs in an area, the one havingthe lowest elevation has the greatest baseflow.

7.2.4. The ground-water system

In this thesis, it is advocated that a re-gional ground-water system exists compris-ing the whole of the Valles–San Luis PotosíPlatform. The mathematical framework ofsuch flow systems has been given byHubbert (1940), Toth (1963), and Freeze

and Witherspoon (1966, 1967). Severalfield studies of regional systems have beenreported, and Toth (1972) has summarizedtheir characteristics. The most interestingwith respect to variety of supportive dataand similarity to the El Abra region are theground-water systems in the carbonaterocks of eastern Nevada, as described byMifflin (1968), Maxey (1968), Eakin(1966), and Maxey and Mifflin (1966). Insuch groundwater bodies, the flow is di-vided into shallow local flow systems anddeep regional flow systems; sometimes anintermediate flow system is added. Theevidences developed here that support theexistence of such a system in the El Abraregion are (1) chemical character of thewater, (2) spring water temperature, (3)aquifer hydraulics as deduced from thehydrochemical modeling (section 7.1.3),(4) spatial variations of average yield, par-ticularly the existence of areas that have nosurface runoff and that are capable ofground-water recharge, (5) the concentra-tion of most of the flow to a few springs ator near the regional base level, (6) the Coyspring, which unequivocally must obtainits discharge from some distant area(s), (7)exploration oil-well records that demon-strate deep solution effects, permeability,and sometimes actual flow, and (8) conti-nuity of the rock body of the supposed aqui-fer over nearly all of the region.

Fundamental elements of most studiesof flow systems in theory and in the fieldare Darcy flow through granular media andthe concept of fluid potential. The ground-water body will have a smooth, continuouspotentiometric surface. Where the poten-tiometric surface lies below the groundsurface recharge may occur, and where ator above, water should appear on theground surface provided there are no con-fining beds. Because the flow of karst sys-tems is so strongly integrated into discreteconduits (particularly in the Valles–SanLuis Potosí Platform) and the flow is com-monly turbulent, the application and inter-pretation in karst regions of the conceptsof phreatic zone, water table, and piezo-metric or potentiometric surface have beenquestioned and debated (Drew, 1966; Whiteand Longyear, 1962; LeGrand and String-field, 1971; and Thrailkill, 1968, to namebut a few). It is clear that spatial and tem-poral variations of aquifer hydraulics andextreme concentration of flow limit the util-ity of the above concepts in many karstaquifers. It is believed that the El Abra aqui-fer has a phreatic zone within which allopenings that have undergone significant

solutional enlargement are conduits filledwith water that may range in size from cavesto fissures and opened bedding planes, allconduits are integrated with karst flow sys-tems, and those openings at shallow depthwithin the phreas have a distribution offluid potential that forms a smooth, con-tinuous surface when viewed at the regionalscale. The supporting evidence, althoughmeager, includes the observations in theSierra de El Abra (discussed in Section 7.3),the existence of regional flow which hasan approximately constant discharge as cal-culated in Table 7.4, the hydrogeologic set-ting and stability of discharge of the MediaLuna spring, the very large number ofphreatic conduits that must occur in thislarge region, and the flooding of deep wells(Carrillo Bravo, 1971). At the micro scale,there will likely be a discontinuity of hy-draulic gradient between slightly openedfissures or bedding planes and large con-duits, and the potentiometric surface wouldbe quite bumpy or even discontinuous.

Based on the criteria listed in Table 7.9,a hypothesized configuration of the dryseason potentiometric surface is shown inFigure 7.11. Cross sections are shown inFigures 7.9 and 7.12. The map is of coursevery crude and highly interpretive, the largesprings and to some extent the rivers beingthe only positive fluid potential data. Theremust be a relationship between the dis-charge of a spring and the size of its re-charge area (also the infiltration rate).Hence, regardless of variations in the aqui-fer hydraulic regime, Figure 7.11 estimatesthe fluid potential available at each pointfor driving the flow system, the level towhich water would rise in a well if a water-bearing conduit were tapped, and the netdirections of regional water movement.Local flow systems may completely disre-gard the gross regional gradients shown.Some local anomalies occur, such as the100 m elevation changes at the Micos can-yon cascades and at the Salto del Agua cas-cades, where a major stream may beperched for a short distance. Unless themajor rivers are perched and receive theirwater from vadose storage, the general pat-tern of the potentiometric surface must besomething like that shown in Figure 7.11.Much more information, such as influentand effluent reaches of the major streams,possible geologic barriers to ground-waterflow, water tracing, cave surveys, and di-rect measurements of fluid potential, isneeded.

The ground-water body of the Platformis divided into a large number of flow

155


Figure 7.11. Hypothesized regional dry-season potentiometric surface.

systems that feed the springs. These sys-tems are classified into three types (see Fig-ure 7.9): local flow where recharge waterinto a range is discharged to an adjacentvalley, intermediate flow where rechargeinto a range is discharged to a non-adjacent,but nearby valley or at great distance alongthe strike or where water from a large karstplateau or massif is discharged at periph-eral springs , and regional flow where re-charge water circulates deeply and passesthrough the tectonic structures for greatdistances before reappearing at the large

springs. (An example of the second type isthe large area of karst ranges and valleyseast of Cd. del Maiz that has no surfacerunoff and likely discharges most of its re-charge to the Gallinas and upper Micosbasins.) Presently, the distribution of sys-tems between the first two types is more aconceptual matter of scale than a provenseparation of flow systems having distinctboundaries; thus local is often used in thisthesis to include local and intermediate. Forthe large eastern margin springs, high sul-fate and magnesium contents in conjunction

with an anomalously high temperature arebelieved to indicate the presence of a re-gional component in the spring discharge.

The potentiometric map and knowledgeof discharge and likely recharge areas, par-ticularly those with deficient runoff, wereused to infer the general source areas anddirections of flow of the local and regionalground-water systems (Figure 7.10). Thearrows indicate probable or possible gen-eral directions of water movement, not spe-cific source-to-spring flow paths, becauseno such relationships are yet known. For

156

AMC

S Bulletin 14 — C

hapter 7

Figure 7.12. N-S cross-section of front ranges of Sierra Madre Oriental with regional base levels, major springs, and potentiometric surface.

Table 7.9Data used to construct the potentiometric map

1. Elevation profile of: Río Tampaón (Santa María)Río VerdeRío GuayalejoRío MicosRío GallinasRío CamandanteRío Moctezuma

2. Elevation and base flow of gauging stations

3. Elevation and base flow of major springs

4. Elevation of land surface in the west and in the karst valleys or valley margins thathave no surface water—gives the maximum potential

5. Caves in the El Abra range

157


example, the recharge water that eventu-ally leaves the karst area east of Cd. delMaiz via the subsurface might flow to theCoy, to springs along the Río Tampaón, RíoGallinas, or Río de los Naranjos (upper orlower Micos basins), or possibly even tothe Choy or Mante springs. The generaldirections of regional ground-water flowmust be to the east or southeast. Local flowsystems may be oriented in any direction.Intermediate systems will normally have aneastward component of movement, but mayalso have a large north or south component.

To summarize, many lines of evidencesupport the existence of regional, as wellas local, components of flow to the largeeastern margin springs of the Platform.Such systems are usually quite difficult toprove by water budget calculations alone,but here the combined evidence seems con-clusive. It is hypothesized that the lime-stone aquifer has a phreatic zone, which atthe regional scale is smooth and continu-ous and indicates the general amount ofhydraulic head available to drive regionalflow systems. At the local scale, flow sys-tem theory may break down if ground wa-ter is concentrated into too few conduits.

The magnitude and amazing stability ofthe Media Luna discharge (Table 7.8) mustresult from the facts that its likely sourceareas lie far to the northwest and that largeprecipitation events are rare compared withthe Valles area. The well water-level dataof J. Lesser (unpublished S.R.H. report)show that the Media Luna discharge is un-related to the flow system of the Río Verdevalley sediments and thus could only bederived from the El Abra and GuaxcamáFormations. The morphology of the springand the concentrated discharge point (ratherthan a long zone of discharge) also supportthis conclusion. The Sierra Madre fold beltcreates a partial barrier, which helps tomaintain a high zone of saturation west ofa N-S line from Tula to Rayón. The MediaLuna spring probably discharges a consid-erable part of the water that infiltrates inthe limestone mountains west of the RíoVerde valley. The surprisingly high ratioQmin/Qavg at Estación Vigas (Table 7.8) iscaused by ground-water seepage from theRío Verde valley sediments (0.5 m3/sec) andreturn irrigation water from the Media Luna(1.8 m3/sec; data from J. Lesser, unpub-lished report). The Nogal Obscuro station(elevation 1027 m, only 27 m above theMedia Luna), which gauges the upper halfof the basin above Vigas and is situated atthe eastern edge of the limestone moun-tains west of the Río Verde valley, has zero

discharge during most dry seasons. It is notknown if there is hydraulic continuity be-tween the valley sediments and the adja-cent El Abra water bodies.

The large fraction of the total dischargeof the Mante, the Coy, and the Choy that isprovided by their base flow is not surpris-ing, considering their elevation and loca-tion; they receive water from both local andregional sources. The Sabinas spring (160 m)has only a modest dry season flow (0.3 m3/sec) because the Frío springs (about 90 m)have the elevation advantage to capturenearly all (95%) of the base flow from themassive Sierra de Guatemala and rangesfarther west (Figure 7.12). The Frío springsalso capture about two thirds of the totaldischarge from that area. This is an exampleof the control of elevation on local and in-termediate to regional flow systems.

During the wet season, the dynamics ofthe aquifer become most complex. Largestorms apply a stress to the aquifer, causingrapidly changing water levels and hydrau-lic conditions. Individual flow channelsbehave according to the amount of inputand their capacity to transmit the water sup-plied. When this transient state exists, thepotentiometric surface is undoubtedly dis-continuous, extremely variable, and gen-erally unmappable. Note that in Figure 7.9the trace of the potentiometric surface hasnot been shown as a subdued version ofthe topography, although it might have sucha profile, because in the high karst area ofthe Sierra Madre Oriental both the ridgesand the valleys act as recharge sites. Thewet season conditions are unknown, willbe very complicated, and need a great dealmore field work. The flood-pulse chemicaldilution analysis of some eastern marginsprings presented in Section 7.1.3 suggeststhat the local and intermediate flow systemsare the dynamic systems, while the regionalflow systems are discrete and relativelyunaffected by wet season events. The verylow ratio of Qmin/Qavg (Table 7.8) for suchbasins as Río Gallinas and Río Sabinasmeans that in those basins there is relativelylittle water in gravity storage above thesprings and most of the flood recharge isquickly discharged. More discussion of lo-cal flow systems is given in Sections 7.3and 7.4.

The presence of high sulfate concentra-tions in such eastern margin springs as theChoy and the Coy is a clear demonstrationof a regional component in their discharge.This may not be true for springs in highranges 50 km to the west. As may be seenin the cross section (Figure 7.9), tectonic

deformation is expected to have raised gyp-sum (or anhydrite) to a few hundred meterselevation above sea level in the high sier-ras, causing a considerable thinning of thephreatic zone. Some of the folds are of suf-ficient breadth to contain a gypsum coreridge. (Carrillo Bravo, personal discussion,suggests that many of the folds are of theinjection type.) Farther north, in a large areaeast of Cd. del Maiz, some of the folds arehigher yet and have a structural and topo-graphic relief of about one kilometer. There,gypsum may be brought up to an elevationof 500 m or more, and would thereforenearly reach the level of the valley bottoms.Under these conditions, it seems quitelikely that many of the intermediate andlocal systems may obtain moderately highsulfate concentrations. The Puente de Díosspring above Tamasopo may represent thiskind of situation, and some of the springsalong the Río El Salto probably will con-tain sulfate. Another effect of the gypsumridges or of longitudinal fracturing mightbe to channel water along the strike to thesoutheast.

The boundaries of the regional ground-water body are not known, and thus will behypothesized. In general, the extent of thesystem is probably delimited by the aerialextent of the El Abra Formation, includingthe forereef, reef, and backreef to lagoonalfacies. In other words, a paleogeologic unit,the Valles–SLP Platform, effectively de-fines the zone of significant ground-watermovement. The deep water sediments of theMesozoic Basin of central Mexico (CarrilloBravo, 1971) are not expected to transmitlarge quantities of water eastward into theEl Abra Formation; some of the volcanicrocks and the valley-fill sediments west ofthe Platform might transmit a small quan-tity into the region. However, it is knownthat there is some ground-water flow in thatarea; Waring (1965) lists a radioactive ther-mal spring 40 km south of San Luis Potosíhaving a very large flow of 5.27 m3/sec andtwo smaller thermal springs in the Queré-taro area. The northern and southern mar-gins of the Platform must be consideredpartially open, because the reef trend con-tinues as a narrow belt to the south, andthe Cuesta del Cura, the deeper waterequivalent of the El Abra to the east andthe north, probably can develop some sec-ondary permeability. However, the princi-pal flow direction is thought to be eastward;hence ground-water gains or losses acrossthe northern or southern margins should notexceed a few percent of the total regionalbudget. Finally, as may be seen in the cross

158


section (Figure 7.9), the juncture of thePlatform and the ancient Gulf of Mexicodeep water sediments clearly forms the east-ern boundary of the flow systems. Somepermeability has been reported in the tran-sition zone from forereef dolomite and reefdebris to the Cuesta del Cura Formation,where some wells have been invaded bywater (Carrillo Bravo, 1971), and in someoil fields near Tampico the TamaulipasFormation possesses some fracture perme-ability. The coastal plain is the spill pointof the aquifer; all ground-water flow mustreturn to the surface at this juncture, be-cause that is the easternmost point at whichit can escape.

7.2.5. Ground-Water Temperature

The analysis of ground-water tempera-ture is placed here rather than in the chemi-cal modeling section because it can be bestunderstood in the context of the local andregional flow system analysis given in thissection, particularly with reference to Fig-ures 7.9 and 7.12.

The observed distribution of spring wa-ter temperatures is given in Figure 5.4.Excluding the thermal springs and the Me-dia Luna, the range was from 19.6 to26.5°C. Springs such as the Choy and theCoy exhibit larger temperature changesduring flooding than they do seasonally.The temperature difference between earlyand late dry season measurements of theSanta Clara and San Juanito springs wasonly 0.1°C, and at Arroyo Seco was 0.2°C(Table 5.1).

There are very few measurements oftemperature as a distributed parameter ofthe aquifer. A few are reported by Mitchellet al. (1977) and Bonet (1953a, 1953b); allare from the Sierra de El Abra, except thosefrom the Xilitla area given in the latter pa-per by Bonet. Furthermore, all except twoor three are from perched cave lakes, manyof which are sufficiently close to a caveentrance to be controlled or strongly influ-enced by the average seasonal temperature.The range seems to be about 19 to 26.5°Cfor the Sierra de El Abra.

As previously indicated, there is a rela-tionship between the temperature of aspring (during dry season base flow), theelevation of the supposed recharge area,and the chemistry of the spring. Thosesprings which yield nearly pure calcium-bicarbonate type water have the expectedinverse relationship between temperatureand elevation of the recharge area. ArroyoSeco (Rancho El Peñon) and the Santa

Clara springs (in the Sierra de El Abra) werevery similar at 23.6 to 24.1°C, whereas theSabinas, Huichihuyán, and San Juanitosprings were cooler at 19.7 to 21.0°C (seeFigure 7.12). The discharge of all thesesprings is interpreted as coming from localflow systems with relatively shallow cir-culation and short residence time. The highsulfate springs, with their regional compo-nent of flow, have higher temperatures de-rived from deeper circulation that has alonger residence time. The flood pulsestudies described in Section 5.3 were par-ticularly valuable in interpreting the thermalcharacterisitcs of the ground-water system.The correlation between the sulfate contentand the temperature of a spring support themodel in which the regional flow is warmedabove the temperature of the local flow. Nooverall direct relationship exists betweenthe chemical and physical properties of thesprings as a group. On the contrary, eachhas its own physico-chemical characteris-tics dependent on the characteristics of itslocal source(s) and of its regional source(s).The Choy and the Mante have nearly iden-tical geochemical paths, and therefore thesource areas of the Choy and the Mantesprings are believed to be similar in char-acter. The springs differ principally in mag-nitude of flow and mixing ratio. The Coyspring has an offset path because its localsource area(s) has different climatic condi-tions, and, as previously shown, its regionalsource has slightly different chemical prop-erties than those of the Choy and Mante,and thus may have a different temperatureas well.

To summarize, the principal water-temperature controls are believed to be lo-cal climatic conditions and geothermal heatflux in the local flow systems and only thegeothermal heat flux in the regional flowsystem. The temperature would vary withposition in each flow system, graduallybecoming warmer from the recharge pointto the discharge points. Other influencesthat are believed to be less important aregeochemical reactions, such as solutionprocesses or mineralogic changes of anhy-drite to gypsum or vice versa and, in somezones of nearly stagnant flow, the oxida-tion of hydrocarbons and the reduction ofsulfate ions, possible anomalies around in-trusive bodies, and lag effects induced bythe reservoir rock as cooler flood pulsewaters alternate with warmer local or mixedflow waters.

In the proposed hydrologic model of theregion, the temperature of the low sulfatesprings is presumed to be the integrated

result of the above effects on type A water.The temperature of the type B water willbe estimated by a mixing equation, withnotation as in Section 7.1.3. Assume T2 = TC,i.e., there is no change in the temperatureof the mixed water before it is discharged.QA′∆TA = –QB′∆TB, where ∆TA = T2–TA,∆TB = T2–TB, and QA′+QB′ = 1. Therefore,∆TB = –(QA′/QB′)∆TA. T2 is measured, QA′and QB′ are calculated as in Section 7.1.3,and TA must be estimated by some means.For the Choy on May 21, 1973, T2 = 26.4°C,TA = 24.8°C by extrapolation from Figure5.9, and QA′ and QB′ are calculated frommodel number 15 (equilibrium model) inTable 7.2. Then ∆TB = –(0.79 / 0.21)×(26.5–24.9) = –5.64°C, and TB = 26.4°C+5.6°C = 32.0°C. If 24.0°C is used for TA,based on measurements of two smallsprings and selected cave lakes, a value of35.4°C is obtained. These and other resultsare given in Table 7.10. Chemical modelsof type B water that are at equilibrium withrespect to gypsum yield temperatures rang-ing from 30 to 37°C. The Coy has distinctlythe lowest calculated TB, as it had with thedissolved species (Section 7.1.3). If thedeep regional flow has not reached chemi-cal equilibrium with respect to dolomite orgypsum, then QB′ must increase and QA′must decrease (Section 7.1.3), causing thecalculated TB to decrease as shown for theChoy in Table 7.10. The differences be-tween the early and late dry season tem-peratures of the high sulfate springs appearsto be greater than the local flow springs;perhaps this is because the mixing ratio ofthe type 2 (high sulfate) springs graduallychanges during the dry season.

7.2.6. Discussion

Figures 7.9, 7.10 and 7.12 summarizemany of the important regional geologic,geomorphic, hydrologic, and hydro-chemical relationships. Erosion first ex-posed the structurally highest parts of theEl Abra and Tamasopo Formations prob-ably sometime in the early to middle Ter-tiary. The El Abra Formation is believed tohave a karstic permeability over the wholePlatform, and it forms a nearly continuousaquifer.

As previously stated, one of the princi-pal contentions of this thesis is that thechemistry and temperature characteristicsof the springs together with their elevationand discharge form a consistent and rea-sonable pattern. The properties listed abovehave been studied for many large and sev-eral smaller springs generally in the eastern

159


part of the Platform. For each of the largesprings, Figure 7.10 lists the elevation, theestimated average dry-season minimumflow, the highest measured sulfate concen-tration (mg/l), and the highest measuredtemperature, so that their geographic dis-tribution and properties can be seen. Ac-cording to the interpretations given here,the local flow systems along the eastern partof the Platform have Ca and HCO3 as theirmajor constituents and a temperature thatreflects the elevation (and climate) of theirrecharge area; examples of springs believedto have only local sources are Santa Clara,Arroyo Seco, Huichihuayán, San Juanito,Sabinas, and Canoas. Some large springscombine local sources with a regionalsource of water that is of calcium-magnesiumsulfate type and is much warmer. Thesemixed flow springs (type 2) exhibit tem-peratures a few degrees higher than nearbylocal flow springs (type 1). Furthermore,the greater the sulfate content, the greaterthe temperature difference; hence thegreater the percentage of regional water inthe mixed output. There is also the veryimportant relationship whereby the springsthat have the highest sulfate contents alsohave the greatest base flows and are thelowest elevation springs in their areas.The Coy is the principal outlet for regionalflow in the southern part of the region andhas the lowest elevation and the greatestbase flow of all the springs; the Mante isits northern counterpart. Based on eleva-tion and the water budget calculations (Sec-tion 4.3), Figure 7.10 shows some of thepossible general directions of local and re-gional ground-water flow. Springs sup-ported only by local flow systems rapidlydischarge flood recharge and retain rela-tively small quantities of water in gravitystorage; they are principally conduit flow

systems. The mixed source springs havegreater stability of flow and have muchlarger quantities of water in gravity stor-age; it is concluded they have conduit typelocal flow systems, but probably have alarge quantity of water stored in moreslowly moving regional flow systems thathave smaller in diameter conduits.

The general concept of the two watertypes, the local and regional ground-watersystems, and the geologic basis are shownin Figure 7.9. The sulfate content of thewater is believed to be derived mostly fromthe massive Guaxcamá Formation. Someof it might be derived from the thin sedi-ment veneer (including gypsum) in the RíoVerde valley, but sulfate water derived fromthe sediment probably would be deficientin Mg. It is hypothesized that water circu-lates deeply through a fractured dolomiteaquifer at the base of the El Abra Forma-tion, where it also contacts and dissolvesgypsum (or anhydrite). The intensely de-formed fold belt is an especially likely placeof gypsum solution, and the broad basinwest of the high sierras should also be fa-vorable.

The real system is undoubtedly morecomplicated than the simple model proposedin Figure 7.3. One possible deviation is thealternate source of sulfate mentioned above.Another deviation would occur if some typeB water leaves the dolomite-gypsum area,but remains in contact with the basal dolo-mite near the reef trend (Figure 7.9); thiscould explain the greater Mg/SO4 ratio inChoy water compared with the Coy. As pre-viously mentioned, in the high sierras gyp-sum in the interior of folds may be raisedto nearly the level of the valley bottoms,and hence some of the local and intermedi-ate flow systems there may contain signifi-cant amounts of sulfate. The Puente de Díos

spring at Tamasopo did have high sulfatecontent, but the Canoas (farther west) didnot. Another complication is loss duringbase flow of some high sulfate water fromthe Río Verde (and possibly the Río de losNaranjos) back into the subsurface. Prob-ably the most important factor not coveredin the model is regional movement of typeA water. Such movement is clearly possible(in fact probable), and it would remain anundetected third source in the two-sourcehydrochemical model developed here.

It has been previously shown that theCoy discharge is strongly dominated by itslocal source during flooding. Because theCoy is the lowest major spring in the re-gion (31 m) and because it does carry localsource water during floods, it is thereforelikely that part of the Coy’s base flow isderived from local sources also. This sug-gests that the chemistry of type B water isprobably nearer the model with the maxi-mum sulfate content than the model withthe minimum (taken to be identical to thelargest value measured at the Coy spring),and that the regional flux of type B wateris nearer the calculated minimum than thecalculated maximum. Temperature consid-erations also support this conclusion. Theregional sources of each type 2 spring haveslightly different concentrations of the ma-jor chemical species, as demonstrated bytheir chemical ratios and chemical dilutionpaths. This might be explained by some ofthe complicating factors discussed in thepreceding paragraph, particularly variationsof flow path, or by differences in residencetime.

Another important point to note in Fig-ure 7.10 is the geographic distribution ofthe high sulfate springs. They encompassall of the major eastern margin springs ex-cept those at the extreme north and south

Table 7.10Calculated temperatures for type B (regional) water

Model used CalculatedSpring T2 (°C) TA (°C) for QA′ and QB′ TB (°C) Source of TA

Choy 26.4 24.9 15 32.0 extrapolation of Choy line in Figure 5.9

24.0 15 35.4 temperature of Arroyo Seco and Santa Clara springs and cave lakes

24.9 20 28.8 extrapolation of Choy line in Figure 5.9

24.0 20 30.2 Santa Clara and Arroyo Seco springs and cave lakes

Coy 24.6 21.4 22 29.5 extrapolation of Coy line in Figure 5.9

20.5 22 30.9 slightly warmer than Huichihuayán and San Juanito springs

Mante 26.5 24.0 15 34.3 Santa Clara and Arroyo Seco springs and cave lakes

Frío 22.6 21.0 15 36.7 Sabinas spring

160


ends. There is a sharp break in the hydro-chemistry between the Pimienta and the SanJuanito–Huichihuayán springs; the separa-tion of flow systems is probably caused bya structural discontinuity between theXilitla and Aquismón areas (Figures 7.10and 7.12). This constitutes the only bound-ary between flow systems known with anydegree of geographic precision. The lattertwo springs are the most important in theintensely karstic Xilitla area, but have sur-prisingly low minimum flows comparedwith the Coy, Choy, and Mante, for ex-ample. At the northern end, the change ismore gradational. The Frío springs have avery large base and total flow, but onlyabout 10% of the base flow is supplied bytype B sources. This great spring group isthe principal outlet for Sierra de Guatemalaground water and probably for nearbyranges to the west. (The Frío springs mayreceive intermediate or regional flow fromwestern ranges.) This suggests that there islittle or no gypsum in the subsurface ofthose ranges.

A few final points are:

1. Despite the potentiometric map givenin Figure 7.11, the permeability develop-ment (i.e., conduits) may be anisotropic.Geologic structure and fractures may oftenfavor strike-line flow, and in the high,folded karst area east of Cd. del Maiz, itappears that water is transmitted along thestrike to the north and the south rather thaneastward to the lower Micos basin.

2. The interpretation given in Section7.1.3 that the regional source (type B) wa-ter to the Coy is diverted during floodingto the Pimienta and perhaps to other springsimplies a well-integrated system of conduitswithin the limestone aquifer. This wouldhave serious consequences to the successof a proposed artificial surface reservoirthat would flood a large area of the Vallesvalley west of El Pujal, including the Coyspring.

3. A minimum value for the average gra-dient of the regional potentiometric surfacemay be determined between the MediaLuna and the Coy springs. The value is(1000–31)m/115km = 8.43 m/km, or0.0084 dimensionless. This is a very highgradient for large karst-conduit systems andindicates that the conduits are often con-stricted or generally poorly developed.

4. Some support for the regional ground-water flow system described is obtainedfrom a report by Lesser Jones (1967) ondeep water wells in the Monterrey area,which obtain large quantities of fresh water

from massive confined Cretaceous lime-stones hundreds of meters below sea level.The probable recharge areas are located 50to 100 km to the west. Water levels in thewells recover very rapidly after pumpingceases, respond quickly to rainfall on therecharge areas, and exhibit seasonal fluc-tuations of 20 to 80 m.

7.3. Hydrology of theSierra de El Abra

One of the principal goals of the thesiswas to collect data to develop a qualitativeand quantitative understanding of the hy-drology of the Sierra de El Abra. Some ofthe information and analysis has been givenin Chapter 4, particularly Sections 4.2 and4.4. This section will utilize previouslydeveloped and other data and interpreta-tions to summarize the presently knownhydrology of the range.

7.3.1. Water budget and distribution ofdischarge of the Sierra de El Abra

Recharge into the Sierra de El Abra oc-curs wherever the El Abra limestone is ex-posed, including numerous places along thewestern margin where streams flowing fromshale catchments are pirated underground;the piracies usually occur a short distancebeyond the point where the streams passonto the limestone. Surface runoff from theEl Abra limestone does not occur exceptalong the steep eastern escarpment andpossibly at a few other minor localities.Thus, precipitation on the range may ef-fectively be divided into evapotranspirationlosses and ground-water recharge.

Discharge from the El Abra aquifer isfrom springs. The two most important arethe Mante and the Choy; they are first mag-nitude springs. Both maintain a large baseflow, estimated to average about 8 and 2.5m3/sec, respectively, late in the dry season(from inspection of S.R.H. data). Subtract-ing the surface runoff included in the gaug-ing station measurements, the long-termaverage discharges of these springs are cal-culated to be about 11.9 and 5.41 m3/sec.However, there are also a number of smallersprings that are significant to the hydrol-ogy. These latter springs have a very smallbase flow; many dry up completely duringthe dry season. The third most importantspring in the range is probably the Naci-miento del Río Santa Clara (actually oneperennial and two temporary springs). Al-though its dry season base flow is onlyabout 0.3 m3/sec, wet season floods are

sufficient to produce a significant yearlytotal discharge (Figure 4.10). The averagedischarge at the gauging station for 1972was 2.53 m3/ sec and for 1973 was 5.09m3/ sec (calculated from S.R.H. stage anddischarge measurements, Section 4.2). Us-ing rainfall data at El Refugio, an S.R.H.climatologic station 12 km northwest of thegauging station, a minimum contributingbasin area was calculated for the majorfloods of 1972 and 1973, assuming thatstorm runoff equalled the volume of pre-cipitation. The pre-flood base flow com-ponent was subtracted from the total flow,but changes in aquifer storage were onlypartially taken into account. The values ofthe minimum contributing total area rangedfrom 82 to 123 km2. From air photos, themaximum surface basin included within thecatchment was measured to be 16.5 km2; itmay be less because there may be artificialdiversions of one of the principal streamsout of the basin. Thus, the infiltration wa-ter into at least 85 km2 of the El Abra lime-stone is required to provide the measuredflow of the Santa Clara springs. These cal-culations demonstrate the extremely dy-namic discharge pattern of the springs inthe Sierra de El Abra and other places inthe region that have small base flows, andalso their capacity to yield important yearlytotal volumes of water. Based on the sizeof their drainage channels, it is suggestedthat many of the other small El Abra springssuch as the Tantoán, San Rafael de losCastros, and Arroyo Seco on Rancho Peñoncarry moderately large floods and have sig-nificant total discharges, although less thanthe Santa Clara springs.

It is desirable to determine the waterbudget for the Sierra de El Abra. Unfortu-nately, there are several problems, includ-ing that there is a complete lack ofraingauge stations on the range (Figure4.3), not all the springs are monitored forrunoff, no drainage divides for the springshave been established, and the basin orcatchment area is not clearly defined, be-cause the El Abra limestone of the Sierrade El Abra connects directly at its north endwith the Sierra de Guatemala and via thesubsurface with other exposures in rangesto the west. Enjalbert (1964), on the basisof a brief field trip in the region and appar-ently having no quantitative data on theregional hydrology, hypothesized that theprincipal source area of the Mante and Fríosprings is a calcareous plateau or moun-tain ranges 30 to 40 km to the west. Heeven gave a cross section showing ground-water flow through caves under the Antiguo

161


Morelos valley to the Mante spring.Mitchell et al. (1977) have argued that allof the discharge from the Sierra de El Abrais derived locally, based on a crude waterbudget and the response of the Choy andMante springs to storms.

An improved water budget calculationis given here. The approximate local sourcearea involved is determined in Table 7.11.For the period 1961–68, the discharge ofthe Mante was 11.77 m3/sec and of theChoy 5.07 m3/sec. Of the two years forwhich data are avilable for the Santa Clarasprings, 1973 was exceptionally wet, but1972 was about average or slightly abovefor the nearby area. Hence, the long termaverage discharge of the Santa Clarasprings should be about 2.5 m3/sec, and thisvalue is taken for 1961–1968.

The total discharge of all the other smallEl Abra springs, including losses from thesouthern end of the range into the RíoTampaón and from the northern end intothe Río Comandante, is assigned a valueequal to that of the Santa Clara springs; thisis likely within a factor of 2 of the realvalue. Any direct runoff from the east faceis not allowed for. Hence, the average dis-charge from the Sierra de El Abra was about21.84 m3/sec. Rainfall in the area is about1000 mm per year near Mante, 1100 to1250 at the ends of the range and along thewestern side, and might be on the order of1300 to 1400 mm on the high central por-tion of the range. Using a basin average of1250 mm per year, the potential runoff is35.89 m3/sec. If all of the observed (or es-timated) runoff came from the basin givenin Table 7.11, runoff would be equal to61%, and evapotranspiration would be only39% of the precipitation. A high amount ofinfiltration into the El Abra aquifer wouldbe expected, because it has spotty soil coverover a highly permeable karst surface; butthe 166 km2 that yield surface runoff(swallet cave catchments plus surface ar-eas included at gauging stations) should beless efficient at producing runoff, like theValles valley (Section 4.3.3), because ofsoil infiltration and other losses. It wasobserved in the field on some occasions thatseveral days of showers did not cause thearroyos of the swallet caves to run. Fur-thermore, dry season showers of as muchas 30 mm seldom cause a noticeable re-sponse at the springs. The percentage ofrunoff calculated above is a very high value,and is thought to be too high.

Using figures from above, the unitarydischarge of the El Abra range is calculatedto be UEl Abra Range = (21.84 m3/sec)/(906 km2)

= 0.0241 m3/sec/km2. This is three timesthe value calculated for the Valles valley(taken to be about 0.008 in a possible rangeof 0.00604 to 0.00955 m3/sec/km2 (seeTable 4.12), and 25% greater than the0.0193 m3/sec/km2 value determined for thehigh rainfall karstic belt in the eastern partof the Valles–San Luis Potosí region. Thislatter area was termed the “reduced region”in Section 4.3.3, and its estimated averageyearly rainfall was 1500 mm. Ergo, the cal-culated hydrology of the Sierra de El Abra,the analysis of the regional hydrology (Sec-tion 7.2, particularly 7.2.4), and thehydrochemical model (Section 7.1) all sup-port the existence of a regional componentof flow in the discharge from the range. Itis thought that the smaller springs are localflow systems, whereas the Choy and Mantesprings have both local and regionalsources. If the minimum and maximumcomponents of regional flow to the Manteand Choy as determined in Table 7.4 aresubtracted from the El Abra discharge, thenew unitary discharges are 0.0201 and0.0146 m3/sec/km2, respectively. Runofffrom the local source areas would then beequal to 51% or 37% of the precipitation.These new figures delimit the range of pos-sible values for the El Abra range hydrol-ogy as determined by the modeling in thisthesis. Based on the hydrochemical model,the regional component of flow to the Choyand Mante represents 17% to 39% of thetotal discharge from the El Abra range,while the local component represents 83%to 61%. Most of the regional flow is fo-cused on the Mante spring, where it com-prises 25 to 60% of the total flow. If thereis regional ground-water flow of type Awater (as discussed in Section 7.2.6) to theMante and Choy springs, then the amountof local discharge would be less still, andthe calculated values above would have tobe further amended. Thus, Enjalbert (1964)was correct when he guessed that there is

regional flow to the Mante and Frío springs,but according to the hydrochemical modeldeveloped here, he may have overestimatedthe amount. Reasonable changes in thequantity of runoff and of precipitation usedin the calculations above will cause onlyabout a 10% and a 15% change in the cal-culated unit runoff and percentage of run-off, respectively.

7.3.2. Analysis and discussion of theground-water systems

This section will examine the data con-cerning El Abra ground water, present mod-els in an attempt to explain the observedhydrology, and indicate some of the prob-lems and uncertainties that remain.

A significant part of the field work wasdevoted to mapping caves, observation ofthe relationship of caves to the hydrology,and mapping water levels to determine hy-draulic gradients, the potentiometric sur-face if possible, and whether the swalletcave siphons were perched or at the watertable. The El Abra limestone itself has es-sentially no capacity to transmit water. Allflow occurs along fractures and beddingplanes that have been enlarged by karstsolution processes. The amount of solu-tional enlargment is highly variable. Fieldobservations of the aquifer and the ex-tremely dynamic response of the springs torainfall demonstrate that recharge rapidlyinfiltrates the aquifer and is integrated intolarge caves (conduit flow systems) that arecommonly 5 to 10 m (or more) in diam-eter. Flood waters should be rapidly dis-charged from the cavern systems. Evidencethat supports the idea of high flow veloci-ties includes emergence of muddy water atEl Choy within a few days after heavy rainscause runoff to the swallet caves, under-saturation with respect to calcite of floodwaters at all springs sampled and at localflow springs such as Santa Clara during the

Table 7.11Drainage area of the Sierra de El Abra “basin”

km2

Outcrop area of El Abra limestone in El Abra range* 740

Surface catchments included at gauging stations 90.4(Mante 69 plus Choy 4.9 plus Santa Clara 16.5)

Basins of western margin swallet caves (Table 3.1) 75.6

Total area 906

* Mitchell et al. (1977) give 718 km2 for the El Abra range up tothe Río Camandante; an additional 22 km2 is allowed for theextension of the range to Puerto Chamilito.

162


dry season, and the tentatively positivespore tracing test from Sótano de Japonésto El Choy. Furthermore, the Santa Claraspring does not have a large quantity ofwater in gravity storage compared with thevolume of its flood discharge; this impliesa short residence time. The discharge fromthe Choy and Mante springs is estimatedto be 82% of the annual flow and 97% ofthe dry season base flow from the El Abrarange (including surface catchments as dis-cussed in Section 7.3.1). Hence, the aqui-fer would appear to be strongly integratedinto two groundwater systems that have avery large amount of water in gravity stor-age. Water discharged during the dry sea-son obviously has been in the aquifer a

minimum of several months—muchlonger than at least part of the flood wa-ters, but still a very short time for groundwater.

Discharge from the Sierra de El Abrausually has some component of lift or riseto it, i.e., the springs are openings at thetop of flooded portions of caves. None ofthe springs are thought to be perched. It isbelieved that the El Abra has a phreatic zonein which all secondary openings are filledwith water. Circulation within the phreaticzone likely reaches several hundred metersbelow regional base level and might extenddownward to the base of the aquifer. Largefractures provide the routes for verticalmovement of water in the aquifer and may

play an important role in horizontal move-ment to the two major springs in this greatlyelongated range (e.g., along longitudinalfractures in the reef zone). The depth ofcirculation is discussed further in Section7.4.

The problems of the phreatic zone in akarst aquifer such as the El Abra are thehydraulics and dynamics of the aquifer, thenature of the potentiometric surface, andthe distribution of storage. It is anticipatedthat the potentiometric surface will behighly irregular, because the range of cross-sectional area of the conduits is extreme.Water in narrow fissures requires a muchgreater hydraulic gradient for a significantflow rate than water in large caves. If onlylarge conduits are considered, the poten-tiometric surface might be reasonablysmooth; perhaps more significant is thehydraulic gradient of each of the large con-duits. Figure 7.13 is a plot of the elevationof most of the more important cave lakesin the southern El Abra range against theirradial distance from the Choy spring (Table7.12 lists cave depths and terminal si-phons). The caves are drawn vertically in-stead of in true profile, but this will notintroduce a large error. Obviously, most ofthe lakes are perched, usually on mud. Asdiscussed in Section 4.4.2, the elevation,the water depth, and most importantly, thehydrograph of the lake in Sótano de Soyateare indicative of a genuine water table lake.However, data through the dry season havenot been obtained, so it is possible that af-ter some drawdown the lake will becomeisolated by a rock barrier in the conduit,causing stabilization of the surface. Astraight line has been drawn from El Choythrough the early January 1972 elevationof the Soyate lake as an approximation ofthe potentiometric profile for large El Abraflow systems. The gradient of this profileis 2.6 m/km, which is greater than a pre-liminary estimate of 1.0 m/km given by Fishand Ford (1973). The geomorphology ofthe swallet caves suggests that all of theirterminal siphons are perched lakes, stillabove the level of permanent water satura-tion in each conduit. Open air passage ispredicted to be beyond their siphons, eventhough several of the large swallets havebeen explored to depths that approach theprojected fluid potential line. Explorationin Sótano del Arroyo was halted by 4% CO2in the air, rather than by a sump. Accord-ing to the profile drawn, it should havereached the phreatic zone, but the cavesurvey, the altimeter reading, or the poten-tiometric profile may be in error. The

Table 7.12Cave survey data and elevations for water-level profile

Surveyed cave Entrance4 Deepest knownCave depth (m) elevation (m) free air point (m)

Choy resurgence pool — — 35

Curva (Ferrocarril) 19.1 131.5 112.5

Palma Seca 52.51? 151.5 99?

Piedras 46.51 145 98.5

Jos 84.5 176 91.5

Soyate 234. 293 59

Sótano de Montecillos 81.52 157.5 76(Pichojumo)

Sotanito de Montecillos 91.52 190 98.5

Tinaja 82. 165.5 83.5

Sabinos 95.5 231 135.5

Arroyo 134.5 192 57.5

Roca (42.)1 240.5 (198)

Tigre 160.0 245.5 85.5

Japonés 139.5 243 103.5

Matapalma 86.53 (210*) (123.5)

Yerbaniz 95.1 241.5 146.5

Prieto (Los Cuates) 32.51 61.5 29

Chica 18.1 49 31

Venadito >183.1 312 <129

Pachón 8.1 210.5 202.5

1. Mitchell et al., 1977.

2. Neal Morris, personal communication.

3. M. Walsh, 1972.

4. The elevations used here, except for El Choy, are those given byMitchell et al., 1977.

* Estimate.

163


hydrographs from Sótano de Jos revealthe character of these perched lakes. Theirsurface levels remain approximately staticexcept when they are in flood.

It was previously stated that the dry sea-son flow from the El Abra range is concen-trated almost exclusively at the Choy andthe Mante springs, and that this implies avery strong integration into two flow sys-tems. However, in previous sections of thischapter it has been shown that at least partof the base flow is derived from regionalsources. Furthermore, all of the springsyield moderate to very large flood pulses.These facts lead to three possible modelsto account for the observed flow behavior:(1) There are two distinct types of flow sys-tem (cave) development and storage in theSierra de El Abra aquifer—one for the twohigh base and total flow springs, the otherfor the smaller springs. The smaller springshave drainage systems that are not inte-grated with the two main systems. A largepart of the base flow comes from El Abrarange storage. (2) The aquifer is almostcompletely integrated into two systems, and

the smaller springs function as spilloversfor high water levels that occur during thewet season. A considerable part of the baseflow comes from El Abra range storage. (3)The dry season flow of the Mante and Choysprings is derived almost completely fromregional sources to the west, and the localflow system behind each spring is essen-tially isolated and supplies the flood wa-ters.

The first model seems unlikely, becauseno adequate causal factors have been ob-served which could account for two dis-tinct types of flow-system development.Basically, the model requires one type tobe capable of vast storage and the other typeto be relatively incapable of storage. Infil-tration into the range should be greatest inthe high central portion of the range, yetthe two largest springs occur near the ends.Water stored in narrow bedding plane orfissure openings and in larger conduitsshould be relatively uniformly distributedthroughout the range, yet the smallersprings, except the Santa Clara (and thethermal springs), dry up completely some-

time during the dry season. The capacityfor three dimensional circulation within theaquifer is believed sufficient that flowshould continue even if some horizontalconduits contain bedrock barriers to move-ment during low water stages; shale barrierscan not be called upon as an explanation.If the systems were truly isolated from oneanother, some base flow would be expectedfrom all the springs in rough proportion totheir catchment areas. Ergo, the first modelis rejected.

The third model implies that there is verylittle water in gravitational storage withinthe Sierra de El Abra. It suggests that eachlocal ground-water flow system is isolated(not connected with other systems) and theyall behave in a similar manner, i.e., as dy-namic wet-season flood-water systems. Thesmaller springs would then be the true in-dicator of the character of the Sierra de ElAbra aquifer. The available data do notadquately test this model—for example,water tracing tests for dispersion and resi-dence time, and pump tests. According tothe hydrochemical calculations (section

Figure 7.13. Comparison of elevation of cave lakes and deepest point known in caves with radial distance to andelevation of El Choy rise pool.

164


7.1.3), regional sources provide at least24% of the combined dry season flow ofthe Mante and Choy (using QBs for May1973 from Table 7.4), and perhaps as muchas 56%. Because it is possible that theremay also be a regional source of shallowcirculating type A ground water, the latterfigure might not be the true maximum con-tribution from regional sources. If a dryseason average of about 8.0 and 2.6 m3/secfor the Mante and Choy, respectively, issubtracted from the total runoff as deter-mined in the previous section, the new uni-tary discharge would be 0.0124 m3/sec/km2,and the local runoff would be 31% of thelocal rainfall. These figures cannot be dis-missed as totally unreasonable, and if cor-rect about one half of the El Abra rangedischarge would be derived from outsidesources. However, data given by Faulkner(1976) show that the great Silver Springsground-water basin of Florida has an aver-age recharge of about 31% of the total rain-fall of 1350 mm (53.2 inches), including2% lost to pumpage. Because the soil coveron the El Abra range is spotty, a slightlyhigher percentage may be expected. Fur-thermore, it seems likely that there is a sig-nificant amount of water in storage withinthe El Abra range that is gradually dis-charged. Evidence for storage includeswater levels in a number of observationwells drilled by the S.R.H. in the southernpart of the range (S.R.H., unpublisheddata), some vadose zone water discussedin Section 4.4.1, and, most importantly, thelarge water-table lake in Soyate (there areprobably a great many in the El Abra). Thewell data suggest that considerable watermay be stored in small solution channelsand that variable hydraulic gradients exist.Finally, the stability of dry season dischargeof the Coy spring has previously been at-tributed to its large component of regionalflow. For the period 1960–1971, the coef-ficient of variation of the yearly minimumdischarge of the Choy is three times greaterthan for the Coy (0.220 versus 0.068, re-spectively). The lesser stability of annualminimum discharge of the Choy suggestsgreater dependence on its local source.

In model number 2, most of the aquiferis integrated into the two main ground-water flow systems. Extremely dynamicground-water conditions occur during thewet season, when storms cause large tem-porary increases in water levels and replen-ish ground-water storage. All of the springsrespond quickly to the increased hydraulicgradients, and more than half of the re-charge water is rapidly discharged. In this

model, the smaller El Abra springs act asoverflow systems for high aquifer waterlevels and develop specifically because oftemporary high local hydraulic gradients.They are left high and dry during the dryseason, as water levels decline in the largeconduits that drain the aquifer and as wa-ter that has been retarded by flow in smallsolution openings seeks larger conduits atlower elevations. The basic premise is thatthere is some degree of physical connec-tion between the large and small springs.Figure 7.14 shows a hypothesized N-S pro-file of dry season ground-water conditionsand surface base levels along the east faceof the Sierra de El Abra, based on nine base-level points (Table 7.13), including theSoyate lake, and on model 2. Springs areshown as open circles, and where their el-evation is unknown, a ? is given. A num-ber of caves are projected eastward into thesection to assist placement of upper limitson the profile; remember, however, thatthey provide only apparent gradients in thesection, but the error is small except forcaves directly west of El Choy. Both Ar-royo and Venadito continue in the vadosezone, and a ? indicates they will attain someunknown greater depth before reaching asiphon. Using their altimeter survey of caveentrances and cave surveys conducted byother cavers and many from this thesiswork, Mitchell et al. (1977) have given asomewhat similar diagram to Figure 7.14for that part of the El Abra range betweenthe Choy and the Mante springs and havegiven a discussion of the hydrology, par-ticularly as it relates to the distribution of

blind fish in the area.Many aspects and problems of the El

Abra range hydrology may be seen in Fig-ure 7.14, keeping in mind the breadth oflimestone outcrop and location of swalletcaves shown in Figure 3.1. Among theproblems are directions of local flow, lo-cation of drainage divides, variable hydrau-lic gradients, the potentiometric surface,and flow system separation or interconnec-tion. In model 2, the regional componentof flow is taken to be within the rangecalculated in Table 7.4. After the regionalcomponent and an estimate for the surfacerunoff included at the gauging stations aresubtracted, the Mante (between 8 and 4 m3/sec) still has the largest discharge from thelocal aquifer, the Choy (4.4 to 3.5 m3/sec)the second largest, and the Santa Clara(2.25 m3/sec) is thought to have the thirdlargest discharge. As pointed out byMitchell et al. (1977), the fact that theMante is the largest poses some interestingproblems, though they assumed or believedthat all of the discharge measured at theMante gauging station was derived solelyfrom the El Abra range. It lies in the nar-row northern part of the range at an eleva-tion of about 80 m, whereas the Choy liesat the southern end of the broad centralportion of the range at an elevation of 35m. The Mante spring obviously taps a con-siderable groundwater body above 80 melevation. Thus, a ground-water dividemust occur that is closer to the Choy thanto the Mante spring (Figure 7.14). Basedon their simple water-budget calculationsand on the assumption that siphon lakes at

Table 7.13Base-level points of the Sierra de El Abra

Base-level point Elevation (m) Notes

Choy resurgence pool 35 altimeter; rise pool rather than outletelevation is used

Taninul Sulfur Pool 39 artificially raised 1 to 2 m; altimeter andS.R.H. topographic data

Río Tampaón 28 S.R.H. gauging station data

Santa Clara Spring (75–100) dry season water level at gauging stationis 70.0–70.2 m

Mante 80 topo map indicates just under 80 m; Mitchellet al. (1977) give 80.5 m; was 1 to 3 m lowerbefore construction of reservoir

Río Comandante 75 S.R.H. gauging station data

Poza Azul (Frío springs) (90) estimated from S.R.H. data

Taninul Limestone Pool 37.5 altimeter

Soyate lake 59 altimeter and cave survey

165

AMC

S Bulletin 14 — C

hapter 7

Figure 7.14. Hypothesized N-S dry-season ground-water profile in the Sierra de El Abra. Base level points are listed in Table 7.13.

166


about the same elevation in three separatepassages in Sótano de Yerbaniz indicate the“water table” of the Mante drainage,Mitchell et al. (1977) predicted that thedivide occurs between Tigre and Japonés.However, the depth attained in Japonés (bythis study, where exploration was termi-nated by what is believed to be a perchedsiphon) proves that their water-level pro-file must be lowered substantially andstrongly indicates that the conformity ofsiphon elevations in Yerbaniz is a coinci-dence or the result of strong structuralcontrol. It is likely that the Yerbaniz-Matapalma-Japonés group of caves drainto the same spring, probably the Choy, butperhaps to one of the smaller springs alongthe east face. The location of the divide asdetermined by water-budget techniques willdepend on the magnitude of regional flowand the flows assigned to the smaller

springs.Other drainage divides must also exist.

Probably a significant part of the rechargesouth of El Choy (35 m elevation) flows tothe bed of the Río Tampaón (28 m eleva-tion) and to wet season springs along theeast face (Figure 7.14). Two caves near theRío Tampaón have free-air lakes severalmeters below the level of the Choy and thuscould not drain to it (Table 7.13). Anothermajor divide evidently is located some-where between the Frío and the Mantesprings. An unusual dry season flood oc-curred in February 1957 at the Frío springs.Monthly rainfall data at a number of sta-tions in the area (see location map in Fig-ure 7.15) show that five to ten times theaverage February precipitation fell duringthis month in the Sierra de Guatemala area,whereas the northern Sierra de El Abra areareceived only a little more than the average

and the upper Río Guayalejo basin (Jau-mave area) received very little precipita-tion. Apparently, most of the precipitationoccurred during a very localized storm onFebruary 4 and 5, causing the strong re-sponse at the Frío springs (Figure 7.15).The fact that there was no response at LaEncantada and virtually none at Mantemeans that, for all practical purposes, thelocal components of their ground-waterflow systems are hydrologically separated,although it is possible that small connec-tions exist. The very small increase in flowat the Mante gauging station on subsequentdays was most likely caused by local show-ers, but might have come from the SierraTamalave or Sierra de Guatemala.

Unfortunately, gauging stations on RíoComandante and Río Sabinas (Figure 4.2)were not operative in 1957. It would alsobe interesting to observe the response ofthe smaller springs between the Mante andthe Frío springs during isolated events suchas the one above.

Studies of the Mante spring have beenlimited in this thesis, and more are required.Figure 4.9 shows a common discharge pat-tern recorded at the gauging station, al-though other floods have lesser flows thanthe 1951 event. After a few small pulses ondays having showers, a large rainfall-runoffevent occurs. In such circumstances, runofffrom the 69 km2 surface catchment includedwith the gauging station measurementswould provide an important part, perhapsas much as 50%, of the large pulse on thefirst day. However, it is clear that the springalso responds rapidly. Following the ini-tial peak, there commonly is a lengthy pe-riod of high, gradually declining dischargethat is derived predominantly from thespring. With the support of additionalshowers and storms, the period of sustainedhigh to moderate flow may last for severalmonths. (Compare the August-September-October rainfall in Table 4.5 with the Mantehydrographs in Figure 4.9.)

Another Mante hydrograph pattern isshown in Figure 7.16, where it may be com-pared with hydrographs of the Santa Claraspring and the Río Comandante (see Fig-ure 4.2 for the relationships of the basins).Also shown is the rainfall at El Refugio,which is located against the western sideof the El Abra range about half way be-tween the Mante and the Santa Clarasprings. For this period, the data avilablewere the mean daily flows of Río Mante,but only occasional discharge measure-ments of the two canals. (Generally theflow of the canals totals 8 to 13 m3/sec, but

Figure 7.15. Hydrographs in February 1957 of Frío springs, Mante, and RíoGuayalejo showing an unusual isolated dry-season flood.

167


during major flood peaks the flow is oftenreduced to nil.) The smooth curve repre-sents the approximate general shape of thehydrograph, from which only small peaksand troughs would be missed. Examinationof the Santa Clara hydrograph shows thatthe mean daily flows would give a reason-ably good representation of its rainfall-flood events. Clearly, there is a significantdifference between the flow pattern of theMante and the other two. Large rainfallson the supposed recharge area of the Mantespring in the El Abra range produced sharpresponses and well-defined events at theSanta Clara spring, but at the Mante sta-tion there occurred only a very broad floodwave which reached its peak more than aweek after the heavy rains. Apparently,local surface runoff did not have a largeeffect on this Mante hydrograph. The dis-charge pattern shown here may have beencaused by the time delays and dispersionof wave form that would be expected of anintermediate or regional flow system pulseof relatively shallow circulating type Aground water from potential source areassuch as the northern Sierra Tamalave andother western ranges, or it may possiblyhave been caused by storms in the south-central part of the El Abra range (comparewith the Choy hydrograph and rainfall forthis period shown in Figure 4.7).

Thus, the true source areas of the Mantespring and to a certain extent the springhydrograph (as opposed to that measuredat the gauging station) remain uncertain.Contrary to Mitchell et al. (1977), the Choyspring probably responds more rapidly thanthe Mante spring and is more peaked. Theircontention that the Choy response lags afew days behind rainfall events was basedon only one hydrograph for which they in-correctly reproduced S.R.H. discharge andrainfall data. About 70 to 80% of the Mantespring discharge may be considered asbaseflow, whereas about 50% of the Choydischarge is produced by floods, and nearlyall the discharge of the smaller springs isflood flow. In model 2, it is hypothesizedthat the Santa Clara and others such as theTantoán rapidly drain away much of thelocal flood waters in the aquifer, so that theMante spring response is lessened. However,this explanation does not seem satisfactoryfor the hydrographs shown in Figure 7.16.It seems problematical to suggest that theprincipal recharge zone of the Mante springlies south of the Santa Clara recharge zone.

Because nearly all of the caves havingsiphons thus far investigated in the Sierrade El Abra are believed to be perched, it

has not been possible to map the top of thewater-filled zone of the aquifer. The onepotentially usable point is Sótano deSoyate, which gives a hydraulic gradientof 2.6 m/km. This compares with values of1.0 m/km average for Florida (2.3 m/kmmaximum) and the extraordinarily low gra-dient of 0.02 m/km for Yucatan (valuesfrom Back and Hanshaw, 1970). Palmer (inMiotke and Palmer, 1972) states that someof the major passages in Floyd CollinsCrystal Cave have gradients of less than1 m/km and believes that they indicate thegradient of the water table. Considering thesize of many of the passages observed inthe Sierra de El Abra, the value calculatedabove for the Soyate lake seems ratherlarge. A minimum gradient between theMante (80 m) and the Choy (35 m) of 0.55m/km may be determined by the differencein elevation and separation of the twosprings. Between the normally stagnantTaninul limestone pool (see Section 5.3)and the Choy there is a gradient of 0.46 m/km.

Several anomalies of fluid potential andflow exist in the El Abra range. Only 90 mfrom the Taninul limestone pool and 1.5 mhigher lies the permanently flowing TaninulSulfur Pool (thermal). However, the Sul-fur Pool probably was raised artificiallywhen it was made into a spa. Nevertheless,the flow systems are not connected. The twoother thermal springs observed have verysmall flows, but are likely permanent. TheRío Tampaón (28 m) and the Río Coman-dante (75 m at the east end of CañónServilleta) provide the lowest possibleground-water discharge at the south andnorth ends of the range, respectively (Fig-ure 7.14). However, there is no dry seasonflow into the latter and probably very littleinto the former (no major boils or springshave been reported). Furthermore, as pre-viously stated, in model 2, the Mante (80m) receives more El Abra ground-waterthan the Choy (35 m). If correct, this hasboth hydrologic and geomorphic implica-tions that should be carefully evaluated infuture studies. Hydrologically, the aboveanomalies suggest that it takes a very longtime for ground-water conduit flow systemsto adjust to changes in base level. The RíoComandante and Río Tampaón clearly havethe elevation advantage, but the Mante andChoy springs have the advantages of well-developed hydraulically efficient conduitsystems and of position in the very elon-gated aquifer. From the morphology of theChoy cave (Section 6.2.1), it is clear thatthe Choy has been a spring for a long time.

Presently, it is probably expanding its drain-age area northward at the expense of theMante spring. It also seems likely that af-ter the initial exposure of the high centralportion of the range, the principal outletshave slowly been migrating toward the endsof the range, where the Río Comandanteand Río Tampaón were established long ago(see Figures 7.12 and 7.16, and Section3.3).

Wet season storms cause very dynamicconditions within the aquifer. In Section4.4.2, some data and discussion of thesetransient situations are given. The Soyatelake hydrographs (stage) correlated verywell with the Choy spring. Some minordiscrepancies indicate that localized eventsoccur which are not felt throughout theaquifer. Fluctuations well in excess of 10 mhappen frequently. A maximum rise ofabout 30 m for extremely large floods issuggested. This upper bound is supportedby the lack of mud on stalagmites and wallsabove that level in Soyate and by the pres-ence of only two mud rings (layers) in a10 cm diameter stalagmite 34 m above theJanuary 1972 water level (or 23 m abovethe stage recorder board). If this conclu-sion is correct, the mud rings indicate floodevents that only happen once in thousandsof years. A rise of 30 m would increase thefluid potential by only 128%, to 5.75 m/km.It is anticipated that water level fluctuationsin the aquifer will vary greatly from oneconduit to the next, depending on the ca-pacity of each to transmit the quantity ofwater supplied to it. The swallet cavescapture runoff from large areas of imper-meable rocks; hence they have very con-centrated inputs to cave systems. The recordedpulses at Sótano de Jos are interpreted tobe passing flood waves in a perched con-duit. However, on some occasions waterrose above the level of the recorder, andMitchell et al. (1977) report an observa-tion of flood waters filling the entrance sinkof Sótano del Arroyo to about half way upthe entrance drop (water depth would prob-ably be 5 to 8 m above the floor of the en-trance passage). It is not certain whetherwater had backed up all the way from thephreatic zone or whether water had backedup behind passage constrictions or ob-stacles. In Jos, there are also many obstaclesand constrictions that retard flow, and floodpulses coming from the surface basin areso sharp or peaked that the conduit maylocally flood to the roof in this youngswallet cave. In Arroyo, there are manyexamples of piles of breakdown, flowstoneaccumulation, and reduction of bedrock

168


passage cross section which would affectflow (Section 6.4.5.2). At the Hotel Taninul,the springs behave in a very peculiar man-ner when it rains (see Section 5.3 for moredetails). The Taninul Sulfur Pool has twosources of water—one is a deep thermal-sulfurous source, the other is the El Abrarange. When it rains, cool non-sulfurousEl Abra waters dominate (Figure 5.11). Partof the flood water flows overland into thenormally stagnant limestone pool, whichacts as a sink for some time (perhaps days)before the flow reverses and it becomes aspring, with a flow reaching at least 0.1 m3/sec. The uncorrelated discharge pattern andanamolous flow in relation to the elevationof the springs indicates that the flow sys-tems are discrete and that water levels inthe aquifer are very erratic under dynamicconditions.

In light of the information and interpre-tations given, a fourth model is suggestedthat will have to be considered along with

the others in future work to better under-stand the hydrology. The fourth model sug-gests that nearly all of the base flow and asignificant part of the flood flow of theMante spring comes from regional ground-water sources outside the El Abra range andthat the Choy has the largest total dischargeof Sierra de El Abra ground water and isthe principal base flow spring of the ElAbra. In model 4, the Mante spring is con-sidered to be the great regional flow outletfor the northern part of the Valles–San LuisPotosí Platform. The regional flow includesnot only the high sulfate water of thehydrochemical model, but also some re-gional flow of type A water that may havea variable discharge. Flood waters at theMante would then be a mixture of localwater, regional type A water from sourcessuch as the Sierra Tamalave, and the re-gional high-sulfate water. The model im-plies that the northern boundary of the Choydrainage area would extend considerably

farther northward than in model 2. Further-more, the potentiometric profile and thesmaller springs between the Choy and theMante could be lowered below that shownin Figure 7.14, possibly lower than theMante spring. Thus, the elevations of thesmaller springs, particularly the Santa Clara,may provide an important test of the models.

In this section, hydrologic data on theEl Abra range have been summarized andused to develop and to test four models thatattempt to explain the observed and inter-preted character of the aquifer. The modelsalso focus attention on basic problems anduncertainties that remain. None seemswholly satisfactory, nor are the argumentsagainst each totally convincing, becauseinsufficient data are available and moreanalysis is necessary in order to chooseamong them. Future work should includea lot of water tracing, discovery and sur-vey of new caves, observation of water lev-els and water-level fluctuations in caves andspecially drilled wells, measurement of theelevations of the smaller springs, pumptests, isotopic analysis of ground water,observation of the flow characteristics ofthe smaller springs, hydrochemical analy-sis of the smaller springs during flooding,more analysis of the Mante hydrograph, andattempts to determine the amount of storagein the aquifer as a function of the cross-sectional area of solution openings. Espe-cially recommended are lycopodium sporetracing tests of Sótanos Venadito, Yerbaniz,and possibly Tigre, with nets placed in theMante, the Choy, and all springs in be-tween. Presently, only a small portion ofthe regional base flow is used for irrigation,and all of the flood waters escape unused.It is suggested that further studies couldshow how to use the El Abra aquifer as anatural reservoir so that its great water-supply potential can be developed for large-scale irrigation of the coastal plain.

7.4. Morphology and Origin of theCaves of the Sierra de El Abra

This section summarizes the morpho-logic features of the El Abra caves andpresents models of the hydrology of thecaves and of their relationship to the sur-face geomorphology.

7.4.1. Character and origin of the eastface caves

The springs of the Sierra de El Abra dis-charge the integrated output of vast cavernsystems within the range. The accessible

Figure 7.16. Comparison of Mante, Santa Clara, and Río Comandantehydrographs and precipitation for July–August 1972.

169

AMC

S Bulletin 14 — C

hapter 7

Figure 7.17. E-W cross-section showing relationships of geology, hydrology, and caves in the Sierra de El Abra.

170


portion of the active spring outlet caves isextremely short. No lengths greater than100 m are known, and sometimes the out-let is completely flooded. This is becausevertical circulation is so important in theEl Abra range; water is discharged from thephreatic zone. At the Choy spring, waterrises at least 26.5 m from its source at thebottom of a lake (Figure 6.1) to flow ontothe coastal plain. The Mante spring is ac-cessible for about 30 m before becomingflooded, and it appears to discharge via alarge horizontal conduit for at least thisshort distance. For most of the smallersprings, only the rise pool can be observed,the exception being the Nacimiento de Ar-royo Seco on Rancho Peñon. At the lattersite, there is a small upper cave and a lowerhorizontal passage that is the water source.This spring is seasonal, which suggests thatit is in a fluctuating water table zone, i.e.,that it is a spill over of wet season higherwater levels in the aquifer. The caves be-hind the presently active El Abra springsare clearly formed within the phreatic zone.

It is believed that all of the east facecaves of the El Abra range that were ob-served were created entirely within thephreatic zone. Morphologic features thatindicate this include broad and deep walland ceiling pocketing, smooth blind domes,and highly irregular ceilings formed by so-lution upward along joints (see the ceilingprofile of Ventana Jabalí, Figure 6.8; ElChoy; Cuates), highly irregular (ungraded)bedrock floors (El Choy, Ventana Jabilí,Ceiba, Taninul No. 2), spongework (TaninulNo. 2), bedrock natural bridges, and othermanifestations of uniform solution underwater-filled conditions. Furthermore,Taninul Nos. 1 and 2 are three-dimensionalmazes or complexes that required develop-ment in the zone of saturation. The east facecaves grew to their present size entirelywithin the phreatic zone and have been onlymodestly altered by clastic sedimentinfilling, guano and stalagmite deposition,and ceiling collapse.

Bonet (1953a) believed the caves in theescarpment were formed in the nuclei oftabular reefs (not like modern coral reefs),but sometimes extended into adjacent peri-reef conglomerates. Although he stressedthe importance of joints in the developmentof all the caves of the range, it seems im-plied that he believed reef porosity to bethe controlling factor. The occurrence ofthe east face caves in reef nuclei as opposedto other facies of the reef zone has not beentested in this study. However, from obser-vations in the caves it appears that very

large joints and bedding (facies) planesprovided the openings for water flowthrough the reef zone and therefore becamethe loci of cave development. Some faciesof the reef zone are probably more solublethan others, so that a passage may vary con-siderably in dimension when passing fromone facies to the next. The caves are classi-fied into two types: short, vertical fissuresor canyons having great height and con-trolled by a major joint or joints, as El Choy,la Ceiba, and Cuates, and relatively hori-zontal conduits that are longer, moreequidimensional in cross section, and con-trolled by a facies plane in combinationwith smaller joints, as exemplified byQuintero and Taninul No. 2. Ventana Jabalíis such a huge chamber (passage ?) that itis difficult to scrutinize. Many large jointswere utilized by circulating water, but otherfactors may also have been important.

The relationships between the geology,the geomorphology, and the east face cavesmay be seen in Figure 7.17. The activespring caves all occur at the base of thescarp, albeit at various elevations along it.Basically, water spills from the phreaticzone of the range onto the impermeablecoastal plain. As previously noted, theremay be a considerable lift, as at El Choy,where the final portion of the ground-waterflow system is up a large joint. A temporalsequence is envisioned in which the young-est spring sites occur at or a very short dis-tance in front of the scarp (e.g. Santa Claraand a spring on Rancho Tampacuala), ElChoy is an older although still active spring,and caves higher up the scarp are fossilspring sites that were active when thecoastal plain stood at much higher levels.The phreatic morphology and height of theChoy cave clearly require it to have beenan active spring for a very long time (Fig-ures 6.1 and 7.17). Flat-topped hills in frontof the cave are remnants of an older coastalplain level that probably correlates withMiddle Entrance (or possibly a fossil exitat the present skylight), and prior to thattime water rose up the vertical pit at theback end of the cave to flow out the UpperCave Entrance onto an even older coastalplain. The total phreatic lift was at least 120m. The phreatic morphology of caves suchas Ventana Jabalí requires ancient coastalplains to have reached the level of the high-est of such features in the caves, and thusto the highest part of the El Abra range.

Models of east face caves and their re-lationship to the coastal plain are shown inFigure 7.18. The paleohydrology of theChoy cave has been discussed above.

Where a major joint or group of joints ex-ist over a large vertical range, as at El Choy,a cave at that location may function as aspring for a very long time. As the coastalplain is gradually lowered, new lower out-lets from the reservoir are developed.Where the fossil spring caves are controlledby a reef facies plane, as at Grutas deQuintero, there is less chance for contin-ued use of the cave as the base level is low-ered. Quintero must have had a simple riseshaft, now missing, in the El Abra lime-stone or coastal plain rocks. The ArroyoSeco spring cave on Rancho Peñon, whichcan be explored a few tens of meters, ap-pears to be a horizontal conduit that is onthe verge of being abandoned as a springsite. The cave above the spring shows thatwater from this same source conduit for-merly flowed up east-dipping facies planesin the reef rock adjacent to the coastal plainto resurge 15 to 20 m above the presentlevel. Caves such as Ventana Jabalí andCueva de La Ceiba are immense voids con-trolled by large vertical joints and perhapsby host rock that was especially favorablefor solution. They also demonstrate greatvertical range of circulation and probablyhad lifting shafts at their outlets as proposedfor Quintero. The source(s) of water toVentana Jabalí is probably buried beneathsediments. La Ceiba, however, has a cen-tral pit series having phreatic morphologythat suggests it was the principal source forthat cave. It would thus have had a flowsystem very similar to El Choy and aphreatic lift of at least 190 m. Also shownin Figure 7.18 is a schematic model of theevolution of the discharge portion of an ElAbra karst flow system. Water is dischargedfrom the aquifer at some established favor-able site, here shown as a series of inter-connected or closely spaced joints. As thecoastal plain is eroded away, the springpoint lowers until the bottom of the joint isreached. After that time, water can nolonger dissolve the conduit downward suf-ficiently rapidly; hence, a new dischargesite must be found. There is no sudden low-ering of base level or sudden draining of aphreatic system. Water gradually abandonsthe upper route. It may flow along a facies(bedding) plane and out, or perhaps inter-sect another joint and then rise upwardalong a joint chimney and through a smallsection of the coastal plain rocks (SanFelipe?). In some instances, lower cham-bers may already have been created, andthe overlying materials may subside into itor be eroded away by flowing water. TheSanta Clara spring may be an example of a

171


diversion some hundreds of meters alongthe strike from a hanging site to a locationat a more favorable elevation and structure.

The Coy spring and Coy caves, in asmaller limestone mass separate from theSierra de El Abra (Figure 6.26), beautifullyillustrate the same kind of phreatic flowsystem, solutional morphology, and evolu-tionary characteristics described above.Above the present Coy spring are caves thatwere former water-filled discharge con-duits, and in at least one of them it is pos-sible to descend to part of the presentlyactive flow system. Many of the other ma-jor springs in the region appear to have atleast some phreatic lift. The most notableis the Media Luna, which discharges fromthree orifices at the bottom of a 75 m deeplake.

One type of cave not observed in thisstudy that may occur along the east face isthe vadose fissure or shaft. These may formwhere small concentrations of water run-ning down the face are captured down ajoint or where a stream has in the pastdrained or presently drains from the top ofthe range and is pirated on the face. Thesecaves will be very small compared with thespring caves of the El Abra.

7.4.2. Character and origin of the crestcaves

Before the thesis research was begun,virtually nothing was known about thecaves on the crest of the El Abra range. Thisstudy and other recent exploration and sur-vey work have greatly increased knowledgeof the area. Because so much effort is of-ten required to reach caves on the crest, thetotal number that have been explored is stillrelatively small; however, many importantfeatures of the caves are now known.

Caves on the crest of the range may beclassified into two basic types, old phreaticcaves and younger vadose systems. Thereis a third type that is considered to be aspecial case, and it will be discussed later.The phreatic caves are very short comparedwith the western margin swallet caves orthose in many other regions of the world.However, what has been observed aremerely short segments of much greatercavern systems. Some of the segments arehorizontal and others are vertical or a com-bination of the two. To illustrate many com-mon aspects of these caves and to indicatetheir importance, a detailed discussion ofthe history and hydrologic implications ofLa Hoya de Zimapán will be given next. Adescription of the cave with photographs

and a map are contained in section 6.3.7.In its formative period, Zimapán was a

great phreatic tunnel, slowly transportingwater in a very deep flow loop. That itsdevelopment was entirely phreatic is shownby the round passage cross section, phreaticdomes, and the smooth sculpturing andpocketing of the walls and ceiling. Waterpassed downward (probably) to a depth ofmore than 300 m, only to return nearly tothe plateau surface and resurge from somefossil spring along the east face of the rangeonto an ancient coastal plain (for example,Ventana Jabalí or Cueva de la Ceiba; Fig-ure 7.17). Thus it is a very old cave, likelyformed soon after or while the top of theEl Abra range was being stripped of itsimpermeable cover and while the coastalplain was approximately at the present crestof the range. It is entirely possible that waterpassed up the conduit instead of down. Ineither case, the cave is a segment of a karstflow system which demonstrates perhaps

better than any other known cave that deepphreatic flow does occur. For that reason,it is considered the most important cavediscovered in Mexico so far.

Then over a period of time (presumablynot instantly), the water supply was some-how cut off, and the conduit ceased tofunction and grow. There may have beenan intermediate period of vadose solutionactivity, but there are no bedrock solutionfeatures to suggest it. Since the flow ceased,erosion has lowered the coastal plain by 400to 600 m. During this time, the cave willhave resembled a well with a standing bodyof water. As the water gradually lowered,vadose seeps were able to reach progres-sively lower into the cave, and stalagmitedeposits began to accumulate. The massivestalagmite deposits in the entrance passagemay be much older than those in theZimapán Room.

How the Zimapán Room developed is apuzzle made difficult because its relationship

Figure 7.18. Models of the east-face caves.

172


to the passage or other passages (?) is hid-den below the sediment floor. If the knowncave is the only passage involved, it seemslikely that the room grew to at least itspresent size in the phreatic zone, while theconduit was an active flow route. Somebreakdown may have been dissolved ordispersed by a later vadose stream, but it isbelieved that this could not have had morethan a minor effect. However, a junction ofpassages below the present floor of theroom could have been a significant factorin promoting its large size. This would beespecially true if the lower passages con-tinued to function long after the upper partof the cave ceased its hydrologic activity.No major fracture was observed in theZimapán Room that would have been im-portant in promoting exceptional rock so-lution or breakdown there. The lower 15 to20 m of the walls (or roof) of the room aresmoothly beveled and pocketed by solution(Plate 6.18); thus, if there has been signifi-cant wall retreat by collapse, the evidenceis lost. The higher part of the roof hasprominent steps left from upward stoping,but it also shows forms indicative of solu-tion under water-filled conditions. Theseeffects can be explained by a fluctuatingwater table or by intermittently pondedvadose water relatively late in the cave’shistory. The Zimapán Room now seems tobe very stable, because there are neithersubsidence trenches in its floor nor a sig-nificant accumulation of collapse boulders.

At some time, a small vadose streaminvaded Zimapán. This may have been theresult of the collapse at the present entrance.It washed soil into the cave that has accu-mulated behind the stalagmite barriers inthe entrance passage, at the base of the thirddrop, and probably as a thick deposit lev-eling the floor of the Zimapán Room.Breakdown and stalagmites may havetrapped the fine-grained sediments, even-tually making the room nearly watertight.Thus, when the vadose stream flowed, itwould be ponded, causing re-solution ofthe stalagmites (Plates 6.19 and 6.20) andperhaps the solution of ceiling rock dis-cussed above. This stream also was respon-sible for making the crawlway through thestalagmite barrier at the top of the first drop,and possibly for the partial erosion of theflowstone in the pit series. These interpre-tations suggest that this ephemeral streamis a very late phenomenon in the cave’s his-tory, and mud in the gour pools and faintchannel marks indicate that a stream stillintermittently invades the cave, althoughit is probably rather small. Vadose drips

enter the cave at all levels, including theZimapán Room, where the whole floor iscovered with gour pools that are active sea-sonally.

The discussion of Zimapán above illus-trates many common morphologic featuresand processes that occur in the crest caves.Each cave was a segment of a much greaterflow system that moved water from sourceson the plateau surface to east face springsvia flow loops in which depth was often animportant dimension. After these great cav-ern systems ceased their hydrologic func-tion, they were separated into small piecesby vadose zone processes such as collapse,stalagmite and flowstone deposition, clas-tic sediment infilling behind barriers andat the bottoms of flow loops, and guanodeposition. Taninul No. 4 and Loros arehorizontal segments of systems; the formeris particularly interesting for its blind jointalcoves and domes and its facies control.Tanchipa and Soyate are other examples ofportions of deep phreatic flow loops. TheSoyate loop must have been at least 234 mdeep.

Access to nearly all of the phreatic cavesis gained by entrances that are usually cre-ated by collapse of high phreatic domes(such as at Loros and Cueva Pinta). Al-though various types of cave sediment mayaccumulate before a collapse entranceforms, the rate of accumulation probablyincrease afterwards. Many of the caves areno more than large rooms where access tothe remainder of the system is blocked(Escalera, Pinta, and many other examplesnot described). Where the original solutionchamber is at considerable depth and theamount of collapse is large, the chamberwill probably be buried by collapse debrisand other types of sediment. This createsthe third type of cave (the special case re-ferred to at the beginning of this section),where no part of the original solution cavecan be reached. Escalera and Cuesta maybe examples of this type. It is hypothesizedthat the “phreatic” re-solution features ofstalagmites and wall rock seen in many ofthe caves are created by water temporarilyponded in chambers that have been madewater-tight by sediment infilling. It appearsthat during storms water from small periph-eral areas flows into the caves. Hence, it isbelieved that the re-solution pocketing isnot associated with fluctuating water tablesor with periods of higher water levels fol-lowing periods of lower water levels thatcould be attributed to geomorphic or cli-matic events.

There are two fundamental problems

concerning the phreatic crest caves thathave not been solved. The first is that sincethe known passage segments are so short,no knowledge of the interrelationships ofpassages have been gained. The second isthat the sources of water or type of rechargethat created the caves are not known. Ofcourse, the two problems are not com-pletely independent. Presently, within thephreatic zone there ultimately must be con-siderable integration of passages, becausethere are a great many more recharge pointsthan there are springs, and the same condi-tion is presumed for the past. However,there could be many instances where waterbecomes divided between two or more dis-tributary channels, which later rejoin orintegrate with other conduits. Thus, thisstudy has shown that very large cavernsdevelop in the phreatic zone, that the flowloops can be very deep, and that structuraland stratigraphic features control the locusof conduit development, but the pattern ofthe phreatic flow system remains unknown.

At La Hoya de Zimapán, there are noknown tributary passages, although vadosedrips and flows have obviously entered thecave at all levels. It would have taken agreat quantity of water and a very long timeto create the cave. There is a fossil surfacearroyo perhaps 200 m long that drained intothe doline at the cave’s entrance. However,unlike the western-margin swallet caves,which were created by pirated arroyos, it issuggested that this arroyo was not the pro-genitor of Zimapán or the entrance doline.Instead, the arroyo may have graduallyevolved as a consequence of the cave col-lapse and may have supplied the vadosestream which created the flowstone depos-its in the pits and transported soil into thecave. The relationship of the surface geo-morphology and hydrology to the cavesduring the period of removal of impermeablerocks from above the El Abra limestoneduring development of karstic permeabil-ity needs further study. The Zimapán en-trance passage is nearly round (Plate 6.16)and 30+ m in diameter. It was createdwholly in the phreatic zone. Any reason-able discharge of water through its 700 m2

cross-sectional area would have such a lowvelocity that there would have been con-siderable sediment accumulation. In fact,sediment would have collected long beforethe present passage size was reached. Apassage which accumulates sediment can-not maintain a circular cross section as itenlarges. Hence, it is concluded thatZimapán was transporting sediment-freewater, and that its source was not a sinking

173


stream off a Méndez or a San Felipe catch-ment. The conduit must have integrated theinfiltration water from a vast number ofsmall openings on a limestone pavementsimilar to the present surface.

Because the coastal plain of Méndez hasbeen denuded much more rapidly than theSierra de El Abra, the topographic reliefon the limestone has increased throughtime. The thickness of the vadose zone hasgrown concomitantly with the relief (Fig-ure 7.17). In the high central portion of therange, the relief and the thickness of thevadose zone may reach 600 m. Solution andcave development in the vadose zoneshould be very important at the presenttime. However, on the limestone surface thecatchment area for the great majority of sinkpoints is quite small. Consequently, only asmall percentage of the openings are hu-manly explorable.

There are many vadose shafts which canbe explored for some distance. Where largerbasins are available, vadose caves such asMonos develop. Monos exhibits strongstructural and stratigraphic controls; thusit appears that those aspects are importantfor the vadose caves as well as the phreaticcaves. There is some integration of flow inthe vadose zone, but the bulk of it occurswithin the present phreatic zone.

7.4.3. Character and origin of the westernflank and western margin caves

On the western flank of the range, somecaves believed to be of phreatic origin areknown. Those observed in this study areCueva de Los Sabinos, Cueva del Prieto,and Sótano de Soyate. It is not anticipatedthat the western flank phreatic caves willbe significantly different from the crestphreatic caves—Soyate was even includedin the discussion of depth of circulation inthe previous section. The caves containphreatic wall pocketing, blind solutiondomes, bedrock pillars, irregular ceilingprofile, uniform solution of wall rock, andjoint networks (in Los Sabinos). Whenmore caves are discovered and studied,subtle differences in solution morphologyand plan-profile characteristics as functionsof structural attitude of strata and of lithol-ogy (and hence distance from the reef) maybe found.

The interesting problems of the phreaticcaves are their sources of water and theirrelationship to the surface geomorphology.All three of the aforementioned caves prob-ably received recharge down joints at theirpresent entrances. Prieto is located on top

of a broad hill of El Abra limestone and isjoint-controlled. At Cueva de los Sabinos,strike and dip measurements (see map, Fig-ure 6.22) indicate that the entrance sink issituated on the crest of an anticline or adome (crest or flank). Some time ago, theRío Sabinos or one of its tributaries passedvery near or over the entrance sink of LosSabinos and probably lost water to the cave.(Río Sabinos is the name given to the an-cestral river which was segmented and pi-rated underground at several places to formswallet caves; see Section 3.1.3 and Figure3.2.) Presently, there is a considerable dif-ference in elevation between the coastalplain and the Río Sabinos valley (Figure7.17). If the upper part (First Level) ofCueva de los Sabinos is truly a phreaticcave (rather than a floodwater cave), thegeomorphic problem is to explain how sucha cave could develop. For any reasonabledischarge, a conduit the size of Los Sabinoswill have a very low hydraulic gradient.This implies that the coastal plain (baselevel) was perhaps 200 m higher when thecave formed, and the difference in eleva-tion between the Río Sabinos valley andthe coastal plain was much less. Possibly,some water leaked through overlying SanFelipe beds on the fold crest before the bedswere completely eroded to expose the ElAbra Formation.

The western margin swallet caves are thelongest and most complex caves known inthe El Abra range. They were initiated longafter crest caves such as Zimapán ceasedfunctioning, and in fact they are still ac-tive. Streams flowing from impermeablecatchments were pirated down the firstmajor joints, typically 50 to 70 m deep, af-ter passing onto the El Abra limestone. Thebasins are large, ranging up to 21.8 km2,and the caves owe nearly all their growthto water supplied from them. The supplyof water is sporadic. Occasional violentfloods enter the caves during the rainy sea-son, and small flows continue for some timeafter each flood. During the remaining time(most of the year), only vadose drips orflows enter the cave.

The swallet caves owe most of theircharacteristics to the combination of highlyvariable discharge of the source and strongstructural and stratigraphic controls; theyare here termed floodwater caves. Passagestend to be tunnels of relatively uniformcross section compared to the highly irregu-lar profiles of phreatic caves such asVentana Jabalí, El Choy, and Taninul No. 4(where blind joint-controlled alcoves arealso common). The floodwater caves exhibit

quasi-phreatic solution features because thepassages fill to the roof during floods. Blinddomes have developed on joints, and shal-low semicircular or ovate domes are formedby solution upward during floods; they areexcellently displayed in Sótano de la Tinaja(see profile, Figure 6.25) and in all the otherswallet caves. It is also clear that many ofthe joints permit percolating waters to enter,so mixing effects might occur, but the mix-ture-corrosion process (Bögli, 1965) shouldnot be important, because the flood watersshould already be considerably undersatu-rated with respect to calcite. Pocketing ofa sort occurs, but the form is not the sameas nor nearly as large as the phreatic pock-eting found in the east face or crest caves.Small scallops might be expected to de-velop, but are very rare. Vadose featuresinclude plunge pools, inlet domes or verti-cal shafts (Pohl, 1955; Brucker, Hess, andWhite, 1972), floor channels, and possiblylateral encroachment on walls by streamsperched on sediments.

Structural and stratigraphic featureshave played important roles in the devel-opment of the caves and in their plan andprofile characteristics. Large fractures ini-tially captured the surface streams and inother places have served as targets forground-water flow, as in Sótano del Arroyo.From sink to spring, water must flowagainst a westerly (on the average) dip anddescend several hundred meters of strati-graphic section. Joints up to 30 m in height,although normally less, provide the meansof crossing or descending the section. Forexample, the Entrance Passage in Tinajaworks against a 5° westerly dip, graduallydescending via small joints along its 525 mlength (projected) until it is physically 24 mbelow the entrance and about 65 m down-section. The initial profile was composedof bedding plane lifts up the dip betweendescents on joints, so that the overall ap-pearance was somewhat similar to the teethof an inverted saw raised at one end. (Thebedding plane segments were of coursemuch longer than the joint segments.)Much of the evidence of the original pro-file has been obliterated, but at least onelarge example remains. From Third Room,water descended via a large joint, thenflowed up the dip on a thin shale layer (2 cmthick) from 5 to 10 m vertically. Thus, an-other common characteristic of these flood-water caves is their irregular profile, i.e.,one that is not continuously descending inthe vadose zone. Floodwater first must fillthe lows of the profile, which then locallyact as pressure conduits. Other examples

174


are the Sandy Floored Passage, where LakePassage must flood to the roof to flow overa bedrock high in the middle of the SandyFloored Passage, and the Wallows andStrickland’s Bad Air Passage in Arroyo.

Various processes cause modificationsof the passage shape and profile. Collapseof the roof, usually assisted by prominentfractures, creates domes not of solutionorigin, rooms, and boulder piles. Collapsedebris occasionally restricts flow and, inall cases, ponds water, creating sedimenttraps for part of the huge quantity of mudand sand swept into the caves. The pondingprobably enhances lateral and upward so-lution. In some instances, the flow is divertedagainst a wall, so that lateral encroachmentoccurs, which further promotes breakdownand the development of a large room. Insome instances, ponded water causes smalldistributary drains to form, which appearto divert water to unknown lower cave pas-sages. The deposition of stalagmites andflowstone is quite prolific in places. Suchdeposits may also restrict flow, backingwater up the passage during floods andcausing ponding. One such instance occursin Neal’s Passage in Sótano del Arroyo,where a large stalagmite or flowstone masshas almost totally blocked this major down-stream passage. In some localities theremay be a rough balance between the rate atwhich trickle waters deposit calcite and therate at which flood waters erode it. Sedi-ment traps may also occur at low spots inirregular floor profiles or behind bedrockbarriers. The sediments found in the flood-water caves are clearly different than thoseobserved in the crest caves (where the lowerparts of the deposits have not been seen).

Most of the swallet caves have each beenformed by only one source of water—thesink of their respective surface arroyos.William H. Russell (see Russell and Raines,1967) has previously advocated that thewestern margin swallet caves were formedin two stages, development of large poorly-connected phreatic voids (rooms and tunnels)by slowly moving water, and integration ofthe voids into large systems by invadingswallet waters. It is believed that his modelplaces too great an emphasis on the amountof phreatic development for these caves. Ofthe swallet caves studied in this work, onlyFirst Room in Tinaja has features suggest-ing that it may owe a significant part (per-haps most in this instance) of its size toprior phreatic solution. It is developed ona large joint (there has been considerableceiling collapse), and water may haveflowed down the joint while it was in the

phreatic zone to create the broad solutionpockets and other features high on thewalls. Alternatively, these solution featuresmight be quasi-phreatic manifestations offloodwater caves. These caves have a dif-ferent solution morphology than the FirstLevel in Cueva de los Sabinos and phreaticcaves on the crest or east face. Hence, it isbelieved that Arroyo, Tinaja, Japonés,Tigre, Jos, and other swallet caves owenearly all of their origin and developmentto the swallet waters.

Sótano de Japonés illustrates the poten-tial that flood waters have to create mazecaves (see detailed discussion, Section6.4.3). As usual for the El Abra, water wasinitially pirated down a joint about 60 to70 m deep. The characteristics of the strataat the base of the joint, in combination withthe high local hydraulic gradient created bydescending flood water, favored develop-ment of a maze from the base of the joint.Those rock characteristics must have beenone or more of the nature of the beddingplane, the jointing, a resistant bed, or strataldip (formation of down-dip tubes). Thewater supply was much greater than theundeveloped aquifer could handle, so thatwater was distributed into a great manyroutes. The system is still relatively young,and water may still back up during largefloods. A second factor which further com-plicates the Japonés system is retrogressivepiracy of the source and of flow within thecave. Sótano de Yerbaniz, a few kilome-ters north, apparently is also a multi-levelmaze system (Mitchell et al., 1977).

Despite the complexity of the Japonéscave, it is considered simpler in origin thanArroyo or Tinaja. Basically, Japonés formedin an isolated area, in bedrock previouslyhaving no more than minor secondary per-meability, when a perched stream forcedwater through many paths essentially radi-ating from one input point. Most ofArroyo’s volume has been created by thesinking arroyo at the entrance of the cave,but it is clear that both Arroyo and Tinajahave numerous small inlets which have in-tegrated with the systems (inlets occur inthe Left Hand Passage, Tracy’s Water Pas-sage, and Steve’s Surprise Passage (?) inArroyo and in Lake Passage, Sandy FlooredPassage, Siphon Pit Series, and Water Pas-sage in Tinaja; see maps in Figures 6.23,6.24). The sources of the inlets are believedto be small sinks or infiltration into the bedof the Río Sabinos dry valley (dry in at leastsome places) and other exposures of the ElAbra limestone over the caves. Althoughthe caves do not exhibit mazes like Japonés,

Arroyo has many distributary channels,including some small ones in Main Passagethat probably formed early in the cave’sdevelopment. Arroyo diminishes in size asthe flood waters are dispersed. Therefore,the Los Sabinos System (Arroyo + Tinaja+ Los Sabinos) must be examined in termsof the principal sinks and their distributarysystems, integrating inlets (which may beenlarged by backflooding) and possiblephreatic voids or enlarged fissures whichcould act as targets for cave development.Sótano del Tigre, north of Arroyo, exhibitsa tendency toward maze and distributarychannel development near the entrance ofthe cave, but nearly all of its flood watersare carried by one major lower-level con-duit.

The relationship of the swallet caves togeology, topography, and base level maybe seen in Figure 7.17. The initial piraciestook place at elevations between 200 m to300 m (before entrenchment of the ar-royos), although the southernmost occurredsomewhat lower; this compares with thepresent elevation of the Choy rise pool of35 m. A careful analysis needs to be madeof the relationships between the ancient RíoSabinos, including the Soyate valleybranch, the swallets, the geomorphologyand profile of the South El Abra Pass, andthe geomorphology of the coastal plain. Theyoungest swallets, such as Jos, Yerbaniz,and Japonés, undoubtedly correlate with thepresent Choy resurgence. The initiation ofthe oldest swallets, most likely Tinaja andArroyo, might correlate with the MiddleEntrance of El Choy and the 77 m coastalplain level, or at least with some highersurface than the present. A stalagmite ly-ing on sediment in the Sandy Floored Pas-sage was dated by the uranium-thoriummethod by Peter Thompson (1973), whoestimated the base of the piece (above thetrue base of the stalagmite) to be 50,000years old. This suggests that these cavesmay be at least a few hundreds of thousandsof years old. Waters entering the swalletsnecessarily must enter the phreatic zonebefore they can be discharged at El Choy.However, the lowest point reached in Ar-royo may be the only passage yet observedin these caves that might have formed be-low a fossil water table. Except for simplevadose zone piracies, the caves are believedto have been created as a single phase ofcontinuous growth; they have been con-trolled by rock structure, sediment infilling,and floodwater conditions, not by one ormore fossil water tables. This conclusionis easily accepted for the younger caves

175


such as Japonés, but the origin of Steve’sSurprise Passage in Arroyo and the SandyFloored Passage pose some problems.

7.4.4. Discussion

In sumary, most of the large caves foundon the east face, the crest, and the westernflank of the Sierra de El Abra were parts ofgreat phreatic flow systems. The locus offlow (and thus of cavern development) wasstrongly controlled by bedding planes andlarge joints, a style of development advo-cated by Ford (1971) for other areas. Thepurity and massive bedding of the lime-stone, the large vertical range of joints, thelack of significant shale or other imperme-able interbeds, and the great thickness ofthe limestone below base level have madedepth an important dimension of the flowsystems. La Hoya de Zimapán was part ofone such system which penetrated at least300 m (possibly much more) below an an-cient water table, thus demonstrating theexistence of deep phreatic flow. The re-charge of a large number of small sinkssomehow integrated, to a considerable ex-tent within the phreatic zone, to form largeconduits that discharged their waters fromfossil spring caves, such as Ventana Jabalí,onto ancient and much higher coastalplains. Thus, these caves are equal in ageto the time that has been required to erodeup to as much as 600 m of the coastal plain.They are some millions of years old ratherthan post-Wisconsin, as suggested byHarmon (1971).

Probably, the cave most similar to thosedeveloped in the El Abra is Carlsbad Cav-erns in the Guadalupe reef complex in NewMexico (Bretz, 1949). The El Abra cavescould well serve as type examples of thedeep phreatic solution features and patternof circulation postulated by Davis (1930)and supported by Bretz (1942). Except lo-cally, the location of flow bears no rela-tionship to the water table (or potentiomet-ric surface). In areas where the jointing isfiner, the bedding usually thinner, andstratigraphic or lithologic controlling bedsor partings more numerous, caves have of-ten been attributed to water table controlor shallow phreatic development. White(1969) has proposed conceptual models forflow systems developed in carbonate rockshaving various hydrogeologic settings andflow properties; the models were for lowto moderate relief terrains. His model foropen, free flow karst systems having cir-culation below valley level emphasizesshallow phreatic development. The caves

in the El Abra area indicate that anothervariable—the structural-stratigraphic prop-erties of the host rock—needs to be in-cluded, because, at least in some areas, deepphreatic flow exists and is important.

The modern flow systems of the El Abrarange proper have much thicker vadosezones than the ancient systems, but circu-lation within the phreas probably is stillquite deep. Vadose zone caves on the rangehave not received much attention in thisstudy, because they generally are small,having only small catchments; however,further exploration may yield some signifi-cant vadose caves.

Swallet caves along the western marginbelong to a special class of vadose cavescalled floodwater caves. The swallets gen-erally lack entrenchment features, potholes,and the continuously descending profileindicative of vadose caves which carrysmall to moderate discharges (in compari-son with the capacity of the passage). In-stead, the structural properties of the lime-stone and the extremely high dischargesprovided during floods create a host of fea-tures that are quasi-phreatic in character.Palmer (1972) distinguishes floodwatercaves as an important class of cave anddescribed several aspects of their morphol-ogy and hydraulics. The El Abra swalletsbelong to that class and add variety to thepassage plan and profile characteristics andthe morphologic features that may be foundin that class. The swallet entrances and pas-sages, especially when primitive, are over-loaded by the flood waters and sometimescreate a maze or a complex of distributarychannels. They are excellent examples ofsome of the speleogenetic principles Ew-ers (Ph.D. thesis in preparation [1982]) hasobserved in his artificial modeling.

A few final points are: (1) Relief has notbeen an important variable in the El Abracaves. The nature of the ground-water cir-culation has not changed, although the re-lief and the thickness of the vadose zonehave increased. (2) No major differencesin the processes of cavern developmenthave been found between caves in the ElAbra and caves in other climatic regions,although the El Abra caves are larger thanthose in many other regions. There may bemore subtle differences, such as large con-centrations of organics washed into thecaves and the growth of stalagmite barriers,which have their effects. (3) It is believedthat a greater percentage of the dissolvedload of the springs is obtained in the sub-surface than in many (perhaps most) karstareas.

7.5. Summary and Conclusions

The region of interest is the Valles–SanLuis Potosí Platform of Cretaceous age and,in particular, the Sierra de El Abra alongits eastern margin. A thick Lower Creta-ceous deposit of gypsum and anhydriteoccurs over a wide area in the interior ofthe Platform. More than 1200 m of themiddle Cretaceous El Abra limestone witha basal dolomite was deposited over theentire Platform. Upper Cretaceous platformlimestones are limited to the central area,and eventually the whole Platform was cov-ered with thin bedded limestones and athick section of shale and sand. The areawas folded and uplifted in the early Ter-tiary (possibly beginning in the latest Cre-taceous).

The principal facts and conclusions ofthis study are summarized below.

Erosion of the Platform region beganprobably in the late Tertiary. Streams gradu-ally removed the impereable rocks from thecrests and flanks of anticlines and thethrust-faulted Xilitla massif, exposing theEl Abra limestone to solutional attack.

A high relief karst has developed, butthe largest features owe their size to struc-tural relief of the limestone rather than tosolution. A large part of the Platform is nowa karst developed on the El Abra andTamasopo Formations.

A very wide range of karst features arefound, including areas with deep karren,large pinnacles, and mogotes or haystackhills that are commonly found in humidtropical karsts. Climate is probably the mostimportant control on the types of karst fea-tures that develop in the region, but slopeand various aspects of rock structure (inthe geomorphic sense) are important also.

The Sierra de El Abra was formerly cov-ered with impermeable rocks which havebeen stripped off. Traces of paleofluvialdrainages exist on the flat crest (3 km wide),dry valleys are quite prominent on lime-stone on the western flank, and active drain-ages occur on impermeable rocks along thewestern margin; some of the latter are pi-rated underground to form large swalletcaves. Karren, shallow solution dolines,and large collapse dolines occur on thecrest, but none of the typical tropical karstforms are found. Soils are thin and discon-tinuous, because the limestone is very pure,and where found contain abundant vegetalmatter.

The coastal plain formerly stood at leastas high as the crest of the El Abra range.

176


The eastern escarpment is partly formed byexhumation of a steep carbonate platformmargin, and probably partly by structuralrelief from folding or faulting.

The El Abra Formation (plus theTamasopo Formation and Lower Creta-ceous platform limestones) forms a mas-sive, continuous regional aquifer. Groundwater passes through solutionally enlargedchannels ranging in size from cracks toconduits (caves) perhaps as large as 30 min diameter. All other rocks are relativelyimpermeable, except thin sediment veneersin western valleys. Recharge to the El Abralimestone is believed to occur throughoutmost of the region, except where imperme-able rocks remain. Nearly all of the rechargeis by infiltration through a myriad of smallsinks or cracks rather than by capture oflarge streams.

Rainfall is strongly spatially distributedinto a wet eastern zone (800 to 2500 mm/year) and semiarid western zone (250 to600 mm/year) and temporally distributedinto a wet season (June–October) and a dryseason (September–May).

Runoff from karstic and non-karstic ba-sins on the Platform closely correlates withannual precipitation, with the season (wetor dry), and with individual precipitationevents. The karst springs are extremelydynamic. Most have discharge variations(Qmax/Qmin ) of 25 to 100 times in responseto major precipitation events. This indicateswell developed conduit flow systems andrapid discharge of infiltrating flood waters.A response at springs usually occurs withina few hours after the initiation of majorstorms, and the crest is usually reachedroughly 24 hours later.

The region has many large springs. Atleast eight and perhaps twice that many arebelieved to have average flows greater than3 m3/ sec (i.e., are first magnitude springs).The Coy (about 24 m3/sec) and the Fríospring group (about 28 m3/sec) are amongthe largest springs in the world. An evenlarger number of springs have peak flowsgreater than 100 m3/sec (3530 c.f.s.).

Several lines of evidence indicate thata large part of the discharge of the springshas had a short residence time in the aqui-fer, ranging from one day to a few months.Some ground water has a longer residencetime.

Cave waters sampled in the Sierra de ElAbra were all of the calcium bicarbonatetype, with high Ca/Mg ratios. Spring wa-ters in the region were of three kinds, smallto large springs having “normal” calciumbicarbonate type water (examples are Santa

Clara, Huichihuayán, and Sabinas), largesprings having very high concentrations ofCa, Mg, and SO4 and approximately nor-mal amounts of HCO3 (examples are Choy,Coy, Mante, Media Luna), and thermal,sulfurous springs having variable chemis-try, including one which has significantamounts of Na and Cl.

Measurement of the physical and chemi-cal characteristics of spring waters throughflood pulses provides a valuable means ofstudying the dynamics of karst aquifers. Atthe Choy and Coy springs, wet season floodpulses cause their chemistries to diminishto calcium bicarbonate type waters similarto other nearby springs. For each spring,there was a strong correlation between Mgand SO4 and between temperature and SO4.

Based on the chemical compositions ofthe springs, the correlated behavior of Mg,SO4, and T, and the distribution of miner-als in the region, it is concluded that thehigh sulfate springs have two sources ofwater, whereas the calcium bicarbonatesprings only have one source (local) ofwater. A geochemical model was developedin which water circulating at moderatedepths within the El Abra limestone dis-solves mostly calcite (type A water) andwater circulating deeply in the central orwestern part of the region dissolves calcite,gypsum (Guaxcamá Formation), and thebasal dolomite of the El Abra Formation(type B ground water) and is warmed bygeothermal heat. Wherever type B waterresurges, it mixes with type A water to forma mixed output.

The chemical data of the springs restrictsthe possible compositions of type B water.Solutions having –0.3 ≤ SIg ≤ 0 and 0 ≤SIc–SId ≤ +0.12 with SIc ≤ +0.3 are satis-factory compositions for type B groundwater. The equilibrium model works wellfor El Choy and has the highest possibletotal sulfate; other springs require type Bwater to have a small difference betweenSIc and SId.

It is concluded that large regionalground-water flow systems exist, based onthe spring water chemistries and the distri-bution of source minerals in the region, thethermal characteristics of the springs, theconcentration of base flow at a few largelow-elevation springs along the easternmargin, the occurrence at higher elevationsof very large areas that have no surface run-off (particularly the whole northwesternquadrant—about 37%–of the Platform) andthat are capable of recharge, the Coyspring, which can only be supplied fromdistant sources, deep solution effects and

permeability recorded in oil explorationwells, and the continuity of the supposedaquifer. The general correlation of param-eters (which may be modified by local fac-tors) is: Springs that discharge only calciumbicarbonate type water have only localsources, lower base flows and the greatestvariability of discharge, a higher elevationthan nearby high sulfate springs, and a tem-perature reflecting the elevation of the lo-cal recharge area. Springs that have highCa, Mg, and SO4 (mixed output) have lo-cal and regional sources, higher base flowsand somewhat lower variability of dis-charge, lower elevation than nearby calciumbicarbonate type springs, and higher tem-peratures than nearby calcium bicarbonatetype springs. The areal extent of the El AbraFormation (the same as the Valles–San LuisPotosí Platform) essentially delimits theregion of active ground-water flow. Thegeneral direction of regional flow is east-erly or southeasterly, and the flow path maybe 100 to 200 km long. The coastal plain isa barrier to flow and acts as the spill pointof the aquifer.

Mixing model calculations using a two-source dissolved load equation and SO4 andMg as conservative parameters show thatthe total regional flow of type B water tothe large eastern springs is 12 to 27 m3/sec,depending on the chemical model selected,that the flow of type B water is approxi-mately constant even during flooding of thesprings (possibly varies slightly with sea-sonal hydraulic head), and that the type Alocal flow is the dynamic component ofspring discharge. It is possible that regionalflow of type A water also exists, but it can-not be detected by the chemical mixingmodel. The Coy is the great regional outletin the southern part of the region, and theMante is its northern, but lesser, counter-part.

Ground water storage in the Sierra deEl Abra appears to lie mainly in the phreaticzone.

The lake in Sótano de Soyate is believedto lie at the water table. However, the cal-culated hydraulic gradient from Soyate toEl Choy is somewhat larger than would beexpected for the size of conduits that havebeen observed in the range (the late dryseason level or decline is not yet known).

Hydrographs of the lake show that largefluctuations of the water table occur dur-ing storms. Hydrographs of the sump in Josin conjunction with the geomorphic char-acter of the swallet caves indicate that alltheir “terminal” siphons are perched andthat free-air passage lies on the other side.

177


The water budget and ground-waterbasins of the Sierra de El Abra are morecomplicated than previously envisioned.The two largest springs, the Mante and theChoy, receive a significant amount of wa-ter from sources outside the El Abra range(possibly more than 50% for the Mantespring). Furthermore, although the smallersprings have very small base flows, theirflood flows are sufficient to form an im-portant part of the total discharge from therange. Much more field work is requiredbefore the many problems can be resolvedand the aquifer developed for water sup-ply.

Most of the large explorable caves onthe crest and east face of the El Abra range

were parts of great deep phreatic flow sys-tems. Circulation within the phreas reachedat least 300 m (and perhaps much more)below ancient water tables.

These flow systems discharged throughfossil spring caves on the east face of theEl Abra onto ancient, much higher coastalplains. Thus, the caves are very old.

The locus of flow was structurally con-trolled. The factors which permitted deepcirculation are believed to be joints of greatvertical range, thickness of bedding andeven greater distance between open bed-ding planes, lack of major and continuousperching layers, great depth of aquifer be-low base level, and possibly proximity tothe reef.

The modern systems are probably simi-lar to those of the past, except that the thick-ness of the vadose zone has increased.

The western margin swallet caves, thelongest caves known in the El Abra, arefloodwater caves. They owe their charac-ter to structural stratigraphic control, to theextreme hydraulic gradients created, espe-cially in the primitive stages when the de-veloping conduits cannot transmit all thewater supplied, and, to a limited extent, tothe large amount of sediment swept into thecaves.

No major differences in the processesof cavern development have been found inthe caves of the El Abra as compared withother regions.

178

AMCS Bulletin 14 — Appendixes

APPENDIX 1

APPENDIX 2

Water samples for springs, cave lakes, andstreams were collected (in duplicate) 10 to15 cm below the surface in polyethylenebottles and sealed tightly before being re-moved. The spring samples were taken di-rectly at the source, except for two instanceswhere that was not possible; the analysesof the latter were relegated to Table 5.1b.

Water temperature was usually taken 5to 10 cm below the surface using a labora-tory thermometer graduated in 0.1°C inter-vals. The thermometer was checked againstothers. The pH of spring waters and sur-face streams was measured on site using aMetrohm portable meter. Care was takento bring the electrode and buffer to the sametemperature as the sample to eliminate drift.The readings were occasionally checkedagainst other buffers. The 1971 samplesusually have an accuracy of 0.05 to 0.1 pHunits, and the 1972 and 1973 samples arebelieved accurate to about ±0.03 units.Except for two samples, the pH values of

the cave lakes were measured at the fieldheadquarters. Tests indicate that not muchdrift would normally occur if the sampleswere processed as soon as possible. Theirvalues are believed accurate to 0.05 to 0.10units. The pH of drip waters could be seri-ously in error, even though those listed weremeasured in the cave.

Alkalinity was determined by a poten-tiometric titration, using 0.0033N and0.0098N HCl. Some samples were analyzedon location, but most were taken in tightlysealed bottles to the field laboratory foranalysis. The results were good when thesample was processed within a few hours.Considerable care was taken to avoid agi-tation when transporting and processing.The precision is probably ±2% for the 1972and 1973 samples, but somewhat worse for1971. Calcium and magnesium were deter-mined using a commercially availableSchwarzenbach titration kit with EDTA.Calcium was determined directly, and

magnesium was calculated as the differencefrom the total hardness. Duplicate analy-ses were made in 1972 and 1973, and if theresults of the trails were significantly dif-ferent, a third measurement was made. Theprecision of the data is considered to bebetter than ±1% for calcium and ±2% formagnesium for 1972 and 1973. Greater er-rors could occur for the 1971 data, whensamples were only tested once. Sulfate wasdetermined at McMaster University by thevisual thorin method (Rainwater andThatcher, 1960). Samples were run twice,and the results are accurate to ±1% for highconcentrations, but the error is larger forlow concentrations. Sodium and potassiumfor a few samples were analyzed using aflame photometer. Chloride was determinedby two methods: specific-ion electrode forsome samples and Hach field kit for roughvalues on the other samples. The sodiumand potassium analyses were made by JillGleed.

The symbols used on the cave maps in thisthesis are shown on the following pages.They attempt to illustrate or “picture” the es-sential features of the caves without becom-ing lost in a multiplicity of fine detail. Themaps were drawn over a period of years, andas new problems or needs were encounteredor a better method of presentation discov-ered, a few symbols were changed. There-fore in some cases where two symbols are

shown, the one on the left is now preferred.Bedrock walls are drawn with a heavier linethan detail for contrast; this also distinguishesbedrock pillars from boulders. A few sym-bols not required here are included to makethe legend of more general use. Lengths anddepths on the maps drawn for this study arein meters, but the three maps provided byNeal Morris use feet.

179

AMCS Bulletin 14 — Appendixes180

AMCS Bulletin 14 — Appendixes 181

AMCS Bulletin 14 — Appendixes182

AMCS Bulletin 14 — Bibliography

BIBLIOGRAPHY

Abbott, Patrick L., 1974, Calcitization ofEdwards Group dolomites in theBalcones fault zone aquifer, south-central Texas. Geology, p. 359–362.

———, 1975, On the hydrology of theEdwards Limestone, south-centralTexas. Jour. Hydrology, v. 24, p. 251–269.

Ashton, K., 1966, The analysis of flowfrom karst drainage systems. CaveRes. Gp. of Great Britain, Trans., v. 7,no. 2., p. 161–203.

———, 1967, The University of Leedshydrological survey expedition toJamaica 1963: Hydrology. Cave Res.Gp. of Great Britain, Trans., v. 9, no. 1,p. 36–45.

Atkinson, T. C., D. I. Smith, J. J. Lavis,and R. J. Whitaker, 1973, Experimentsin tracing underground waters inlimestones. Jour. Hydrology, v. 19, p.323–349.

Back, William, and Bruce B. Hanshaw,1970, Comparison of chemicalhydrogeology of the carbonatepeninsulas of Florida and Yucatan.Jour. Hydrology, v. 10, p. 330–368.

———, 1974, Hydrogeochemistry of thenorthern Yucatan Peninsula, Mexico,with a section on Mayan waterpractices. p. 45–77, in Field Seminaron Water and Carbonate Rocks of theYucatan Peninsula, Mexico, ed. A. E.Weidie, New Orleans GeologicalSociety for Field Trip 2, 1974 AnnualMeeting, Miami, G.S.A., 274 p.

Bassett, John, 1971, Cueva de Ceiba.Bloomington Indiana Grotto News-letter, v. 10, no. 1, p. 14–16.

———, 1976, Hydrology and geochem-istry of the Upper Lost River drainagebasin, Indiana. Nat. Speleo. Soc. Bull.,v. 38, no. 4, p. 79–87.

Bögli, Alfred, 1965, The role of corrosionby mixed water in cave forming, p.125–132. in Problems of the Speleo-logical Research, ed. 0. Stelcl,Proceedings of the InternationalSpeleological Conference held inBrno, June 29–July 4, 1964, Aca-demia: Publishing House of theCzechoslovak Academy of Sciences,Prague, 220 p.

Bonet, Federico, 1952, La faciesurgoniana del Cretacico Medio de la

Brucker, Roger W., John W. Hess, andWilliam B. White, 1972, Role ofvertical shafts in the movement ofground water in carbonate aquifers.Ground Water, v. 10, no. 6, p. 5–13.

Burdon, David J., and Nicolas Papakis,1963, Handbook of Karst Hydro-geology with Special Reference to theCarbonate Aquifers of the Mediterra-nean Region. F.A.0. of UNESCO,276 p.

Burdon, David J., and Chafic Safadi,1963, Ras-El-Ain: the great karstspring of Mesopotamia, a hydrogeo-logical study. Jour. Hydrology, v. 1, p.58–95.

Carrillo Bravo, José, 1963, Geology ofthe Huizachal-Peregrina Anticlinoriumnorthwest of Ciudad Victoria,Tamaulipas. in Peregrina Canyon andSierra de El Abra, Corpus ChristiGeological Society Annual Field Trip,Corpus Christi, Texas, p. 11–23.

———, 1968, Exploración geologica yposibilidades petroleros de laPlataforma Valles–San Luis Potosí(Sierra Madre Oriental–AltiplanoMéxico). in Mesa Redonda no. 6,Problemas de Exploración en AreasPosiblemente Petroliferas de laRepublica Mexicana, paper no. 5, 20 p.

———, 1971, La plataforma Valles–SanLuis Potosí. Bol. Asoc. MexicanaGeologos Petroleros, v. 23, no. 1–6,p. 1–113.

Coogan, A. H., D. G. Bebout, and CarlosMaggio, 1972, Depositional environ-ments and geologic history of GoldenLane and Poza Rica trend, Mexico, analternative view. Am. Assoc. PetroleumGeologists Bull., v. 56, no. 8, p. 1419–1447.

Cowell, Daryl, 1976, Karst Geomorphol-ogy of the Bruce Peninsula, Ontario.M.A. Thesis, Dept. of Geography,McMaster University, Hamilton,Ontario.

Dalrymple, Tate, 1960, Flood-frequencyanalyses, manual of hydrology: Part 3.Flood-flow techniques. U.S. Geol.Survey Water-Supply Paper 1543-A,80 p.

Davies, William E., 1960, Origin of cavesin folded limestone. Nat. Speleo. Soc.Bull., v. 22, no. 1, p. 5–18.

región de Tampico. Bol. Asoc.Mexicana Geólogos Petroleros, v. 4,no. 5–6, p. 153–262.

———, 1953a, Datos sobre las cavernasy otras fenómenos erósivos de lascalizas de la Sierra de El Abra. Mem.Cong. Cient. México, Cien. Fis., Mat.,Geol., 5, p. 238–273.

———, 1953b, Cuevas de la SierraMadre Oriental en la región de Xilitla.Bol. Univ. Nac. Aut. Mex., Inst. Geol.,v. 57, 96 p.

———, 1963, Biostratigraphic notes onthe Cretaceous of eastern Mexico. inPeregrina Canyon and Sierra de ElAbra, Corpus Christi GeologicalSociety Annual Field Trip, CorpusChristi, Texas, p. 36–48.

Boyd, Don R., 1963, Geology of theGolden Lane trend and related fieldsof the Tampico embayment. inPeregrina Canyon and Sierra de ElAbra, Corpus Christi GeologicalSociety Annual Field Trip, CorpusChristi, Texas, p. 49–56.

Boyd, D. R., J. Carrillo Bravo, E. K.Krause, R. B. Mixon, and G. E.Murray, 1963, Generalized CorrelationChart. in Geology of PeregrinaCanyon and Sierra de El Abra,Guidebook for Eleventh Annual FieldTrip, Corpus Christi Geological Soc.

Bretz, J H., 1942, Vadose and phreaticfeatures of limestone caverns. Jour. ofGeology, v. 50 (L), no. 6, p. 675–811.

———, 1949, Carlsbad Caverns andother caves of the Guadalupe block,New Mexico. J. of Geology, v. 57, no.5, p. 447–463.

Brook, George, 1976, Geomorphology ofthe North Karst, South Nahanni RiverRegion, Northwest Territories,Canada. Ph.D. thesis, Dept. ofGeography, McMaster University,Hamilton, Ontario, 627 p.

Brown, Charles, 1970, Karst Geomor-phology and Hydrology of the LowerMaligne Basin, Jasper, Alberta. Ph.D.thesis, Dept. of Geography, McMasterUniversity, Hamilton, Ontario, 178 p.

Brown, M. C., and D. C. Ford, 1971,Quantitative tracer methods forinvestigation of karst hydrologicalsystems. Trans. Cave Research Groupof Great Britain, v. 13, p. 37–51.

183


Davis, W. M., 1930, Origin of limestonecaverns. Geol. Soc. Am. Bull., v. 41, p.475–628.

Deike III, George H., 1960, Origin andgeologic relations of Breathing Cave,Virginia. Nat. Speleo. Soc. Bull., v. 22,no. 1, p. 30–42.

Drake, J. John, 1974, Hydrology andKarst Solution in the SouthernCanadian Rockies. Ph.D. thesis, Dept.of Geography, McMaster University,Hamilton, Ontario, 222 p.

Drew, D. P., 1966, The water tableconcept in limestones. BritishSpeleological Association Proc. 4, p.57–68.

Drew, David P., and David I. Smith,1969, Techniques for the tracing ofsubterranean drainage. Tech. Bull. No.2, British Geomorphological ResearchGroup, 36 p.

Eakin, Thomas E., 1966, A regionalinterbasin groundwater system in theWhite River area, southeasternNevada. Water Resour. Res., v. 2, no.2, p. 251–271.

Elliott, William R., 1970, El Sótano deSoyate. Texas Caver, v. 15, p. 63–66.

Enjalbert, H., 1964, Phenomeneskarstiques Mexique et au Guatemala.Bull. Assn. Geog. Francais, 324–325,p. 30–58.

Enos, Paul, 1974, Reefs, platforms, andbasins of middle Cretaceous innortheast Mexico. Am. Assoc.Petroleum Geologists Bull., v. 58, p.800–809.

Ewers, Ralph O., 1982, Cavern Develop-ment in the Dimensions of Length andBreadth. Ph.D. thesis, Dept. ofGeography, McMaster University,Hamilton, Ontario, 398 p.

Faulkner, Glen L., 1976, Flow analysis ofkarst systems with well developedunderground circulation. in KarstHydrology and Water Resources, ed.Vujica Yevjevich, v. 1, p. 137–164,from Proceedings of the U.S.-Yugoslavian Symposium, Dubrovnik,June 2–7, 1975. Water ResourcesPublications, Fort Collins, Coloradao.

Ferguson, G. E., C. W. Lingham, S. K.Love, and R. O. Vernon, 1947,Springs of Florida. Florida Geol.Survey Bull. 31, 196 p.

Fish, John E., 1968, Cave description andspeleogenesis. p. 3–5, in Sótano de lasGolondrinas, Bull. 2 of the Assoc. forMexican Cave Studies, ed. T. W.Raines, 20 p.

———, 1974, La Sistema de Los

Sabinos, Mexico’s longest cave.Canadian Caver, v. 6, no. 1, p. 3–20.

———, 1977, The Circum-Gulf KarstBelt. in Karst Hydrogeology, Proc.Twelfth Int. Congress of Int. Assoc.Hydrogeologists, 1975, eds. J. S.Tolson and F. L. Doyle, AlabamaGeol. Survey, 578 p.

Fish, John, and D. C. Ford, 1973, Karstgeomorphology and hydrology of theSierra de El Abra, S.L.P. and Tamps.,Mexico. Proc. 6th Int. Congress ofSpeleology, Olomouc, CSSR, v. 2,sub-section Ba, p. 151–156.

Fish, John, and William Russell, 1972,“Preliminary results on the groundwater geochemistry of the Sierra de ElAbra region, north central Mexico,”discussion. Nat. Speleo. Soc. Bull, v.34, no. 3, p. 111–112.

Ford, Derek C., 1965, The origin oflimestone caverns: a model from theCentral Mendip Hills, England. Nat.Speleo. Soc. Bull., v. 27, no. 4, p.109–132.

———, 1968, Features of caverndevelopment in Central Mendip. CaveRes. Gp. of Great Britain, Trans., v.10, no. 1, p. 11–25.

———, 1971, Geologic structure and anew explanation of limestone caverngenesis. Cave Res. Gp. of GreatBritain, Trans., v. 13, no. 2, p. 81–94.

Freeze, R. Allan, and P. A. Witherspoon,1966, Theoretical analysis of regionalgroundwater flow: 1. Analytical andnumerical solutions to the mathemati-cal model. Water Resour. Res., v. 2,no. 4, p. 641–656.

———, 1967, Theoretical analysis ofregional groundwater flow: 2. Effectof water-table configuration andsubsurface permeability variation..Water Resour. Res., v. 3, no. 2, p. 623–634.

Garrels, Robert M., and Charles L.Christ, 1965, Solutions, Minerals, andEquilibria. Harper & Row, New York,N.Y., 450 p.

Griffith, L. S., Max G. Pitcher, and G.Wesley Rice, 1969, Quantitativeenvironmental analysis of a LowerCretaceous reef complex. in Deposi-tional Environments in CarbonateRocks, A Symposium, ed. G. M.Friedman , Soc. Econ. Paleontologistsand Mineralogists, Spec. Pub. no. 14,p. 120–138.

Gunnerson, C. G., 1967, Stream flow andquality in the Columbia River basin. J.Sanitary Eng. Div., Amer. Soc. Civil

Eng., v. 93 (SA6), p. 1–16.Hall, Francis R., 1970, Dissolved solids–

discharge relationships, 1. Mixingmodels. Water Resour. Res., v. 6, no.3, p. 845–850.

———, 1971, Dissolved solids–discharge relationships, 2. Applica-tions to field data. Water Resour. Res.,v. 7, no. 3, p. 591–601.

Harmon, Russell S., 1971, Preliminaryresults on the ground-water geochem-istry of the Sierra de El Abra region,North Central Mexico. Nat. Speleo.Soc. Bull.. v. 33, no. 2, p. 73–86.

Harmon, Russell S., William B. White,John J. Drake, and John W. Hess,1975, Regional hydrochemistry ofNorth American carbonate terrains.Water Resour. Res., v. 11, no. 6, p.963–967.

Harmon, R. S., J .J. Drake, J. W. Hess, R.L. Jacobson, D. C. Ford, W. B. White,J. Fish, J. Coward, R. Ewers, and J.Quinlan, 1973, Geochemistry of karstwaters in North America. Proceedingsof the 6th International Congress ofSpeleology, Olomouc, CSSR. v. 4, p.113–114.

Harmon, R. S., J. W. Hess, R. W.Jacobson, E. T. Shuster, C. Haygood,and W. B. White, 1972, Chemistry ofcarbonate denudation in NorthAmerica. Cave Res. Gp. of GreatBritain, Trans., v. 14, no. 2, p. 96–103.

Heim, A., 1940, The front ranges of theSierra Madre Oriental, Mexico, fromCiudad Victoria to Tamazunchale.Eclogae Geol. Helvetiae, v. 20, no. 2,p. 313–352.

Hem, John D., 1970, Study and interpre-tation of the chemical characteristicsof natural water. U.S. Geol. SurveyWater-Supply Paper 1473, 363 p.

Hendrickson, G. E., and R. A. Krieger,1964, Geochemistry of natural watersof the Blue Grass Region, Kentucky.U.S. Geol. Survey Water-Supply Paper1700, 135 p.

Hostetler, P. B., 1963, Complexing ofmagnesium with bicarbonate. Jour.Phys. Chem., v. 67, p. 720–721.

Hubbert, M. K., 1940, The theory ofgroundwater motion. J. of Geology, v.48, p. 785–944.

Jennings, Joseph N., 1971, Karst. M.I.T.Press, Cambridge, Mass., 252 p.

Lange, Arthur L., 1960, Geometricalbasis for cave interpretation. Nat.Speleo. Soc. Bull., v. 22, no. 1, p. 77–83.

184


Langmuir, Donald, 1971, The geochemis-try of some carbonate ground watersin central Pennsylvania. Geochimicaet Cosmochimica Acta, v. 35, p. 1023–1045.

Le Grand, Harry, and V. T. Stringfield,1971, Water levels in carbonate rockterranes. Ground Water, v. 9, no. 3, p.4–10.

Lesser, Juan Manuel, c. 1973, Interpre-tación geoquímica del agua subte-rránea del valle de Ríoverde, San LuisPotosí. unpublished report of theSectretaría de Recursos Hidráulicos,23 p.

Lesser-Jones, H., 1967, Confined freshwater aquifers in limestone, exploitedin the north of Mexico with deep wellsbelow sea level. in Hydrology ofFractured Rocks, v. 2, Proc. ofDubrovnik Symposium, 1965,UNESCO and Int. Assoc. Sci.Hydrology, p. 526–539.

Lucia, F. J., 1961, Dedolomitization inthe Tansill (Permian) Formation. Geol.Soc. Am. Bull., v. 72, p. 1107–1110.

Martin, P. S., 1958, A biogeography ofreptiles and amphibians in the GómezFarías Region, Tamaulipas, México.Mus. Zod., Univ. Michigan, Misc.,Publ. 101 , 102 p.

Maxey, George B., 1964, Hydrostra-tigraphic units. Jour. Hydrology, v. 2,p. 124–129.

———, 1968, Hydrogeology of desertbasins. Ground Water, v. 6, no. 5, p.10–22.

Maxey, George B., and Martin D.Mifflin, 1966, Occurrence andmovement of ground water incarbonate rocks of Nevada. Nat.Speleo. Soc. Bull., v. 28, no. 3, p.141–157.

Mercado, A., and G. K. Billings, 1975,The kinetics of mineral dissolution incarbonate aquifers as a tool forhydrological investigations, I.Concentration-time relationships.Jour. Hydrology, v. 24, p. 303–331.

Mifflin, M. D., 1968, Delineation ofGround-Water Flow Systems inNevada. Hyd. and Water Res. Pub.No. 4, Cen. Wat. Res. Research,Desert Research Inst.,Univ. NevadaSystem, Reno, 111 p.

Mitchell, Robert W., William H. Russell,and William R. Elliott, 1977, MexicanEyeless Characin Fishes, GenusAstyanax: Environment, Distribution,and Evolution. The Museum, TexasTech. Univ. Special Publications, no.

12, 89 p.Mixon, Robert B., 1963, Geology of the

Huizachal Redbeds, Ciudad Victoriaarea, southwestern Tamaulipas. inPeregrina Canyon and Sierra de ElAbra, Corpus Christi GeologicalSociety Annual Field Trip, CorpusChristi, Texas, p. 24–35.

Morris, Neal, 1976, The Otates MineArea, Sierra de El Abra, Tamaulipas,Mexico. Assoc. for Mexican CaveStudies Activities Letter, no. 4, p. 17–22.

Muir, J. M., 1936, Geology of theTampico Region. Memoir, Amer.Assoc. Petroleum Geologists, Tulsa,Oklahoma, 280 p.

Nigra, J. O., 1951, El Cretacico Medio deMeéxico, con especial referencia a lafacies de caliza arrecifal del Albiano-Cenomaniano en la Cenobahia deTampico-Tuxpan. Bol. Asoc. MexicanaGeólogos Petroleros, v. 3, nos. 3–4, p.107–175.

Palmer, Arthur N., 1969, A HydrologicStudy of the Indiana Karst. Ph.D.thesis, Dept. of Geology, IndianaUniversity, Bloomington, Indiana.

———, 1972, Dynamics of a sinkingstream system: Onesquethaw Cave.Nat..Speleo. Soc. Bull., v. 34, no. 3, p.89–110.

———, 1972, Origin of cave levels inMammoth Cave National Park. inMiotke, F. D,. and A. N. Palmer,Genetic Relationship Between Cavesand Landforms in the Mammoth CaveNational Park Area. TechnischenUniversität Hannover, Hannover, W.Germany, 69 p.

———, 1975, The origin of maze caves.Nat. Speleo. Soc. Bull., v. 37, no. 3, p.56–76.

Pinder, George F., and John F. Jones,1969, Determination of the ground-water component of peak dischargefrom the chemistry of total runoff.Water Resour. Res., v. 5, no. 2, p. 438–445.

Plummer, L. N., Jr., 1972, Rates ofMineral–Aqueous Solution Reactions.Unpubl. Ph.D. thesis, NorthwesternUniversity, Illinois.

Pohl, E. R., 1955, Vertical shafts inlimestone caves. Nat. Speleo. Soc.,Occ. Pap., no. 2, 24 p.

Raines, Terry W. (ed.), 1968, Sótano delas Go1ondrinas. Assoc. for MexicanCave Studies, Bull. 2, 20 p.

Rainwater, F. H., and L. L. Thatcher,1960, Methods for collection and

analysis of water samples. U.S. Geol.Survey Water-Supply Paper 1454, 301p.

Roglic, J., 1972, Historical review ofmorphologic concepts. p. 1–18., inKarst: Important Karst Regions of theNorthern Hemisphere, ed. M. Herakand V. T. Stringfield, ElsevierPublishing Company, Amsterdam,551 p.

Rose, Peter R., 1963, Comparison of typeEl Abra of Mexico with “EdwardsReef Trend” of south-central Texas. inPeregrina Canyon and Sierra de ElAbra, Corpus Christi GeologicalSociety Annual Field Trip, CorpusChristi, Texas, p. 57–64.

Russell, William H., 1968, RegionalGeology. p. 6–15, in Sótano de lasGolondrinas, Assoc. for MexicanCave Studies, Bull. 2, ed. T. W.Raines, 20 p.

———, 1972, Alphabetical listing ofcaves of the Sierra de El Abra: El Abralist number one, July 1, 1972. Assoc.for Mexican Cave Studies Newsletter,v. 3, no. 6, p. 129–132.

Russell, William H., and Terry W. Raines,1967, Caves of the Inter-AmericanHighway, Nuevo Laredo, Tamaulipas,to Tamazunchale, San Luis Potosí.Assoc. Mexican Cave Studies, Bull. 1,126 p.

Searcy, James K., 1959, Flow-durationcurves, manual of hydrology: Part 2.Low-Flow techniques. U.S. Geol.Survey Water-Supply Paper 1542-A,33 p.

Sectretaría de Recursos Hidráulicos, (nodate), Datos Hidrometricos Cuencade] Río Panuco. Bol. Hidrológico No.32, Tomos I & II, (data through Dec.1968).

———, 1962, Datos de la Región delBajo Panuco. Bol. Hidrológico No.19, 288 p.

Segerstrom, Kenneth, 1961, Geology ofthe Bernal-Jalpan Area, Estado deQuerétaro, México. U.S. Geol. SurveyBull. 1104-B, p. 19–86.

———, 1962, Geology of south-centralHidalgo and northeastern México,Mexico. U.S. Geol. Survey Bull.1104-C, 162 p.

Shuster, Evan T., and William B. White,1971, Seasonal fluctuations in thechemistry of limestone springs: apossible means for characterizingcarbonate aquifers. Jour. Hydrology, v.14, p. 93–128.

Stephens, John Lloyd, 1842, Incidents of

185


Travel in Yucatan, 2 vols. ed. V. W.von Hagen, University of OklahomaPress, Norman, 349 & 347 p.

Stumm, Werner, and James J. Morgan,1970, Aquatic Chemistry: An Intro-duction Emphasizing ChemicalEquilibria in Natural Waters.Wiley-Interscience, New York, N.Y.,583 p.

Thompson, Peter, 1973, Speleochrono-logy.and Late Pleistocene ClimatesInferred from O, C, H, U and ThIsotopic Abundances in Speleothems.Ph.D. Thesis, Dept. of Geology,McMaster University, Hamilton,Ontario, 352 p.

Thrailkill, John, 1968, Chemical andhydrologic factors in the excavation oflimestone caves. Geol. Soc. Am. Bull.,v. 79, p. 19–46.

———, 1972, Carbonate chemistry ofaquifer and stream water in Kentucky.Jour. Hydrology, v. 16, p. 93–104.

Tóth, J., 1963, A theoretical analysis ofgroundwater flow in small drainage

basins. Jour. Geoph. Res., v. 68, no.16, p. 4795–4812.

———, 1972, Properties and manifesta-tions of regional groundwatermovement. 24th InternationalGeological Congress, Section II,Hydrogeology, p. 153–163.

Trainer, Frank W., and Ralph C. Heath,1976, Bicarbonate content of ground-water in carbonate rock in easternNorth America. Jour. Hydrology, v. 31,p. 37–55.

Viniegra, O. F., and C. Castillo-Tejero,1970, Golden Lane fields, Veracruz,Mexico. in Geology of Giant Petro-leum Fields, ed. M. T. Halbouty, Am.Assoc. Petroleum Geologists Mem.14, p. 309–325.

Walsh, M., 1972, Mexican Caving, 1966–1971. Southwest Texas Grotto, SanMarcos, Texas, 146 p.

Waring, Gerald A. (revised by ReginaldR. Blankenship and Ray Bentall),1965, Thermal Springs of the UnitedStates and Other Countries of the

World—A Summary. U.S. Geol.Survey Prof. Paper 492, 383 p.

White, W. B., 1969, Conceptual modelsfor carbonate aquifers. Ground Water,v. 7, p. 15–21.

White, W. B., and J. Longyear, 1962,Some limitations on speleo-geneticspeculation imposed by the hydraulicsof groundwater flow in limestone.Nittany Grotto Newsletter, v. 10, no. 9,p. 155–167.

Wigley, T. M. L.,1972, A computerprogramme for water quality analysis.Tech. Note 15, Dept. of Mech. Eng.,Univ. of Waterloo.

Wigley, T. M. L., 1973, Chemicalevolution of the system Calcite-Gypsum-Water. Can. J. Earth Science,v. 10, no. 2, p. 306–315.

Wilford, G. E., 1964, The Geology ofSarawak and Sabah Caves. Bull. 6,Geological Survey, Borneo Region,Malaysia, 181 p.

186

Documents

KARST HYDROLOGY OF THE SIERRA DE EL ABRA, MEXICO3.2. Karst geomorphic features in the region 42 3.3 Summary 43 4. Hydrology and hydrogeology 45 4.1. Introduction 45 4.2. Temporal aspects