Origin of nodules in mixed siliciclastic-carbonate sandstones, the Lower Eocene Roda Sandstone Member, southern Pyrenees, Spain

Sedimentary Geology, 66 (1990) 277-293 277 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

Origin of nodules in mixed siliciclastic-carbonate sandstones, the Lower Eocene Roda Sandstone Member,

southern Pyrenees, Spain

N. MOLENAAR and A.W. MARTINIUS

Sedimentology Division, Institute of Earth Sciences, State University of Utrecht, Budapestlaan 4, Postbox 80.021, 3508 TA Utrecht (The Netherlands)

Received August 7, 1989; revised version accepted December 7, 1989

Abstract

Molenaar N. and Martirtius, A.W., 1990. Origin of nodules in mixed siliciclastic-carbonate sandstones, the Lower Eocene Roda Sandstone Member, southern Pyrenees. Sediment. Geol., 66: 277-293.

Calcite cement is heterogeneously distributed in the Eocene Roda Sandstone Member. Strongly cemented nodules and layers, intercalated within less well-lithified sandstones, are common. The nodules yield valuable information about the factors that controlled the onset and the degree of cementation. All nodules are associated with abandonment surfaces, formed when sedimentation ceased for a variable length of time. The nodules represent a first stage in the development of continuous hardgrounds. The initiation of early cementation was dependent on a combination of several parameters, such as grain-sorting and the presence and amount of carbonate nucleation centers. Early cementation began at distinct points, now manifested by nodules, aggrading into layers in cases where sedimentation had ceased for a sufficient length of time. The resulting variability in the amount of cement determines the present-day difference in weathering behaviour between nodules and the adjacent, less well-cemented sandstones.

Introduction

Nodula r features in carbonate and fine-grained siliciclastic sediments have received much atten- tion. Most nodules are richer in carbonate than

the surrounding sediments. Some authors con-

eluded that nodules in carbonate formed late in diagenesis (e.g., Bjorlykke, 1973), a l though an early

diagenetic origin has been put forward more fre-

quently (e.g., Jenkyns, 1974; Mullins et al., 1980;

M~Sller and Kvingan, 1988). In siliciclastic pelites,

early diagenetic processes related to decomposition of organic material are thought to be the cause of the development of nodules or concre-

tions (e.g., Curtis and Coleman, 1986). Concre- tions in siliciclastic pelites may grow during early diagenesis (Weeks, 1957; Rahmani , 1970), but also

during burial and mechanical compact ion (Rais- well, 1971).

So far, less a t tent ion has been paid to the

occurrence of carbonate-r ich nodules and lenses in

intermediate- and coarse-grained sandstones (e.g., Garr ison et al., 1969; Rahmani , 1970; Roberts

and Whelan, 1975; Chafetz, 1979; Hudson and

Andrews, 1987; Pirrie, 1987; McBride, 1988). With

exception of Hudson and Andrews (1987), these

authors found evidence for a very early format ion

of nodules in sandstones.

The Lower Eocene R o d a Sandstone Member is an example of a sandstone displaying nodular layers and horizons with isolated nodules (Fig. 1),

indicating locally inhomogeneous lithification. This inhomogeneous distr ibution of carbonate ce-

ment was studied in order to better unders tand

0037-0738/90/$03.50 © 1990 Elsevier Science Publishers B.V.

278 N. MOLENAAR AND A.W. MARTINIUS

Fig. 1. Outcrops of the Roda Sandstone Member showing nodular features. The photograph shows a well-lithified layer with intern

nodular structures that are even more resistant to weathering.

the causes of sandstone cementation. An irregular

distribution of cement, such as displayed by ce-

mented nodules, lenses or horizons alternating with

non- or less-cemented horizons, can provide infor-

mation about the factors controlling cementation.

It is likely that sediment parameters and/or en-

vironmental conditions favoured cementation at

distinct spots, now represented by nodules, or

inhibited the development of cement elsewhere. If

such inhomogeneities are visible in outcrops by

differential weathering patterns, it is possible to

accurately sample the variably cemented parts of a

sandstone body. If nodules are indeed caused by

differences in diagenetic processes, their presence

must reflect the variability of local sedimentary

parameters, such as grain-sorting and the amount

of carbonate cement nuclei present.

GeologIcaI setting

The Tremp-Graus Basin forms part of the

southern Pyrenean foreland basin which was frag-

mented due to southward thrusting during the

Paleogene (Williams, 1985) (Fig. 2). Thrusting took

place contemporaneously with sedimentation in

the Tremp-Graus Basin (Ori and Friend, 1984).

UNDIFFERENTIATED $

I z

jz?J &

HERCYNIAN BASEMENT

n AREA STUDIED

Fig. 2. Location and general geologic setting of the Tremp-

Graus Basin in Spain. Sample location are along the river

Ishbena between Roda d’Iskbena and Serraduy.

NODULES IN MIXED SILICICLASTIC-CARBONATE SANDSTONES 279

MEGA-RIPPLE CROSS-BEDDING

v NODULES 0

k4 A 0 A LARGE-SCALE

. .._.. __.-. -

* -

IWiH-ANCiLt

CROSS -BEDDING

VI 0 0

MEGA-RIPPLE CROSS-BEDDING

=a q NODULES 0

LARGE-SCALE LOW-ANGLE CROSS-BEDDING WmH INTERNAL MEGA-RIPPLES

~~ MEGA_R,PPLE

CROSS-BEDDING

fi HIGH-ANGLE ; CROSS-BEDDING

w 0

NODULES IARGE-SCALE LOW-ANGLE CROSS-BEDDING

0 WITH

NODULES A INTERNAL MEGA-RIPPLES

m CALCAREOUS SANDSTONE m SANDSTONE m SILTY-SANDY MARL B NODULES

v INTENSIVE BIOTURSATION q HARDGROUND A SUB-HARDGROUND 0 NON-HARDGROUND

Fig. 3. Generalized lithologic column of the Roda Sandstone Member near Roda d’kkbena.

280 N. M O L E N A A R A N D A.W. M A R T I N I U S

This tectonic movement probably triggered major influxes of clastic sediments into the basin, includ- ing those which formed the Roda Sandstone Member. The Lower Eocene Roda Sandstone Member consists of a sequence of tide-dominated shallow-marine mixed siliciclastic-carbonate sandstones separated by sandy or silty marls (Nij- man and Nio, 1975). The basal part consists of silty to sandy marls with thin sandy limestone layers. The upper part has a thickness between 40 and 65 m and contains up to five vertically stacked sandstone bodies separated by silty and sandy marls and siltstones (Fig. 3). These sandstone bodies, with an individual thickness of up to 30 m, are composed of cosets of large-scale low- and high- angle cross-beds. Deposition occurred from south- west to northwestward prograding ebb-tidal delta lobes, transverse tidal bars, and lateral accretion along meandering ebb-tidal channels (Nio and Yang, 1983). The different depositional units are separated by major abandonment surfaces, typi- fied by burrowed bioclastic and matrix containing arenites or wackestones. This indicates that sedimentation was strongly intermittent due to inter- ruptions in the progradation of delta lobes and tidal bars and to the eventual total abandonment of the depositional area. The upper part of depositional lobes may have been cut and transformed into regular planar surfaces by wave action.

Methodology

The texture, composition and diagenetic features of 60 samples of nodules and 129 samples of host rocks from eight sections have been qualita- tively determined by means of thin-section mi- croscopy and quantified by pointcounting.

The grain-size distributions of the siliciclastic components of 61 samples, separated from the sediment through hydrochloric acid treatment, were measured in duplicate using a Malvern 3600D laser particle sizer.

Oriented glass-mounted preparations of the acetic acid insoluble residue less than 10/~m fraction of eight samples were analyzed for clay minerals by standard X-ray diffraction (Cu-K, radiation and a Ni-filter).

Early diagenesis: hardground formation

In outcrop, three main types of sandstone can be recognized based on differences in composition and resistance to weathering. Several, laterally continuous, horizons are bioclast- and matrix-rich, well-lithified calcareous sandstones. All nodules as well as occasionally large-scale cross-beds are also weU-lithified and resistant to weathering. The major part of the sandstone appears as friable and darker coloured layers, having a greater amount of peloids and dark or opaque clasts.

During periods of low or no sedimentation, texture and composition of the sediments were modified by several processes. The intensity of modification is related to the length of the abandonment periods.

Firstly non-biogenic processes were active. A matrix, mainly consisting of carbonate, was introduced into the sediment through mechanical (hydrodynamic) infiltration. Matrix infiltration re- suited in the partial geopetal filling of pores and a bimodal grain-size distribution of the sandstone. Matrix infiltration was a rapid process, which could only continue due to mixing of the sediment by burrowing, since it easily blocked the pore connections for further infiltration. Most probably, the matrix was introduced into the sands by pumping action of the tides, which had a range of about 4 m (Yang and Nio, 1985).

After matrix infiltration, a fauna populated the barren surface of the sands and burrowing was initiated. The surface was colonized by a partially infaunal association of benthonic foraminifers, echinoids and gastropods. This resulted in bioturbation of the sediment, and the production and accumulation of bioclasts. Through bioturbation, the primary alignment of clasts parallel to lamination was disturbed, laminae were mixed, and carbonate matrix material and bioclasts were mixed with the primary sediment. Matrix introduced by bioturbation displays a irregular distribution but, where present, completely fills interstitial pores. Here, the framework of the sandstone commonly is matrix-supported, and the grain-size distribution of framework components often is polymodal.

In addition, a fringe-cement precipitated around


Fig. 4. Photomicrograph of a thin section of a sub-hardground. The sandstone has a clast-supported framework without compactional features. Mechanically infiltrated matrix (IM), mainly consisting of carbonate, occurs on top of and between clasts. Matrix infiltration was followed by an abundant early diagenetic fringe-cementation (RC), which borders infiltrated matrix and carbonate clasts (out of the picture) and in places also siliciclastic grains (Q = quartz grain). A second generation of cement, consisting of sparry low-magnesian calcite, filled the residual pores. Carbonate clasts, matrix and fringe-cement were replaced by low-magnesian calcite contemporaneous with this second cementation phase. In this sample the fringe-cement has been replaced pseudomorphously by low-magnesian calcite. The arrow indicates the stratigraphic facing. The scale

bar is 0.5 mm; nicols are partly crossed.

clasts, especially around carbonate clasts. This resulted in the formation of hardground-like layers. Fringe-cementation appears to be a later process than matrix infiltration (Fig. 4). Apart from syn- sedimentary escape burrows, fringe-cementation seems to be an earlier process than burrowing.

Overall, a siliciclastic sand was changed into a cemented matrix- and bioclast-rich arenite or eventual wackestone, with poor sorting due to the

cumulative effect of burrowing, matrix addition and introduction of (large) bioclasts (Molenaar et al., 1988).

Based on the extent of early modifications, a distinction was made between major abandonment periods (resulting in hardgrounds) and minor abandonment periods (resulting in sub-hardgrounds) (Molenaar et al., 1988). Sub-hardgrounds are an initial stage of hardground development and occur as individual layers, or constitute the lower parts of hardgrounds. They formed when the pauses in sedimentation were of relatively short duration. Sub-hardgrounds and hardgrounds are at most 0.6 m thick, which is also the maximum thickness of the nodules.

Cementation in the Roda Sandstone Member occurred in two distinct phases. An early, pre- burial fringe-cement formed exclusively below abandonment surfaces (Molenaar et al., 1988). Because of the fibrous morphology of crystals, an aragonite mineralogy can be assumed for fringe- cement. A late calcite cementation occurred after burial and mechanical compaction, simultaneously with replacement of aragonite and high-magnesian calcite components. Apart from the near-surface early cement, which originated near or at the sediment-water interface, an inhomogeneous spa- tial distribution of this early cement might be expected as well, because flow of interstitial water was restricted to the sediment-water interface and must have been influenced by slight lateral differences in sediment characteristics. Evidence for such fluid flow is given by the mechanically infiltrated matrix. Moreover, frequently insufficient time was available for complete cementation, since pauses in the sedimentation were of limited duration in the shallow-marine environment.

Fringe-cement is commonly found in association with infiltrated matrix, which is only present below former abandonment surfaces. Moreover, compactional features are absent in sandstones with fringe-cement. Both factors indicate that fringe-cement must have precipitated at or very close to the sediment-water interface. Submarine erosion could have removed a soft sand layer on top of a concealed (partially) lithified incipient hardground layer. The top of some major hardgrounds have been truncated. However, erosional

282 Y M O L E N A A R A N D A.W. M A R T I N I U S

contacts between sub-hardgrounds and overlying sediments have not been observed.

Later diagenesis

In samples without the fringe-cement stabilizing the framework, mechanical compaction re- duced the porosity from approximately 38% to about 21%, as calculated from percentage of cement and matrix (Table 1). Burial depth attained a maximum of approximately 900 m. Mechanical compaction was effective in reducing the primary porosity because of a high percentage of ductile grains, such as peloids. Most of the porosity reduction then was accomplished by plastic deformation of these soft grains between rigid framework components. A second generation of cement precipitated after compaction. It consists of blocky, more or less equidimensional low-magnesian calcite, ranging in crystal size from 20 to 1000 /zm. Larger crystal sizes occur in non-hardgrounds, where the cement is sub-poikilotopic. The percentage of cement is inversely related to the degree of compaction except in case of hardgrounds that contain large amounts of matrix. Nodules contain a high percentage of calcite cement (29.5% in nodules, as contrasted with 27.2% in sub-hardgrounds and 17.0% in the heavily compacted non- hardgrounds; Table 1).

At present, low-magnesian calcite is the only carbonate mineral present. Later diagenesis resulted in neomorphic replacement by low-magnesian calcite of all former aragonite and high-mag-

TABLE 2

Stable carbon isotope compositions of nodules and host rocks. Asterisks mark mean carbon isotope values excluding the two extreme values (-0.59%o and - 4 . 7 0 ~ , respectively). Dif- ferences are not statistically significant

~13C Nodules Host rocks

M e a n ( q ) - 1 . 1 8 ( - 1 . 2 4 *) - 1 . 5 1 ( - 1 . 3 8 *) Range (%0) - 0.59 to - 1.54 - 1.00 to - 4.70 St. dev. (%o) 0.25 0.74

Number of analyses 10 25

nesian calcite, which primary were present in bioclasts, matrix and fringe-cement. Replacement of fringe-cement was often pseudomorphous, whereas the primary textures of bioclasts were destructed. The pervasive replacement destroyed all chemical and isotopic evidence about the primary and early diagenetic carbonate components. The stable carbon isotopic compositions of cement and total carbonate fractions are similar throughout the sandstone bodies (Table 2).

Form and location of nodules

Nodules are a common feature in outcrops of the Roda Sandstone Member. They occur in all types of sandstone, but are most common in the non-hardgrounds. Nearly all nodules display features typical for sub-hardground and hardgrounds such as infiltrated or burrowed matrix, fringe-ce-

TABLE 1

Mean percentages and standard deviations between brackets of clastic components and diagenetic constituents obtained by

pointcounting 250-400 points per thin-section

Constituents Nodules Host rocks

non-hardgrounds sub-hardgrounds hardgrounds

Matrix (%) 8.0 (7.8) 3.8 (9.1)

Fringe-cement (%) 7.8 (3.7) 0.0 Sparite cement (%)

(second generation of cement) 21.7(8.6) 17.0 (6.9) Bioclasts (%) 9.6 (7.0) 5.5 (3.2) Total carbonate clasts (%) 25.8 (6.9) 23.9 (8.7) Total clasts (%) 62.0 (6.1) 79.2 (10.5) Total cement and matrix (%) 37.3 (5.5) 20.8 (10.5)

N umber of analyses 34 41

4.9 (4.7) 21.0 (12.1) 2.2 (1.5) 3.9 (3.6)

25.0 (6.8) 10.9 (8.9) 9.8 (1.8) 20.3 (13.1)

29.0 (11.0) 27.4 (14.1) 63.6 (17.8) 64.2 (8.4) 29.0 (11.0) 35.8 (8.5)

13 21


TABLE 3

The distribution of nodules in the various types of host rocks and the percentage of the total amount of sampled nodules. Note that nodules occurring in a non-hardground host rock represent the most abundant type

Type Nodule Host rock N a Percentage Appearance characteristics characteristics

I Hardground sub-hardground 1 3.2 ) nodules with 16.1 diffuse outlines

non-hardground 4 12.9 nodules with distinct boundaries

II Sub-hardground hardground 2 6.5 ) nodular layers

sub-hardground 2 6.5 / 80.7 nodules with diffuse outlines

non-hardground 21 67.7

III Non-hardground non-hardground 1 3.2 nodules with distinct boundaries

a Number of analysis.

Fig. 5. Nodules occur in a horizon which is parallel to the coset boundaries. The individual nodules are lensoid and parallel to the mega-cross-bedding. The layer has a thickness of 75 cm.

2 8 4 N. MOLENAAR AND A.W. MARTINIUS

ment and the absence of compactional features, indicative for early modifying processes related to pauses in sedimentation. Only one sampled nodule did not display any of the characteristic features typical for sub-hardgrounds or hardgrounds. The main distinctive features are the percentage of fringe-cement and the total amount of cement. Two major kinds of nodules can be distinguished (Table 3 ):

--Type I ( 5-16% of the sampled nodules): nodules displaying hardground characteristics (burrowed and/or infiltrated matrix, increased bioclast content and fringe cement), occurring in sub- hardground or non-hardground intervals;

--Type II (+81%): nodules displaying sub- hardground characteristics (i.e., infiltrated matrix and fringe-cement), mostly occurring in non- hardground intervals ( + 68%).

Zones of nodules are parallel to layering, the large-scale low- and high-angle foresets (Fig. 5), and to abandonment surfaces (Fig. 1). The nodules in sub-hardgrounds and non-hardgrounds display distinct outlines and have a flattened ovoidal or more irregular form, with their long axes, 10 up to 180 cm in length, parallel to the bedding. The thickness of nodules does not exceed that of the layers in which they occur. The mean ratio between the thicknesses of nodules and host rock is 0.5 + 0.2. Most nodules in hardgrounds have an irregular form. They have diffuse outlines and grade into the surrounding host rocks. The pertinent intervals are better described as nodular layers. Because of their well-lithified nature, nodules are usually more resistant to weathering than the surrounding sandstones, especially nodules in sub-hardgrounds and non-hardgrounds. Nodules may lie detached on exposed and weathered surfaces.

C ~ of nodules mH! host rocks

Siliciclastics

The primary elastics of the Roda Sandstone Member are largely s'diciclastic grains and minor extrabasinal dark grey carbonate fragments. Rela- tive amounts and ranges of variations of various

major types of siliciclastic components are similar for nodules and host rocks (Fig. 6).

Bioclasts

Bioclasts and intrabasinal light-coloured carbonate grains are especially associated with abandonment surfaces. The primary content of carbonate bioclasts is approximately 5.5_ 3.2% (i.e. the content in non-hardgrounds which remained unaffected by hardground processes). For a large part, they were introduced into the sediment postdepositionally, as indicated by their re- lationship with burrows and burrowed sediments. The total content of bioclasts increased during hardground formation (Table 4). The content of bioclasts in nodules and sub-hardgrounds is significantly higher (9.6 + 7.0 and 9.8 5- 1.8%, respectively) than in non-hardgrounds. The content in hardgrounds is even higher (20.3 +_ 13.1%). This confirms the continuation of bioclast addition through bioturbation. In nodules, 41.6% of the grains are carbonates, whereas in non-hardgrounds only 30.2% of the grains are carbonates. There is no evidence for local dissolution of bioclasts or other carbonate clasts.

The assumed primary mineralogical composition of the bioclasts in the various types of sandstones and nodules is shown in Table 4. Gastro- pods and part of the bivalves were probably composed of aragonite; alveolinids, miliolinids and echinoids of high-magnesian calcite; and nummulitids, assilinids, discoeyclinids and another part of the bivalves of low-magnesian calcite. The aragonite bioclast content is similar throughout all types of sandstone. Nodules contain statistically significantly more low-magnesian bioclasts than non-hardgrounds and sub-hardgrounds. However, there is seemingly no consistent trend with respect to the mineralogy of the bioclasts.

Primary grain-size distribution

Results of grain-size analyses are summarized in Table 5. The fraction smaller than 10 /~m, derived from clayey material in matrix and peloids, consists of quartz, feldspars, smectite, illite and kaolinite.

N O D U L E S I N M I X E D S I L I C I C L A S T I C - C A R B O N A T E S A N D S T O N E S 2 8 5

QUARTZ

I'i °s '.ocK I

SUBARI

[]

90%

SUBLITH NITE

75%

N=67

FELDSPARS

o oo

o

: /

0 []

[]

o

o

.-d

(QUARTZOSE)

1:3 1 :1 3:1

NODULES HOST ROCKS

QUARTZ 68.3 (9.5)

LITHICS

67.7 (8.5)

6.2 (3.5) UTHICS 6.2 (2.9)

FELDSPARS 25.5 (9.5) 26.1 (8.6)

NUMBER OF ANALYSES 13 54

Fig. 6. The siliciclastic mineralogical compositions of nodules and host rocks. Percentages were obtained by pointcounting 300-400 points of 67 thin-sections. Top: Diagram displaying compositions of nodules and host rocks with respect to primary silieiclastie grains in terms of percentages quartz (mono- and polycrystaUine-quartz), lithics (mixed siliciclastic-carbonate clasts, extrabasinal carbonate dasts, siltstone and sandstone dasts, biotite, chlorite and chert grains) and feldspars. The triangular diagram is constructed according to the modified classification of Folk (1968). Bottom: Mean and standard deviations of percentages of quartz, lithics and feldspars of nodules and host rocks. The relative amounts of the various terrigenous components is the same in both nodules and host

rocks. This indicates that the primary elastic supply remained constant during the deposition of the sandstones.

The general charac te r of the sil iciclastic grain-

size popu l a t i on of nodules and hos t rocks is s imi-

lar. Ear ly -cemented layers or nodules d i sp l ay a

mean grain-size range s imilar to the su r round ing

rocks (Fig. 7A). On ly the skewness and quar t i le

devia t ions of sil iciclastics in nodules and hos t

rocks are s ta t i s t ica l ly s igni f icant ly different . How-

ever, a c ompa r i son of samples f rom the same

sect ion shows subt le bu t s ta t is t ical ly s ignif icant

d i f ferences in the charac ter i s t ics of the g a i n - s i z e

d i s t r ibu t ions (Fig. 7B and C). The kurtosis , skew-

ness and sor t ing coeff ic ients ind ica te a sl ightly

286 N. MOLENAAR AND A.W. MARTINIUS

TABLE 4

Variations in the total amount of bioclasts and in the inferred original mineralogical composition of bioclasts subdivided into aragonite (gastropods and part of the bivalves), high-magnesian calcite (alveolinids, miliolinids and echinoids), and low-magnesian calcite (nummulitids, assilinids, discocyclinids and part of the bivalves). Mean values and standard deviations were calculated for a 100% bioclasts composition. Percentages were obtained by pointcounting 250-400 points per thin-section

Original mineralogy of bioclasts (%)

total % of aragonite high-Mg calcite low-Mg calcite bioclasts

Number of analyses

Nodules 8.2 (3.4) 7.1 (6.9) 53.6 (20.0) 39.3 (18.7) 21 Non-hardgrounds 5.5 (3.2) 5.7 (10.5) 67.6 (19.4) 26.7 (15.6) 19 Sub-hardgrounds 9.8 (1.8) 6.8 (6.5) 69.6 (5.8) 23.7 (6.0) 5 Hardgrounds 17.9 (10.8) 10.2 (19.3) 56.7 (23.1) 33.1 (18.2) 15

better sorted sand in the nodules than in the surrounding sandstones. Within nodules, the range of grain-sizes is narrower, i.e. the standard deviation is smaller, whereas the range is more symmet- rically distributed around the mean. Moreover, commonly the mean grain-size of nodules is slightly larger compared to the directly surrounding host rock (Fig. 7D).

The amount of infiltrated and burrowed matrix

Percentages of matrix, i.e. non-framework components smaller than 50 gm, are given in Table 1. The amounts of infiltrated and burrowed matrix in sub-hardgrounds and nodules are similar. In contrast, hardgrounds contain a significantly greater percentage of matrix. Large amounts of matrix in the hardgrounds are consistent with the hypothesis that burrowing and matrix infiltration continued for a considerable amount of time.

Nodules did not experience such intense bioturbation.

Discussion

The total amount of cement is the combined result of early diagenesis (fringe-cementation) and late burial diagenesis (mechanical compaction and post-compactional calcite cementation). However, merely the absence or presence and the amount of early stabilizing fringe-cement below abandonment surfaces, and thus the amount of open pore space left after mechanical compaction, determined the amount of late cement in the Roda Sandstone Member. The increased percentage of calcite cement of nodules with respect to the surrounding sandstone and the resulting well-lithified nature is undoubtedly the main reason for their enhanced resistance to weathering. The origin of nodules, which contain distinctly higher amounts

TABLE 5

Results of grain-size measurements by a laser particle sizer. Grain-size distribution parameters of nodules and host rocks are characterized by mean values and standard deviations

Grainsize parameters

Nodules Host rocks Reference

non-hardgrounds sub-hardgrounds hardgrounds

Median (~) 1.92 (0.52) 1.91 (0.61) 1.93 (0.58) 1.84 (0.45) Mean (~) 1.99 (0.54) 2.13 (0.68) 2.11 (0.52) 2.14 (0.51) Sorting (~,) 1.27 (0.31) 1.42 (0.26) 1.31 (0.07) 1.66 (0.48) Skewness (~) 0.22 (0.13) 0.35 (0.09) 0.38 (0.10) 0.37 (0.10) Kurtosis (~) 1.48 (0.23) 1.55 (0.22) 1.75 (0.25) 1.42 (0.39) Quartile deviation (,~) 0.67 (0.13) 0.73 (0.14) 0.63 (0.03) 0.93 (0.34)

Number of analyses 16 29 4 9

Trask (1930) Folk and Ward (1957) Folk and Ward (1957) Folk and Ward (1957) Folk and Ward (1957) Krumbein (1936)

NODULES IN MIXED SIL1CICLAST1C-CARBONATE SANDSTONES

aoq / , '

e 7°1 /, ' /,'

1 / ' : ? J , . . . . . . .

- 1 0 1 2 4 5 6 7 8 9

2 .3

2.1

1 .9

1.5

1.3

1.1

0.9 -0. :

I i I i I

NODULI

NON-NARDGROUND •

I 0.0

o • 0

• o 0

0 o 0 o

• • 0 0 0

ooo,~ o o

o • • o

o • •

o.'~ o.'4 S K E F N E S S

287

i

B

0.6

1 .30

1 .20

1.10

~1 .00

~ 0 . 9 0

~ O.aO

¢~ 0.70

0 .60

i i i i i i I i i i i I I

Io,oo , o NON-~ROUND [

0

o • o •

• o e l , 0 0 0 0

0 0 0 0 •

• 0 0 0

o~ 0 • 0

• 0 0 O 0 •

• ° ° o oO

L'5 'L'7 'L'o SORTING

g &

f o

C

2.3

0.76 ' ' ' O NODULE

0.74 HOST ROCX

0.72

0.70

~ 0 . 0 8

0 . 0 e

0 . 5 4

~--.~0.62 ~; aA ~ 0 . 6 0 c •

~" 0 . 5 8

b a 0 . 5 6 g •

0.54 f •

0 . 5 2 0 . 9 0 1.(}0

i i i i i i

g o

d o e& a o

c o bO

e o

d& D

0.50 o.9 L'I 1.'3 '2.'1 L lo 1.~.o L~o ' L,io

SORTING Fig. 7. Grain-size properties of nodules and host rocks as measured with a laser particle sizer. A. Mean cumulative frequencies of the grain-size distributions of nodules (N = 16) and host rocks (non-hardgrounds) (N ffi 29). Nodules are better sorted and have less fine-grained material (less positively skewed). The similarity of frequency distributions suggests one common siliciclastic source during deposition of these types of sandstone. B. Sorting coefficients (Folk and Ward, 1957) are plotted against skewnesses (Folk and Ward, 1957) for nodules and host rocks (non- hardgrotmds). The range of sorting values is similar, although nodules are better sorted and have a less positive skewness. C. Sorting coefficients (Folk and Ward, 1957) are plotted against quartile deviations (Krumbein, 1936) for nodules and host rocks. The range of values is similar. D. Sorting coefficients (Folk and Ward, 1957) are plotted against quartile deviations (Krumbein, 1936) for pairs of nodules and their direct host rocks. Nodules have a distinctly better sorted grain-size distribution as is evident from smaller sorting coefficients and lower quartile deviations, the latter with exception of

samples "e".

o f f r i n g e - c e m e n t t h a n the s u r r o u n d i n g hos t rocks ,

is thus l i n k e d to t he p roces s o f f r i n g e - c e m e n t a t i o n

d u r i n g pauses in t he s e d i m e n t a t i o n . L o c a l va r i a -

t ions in the f ac to r s c o n t r o l l i n g this f r i n g e - c e m e n -

t a t i on c a u s e d d i f f e r ences in i ts a m o u n t a n d t im-

ing. A " s p o t - l i k e " i n i t i a t i o n o f f r i n g e - c e m e n t a t i o n ,

a l t h o u g h a lways l i n k e d to a b a n d o n m e n t surfaces ,

c o u l d h a v e t a k e n p l a c e w h e r e v e r o n e o r m o r e

p a r t i c u l a r s e d i m e n t a r y p a r a m e t e r s were m o r e

f avou rab l e . P a r a m e t e r s w h i c h m i g h t h a v e con-


trolled the beginning of fringe-cementation are texture, mineralogical composition, permeability and porosity, degree of bioturbation and matrix content. One or a combination of these parameters must have controlled the initiation and eventual degree of early fringe-cementation, and thereby the total amount of compaction and cementation. Differences in composition and/or texture may be primary, related to depositional processes, or, more likely, secondary, and due to early postdepositionai modifying processes. Sever- al possibilities will be discussed.

Differences in primary sificiclastic composition

Local variations in hydrodynamic conditions cause differences in grain-size and grain-size distribution as well as variations in the mineralogical composition of sands. The availability of clastic grains in the various size fractions and their specific grain density consequently determine the composition of the sediment as a result of changes in the hydrodynamic conditions during deposition and grain (or mineral) sorting during transport.

The similarity in primary siliciclastic composition of nodules and host rocks throughout the Roda Sandstone Member suggests that this is not a factor significant for the formation of nodules.

Primary bioclastic composition

Where slight differences in the siliciclastic grain-size distributions between nodules and host rocks are the result of depositional processes, then these differences could have been associated with slight variations in the content of extrabasinal and primary intrabasinal carbonate grains, Most carbonate clasts in the Roda Sandstone Member consist of skeletons with large internal voids. Such bioclasts will have a similar grain density and fall velocity as a smaller quartz or feldspar clast. A result could be that carbonate grains were con- centrated in specific locations and/or levels of the sedimentary sequence. There they could have preferentially induced nucleation of cement without any substantial secondary enrichment of bioclasts during pauses in sedimentation. Such primary hy- draulic accumulations of bioclasts in clusters or

layers can cause localized or differential carbonate cementation (e.g., Chafetz, 1979; Kantorowicz et al., 1987; Bryant et al., 1988; McBride, 1988). Although these differences can not be totally ex- cluded, they do not seem to be very likely consid- ering the similar skewnesses of nodules and host rocks.

Postdepositionally introduced bioclasts

Apart from the availability of time needed for early fringe-cementation, a second restricting con- dition, is the presence of suitable nuclei for precipitation of cement. If nuclei are absent, the degree of supersaturation needed is too high for most natural environments. Most of the biogenic carbonate clasts in the Roda Sandstone Member were introduced into the sand during pauses in sedimentation. The amount of postdepositionally introduced carbonate clasts is approximately pro- portional to the time of non-deposition. Carbonate clasts were predominantly produced within the basin, or upon the pertinent abandonment surfaces. The main difference in the composition of various types of sandstones was caused by the postdepositional change in composition due to the introduction of bioclasts by burrowing. These carbonate bioclasts and also micritic intrabasinal carbonate clasts, such as peloids and aggregate clasts, were preferentially used as nucleation sites during early fringe-cementation. Siliciclastic grains were covered with fringe-cement only when fringe-cementation was very intense and the percentage of fringe-cement is very high. This probably occurred when the pause in sedimentation was relatively long. As the content of bioclasts increased in proportion to time of non-deposition, nuclei for fringe-cementation became more abundant. A locally enhanced bioclast content probably induced an increase in the degree of fringe- cementation.

Apart from the dttration of non-deposition, the amount of introduced carbonate clasts is dependent on the environmental factors which determine the development of an epi- and infauna association. Also the fauna association, and thus the mineralogy of the bioclastic components, may change depending on the environmental condi-


tions. The mineralogy, aragonite or high-magnesian calcite, determines the suitability of bioclast for nucleation purposes. Bioclasts can also act as nucleation sites for precipitation of the second cement generation. Moreover, an increased content of aragonite induced local supersaturation with respect to calcite during introduction of meteoric water and subsequent dissolution of aragonite. Replacement and second phase cementation is likely to have started on places with a high content of aragonite. Both the presence of more nuclei and the resulting local higher supersaturation explain the smaller crystal sizes of sparry calcite cement in sub-hardgrounds and hardgrounds as compared with non-hardgrounds.

Primary grain-size distribution and differences in permeabifity

Slight variations in the porosity and permeability, caused by differences of the grain-size distribution, are a possible explanation for local variations in the amount of early fringe-cement. Per- meability controlled the flow of pore water, and thus the cementing agents, through the sandstone. Both porosity and permeability are dependent on sorting and mean grain-size. Porosity is maximal in medium sand, and increases with increasing sorting. Permeability increases with the degree of sorting and increasing median grain-size (Beard and Weyl, 1973; Leeder and Park, 1986, fig. 3). Local or regular variations in grain-size and grain-size distributions might have generated zones with a maximum permeability, through which a preferential flow of interstitial water would have occurred. In such a case, early cementation would have been initiated in zones with a high porosity and permeability.

Primary differences in permeability in the Roda Sandstone Member can have been caused by the sedimentary depositional structures and sequences. Different sites on the large transverse or longitudinal bars were dominated by characteristic hydrodynamic conditions and, according to these conditions, they have different grain-size distributions (e.g., Pryor, 1972, 1973). For example, toe- sets of giant foresets contain sediment which has accumulated by suspension fall out as well as

TABLE 6

Calculated primary permeability of nodules and non- hardgrounds. For nodules a porosity of 37-38% was used for permeability calculation. This porosity was obtained from the pointcounted amount of infiltrated matrix, fringe-cement and late cement. For non-hardgrounds a slightly lower porosity value of 35-36% was used, inferred from median grain-size and sorting (Beard and Weyl, 1973; Leeder and Park, 1986, fig. 3)

Calculated primary permeability (darcies)

Van Baaren Berg (1970) (1979)

Nodules 6.3-7.3 5.9-6.6 Non-hardgrounds 4.2-4.8 2.3-2.7

Rat io 1.5 2 .4-2 .6

coarser gains deposited in avalanching foresets. As a result, the permeability should be slightly less in toe-sets than higher up in the same sets, where only traction transported grains accumulated. In the top part, individual foresets have a greater height because of a shallower depth (Nio and Yang, 1983) and consequently higher flow rates caused deposition of coarser-grained sand. Where preserved, the upper parts of large bed forms should be more suitable for the development of early cement. These persistent porosity and permeability trends in the depositional sequences probably constrained the flow of interstitial water during pauses in sedimentation and limited fringe-cementation to the uppermost parts of the sandstone bodies. Such a dependency between early cementation and primary textures has been demonstrated by James (1985).

The sand in nodules is better sorted and often slightly coarser than the sand laterally in the same horizon, suggesting that nodules developed at locations with a higher primary permeability (Table 6). Sub-hardgrounds and especially hardgrounds have the worse sorting. This agrees with the more intense bioturbation in sub-hardgrounds and hardground levels. Therefore, the nodules may represent parts of a horizon at or just below the sediment surface where a relatively high primary porosity and permeability could have favoured fringe-cementation if other boundary conditions were suitable. If clastic sedimentation ceased for a

290 N M O L E N A A R A N D A.W. M A R T I N I U S

long time, then burrowing and matrix infiltration prolonged in the host rocks, causing a worse sorting due to the addition of fine-grained material and especially the homogenizing effect of sediment bioturbation. The addition of fine-grained material to the sediments surrounding the nodules would have retarded precipitation of fringe-cement or could have provided nucleation sites. The consistent trend in sorting from non-hardgrounds to nodules and eventually to hardgrounds suggests that the differences in grain-size distributions between nodules and surrounding rocks are a primary property.

In some cases nodules are arranged in horizons at the top of the sets, with the long axes of individual nodules aligned according to large-scale foresets (Fig. 5). The form of a nodule may be determined by directional differences in permeability (see also Pirrie, 1987). Directional an- isotropy can be the result of layering, lamination, primary alignment of elongate or platy clasts. An elongate oblate form has been observed in sub- hardground nodules, where part of the primary clast orientation parallel to the cross-bedding have been preserved. Nodules with hardground characteristics do not show any regular form. This probably is a consequence of intensive bioturbation that disturbed the primary clast orientation and lamination completely and/or of the irregular nature of the burrowed zone.

Differences in the amount of matrix

Matrix reduces the porosity, eventually blocks the connections between pores and thus decreases the permeability. The presence of infiltrated or burrowed matrix should thus increase the time needed for the development of fringe-cement, since matrix infiltration was an earlier process than fringe-cementation. On the other hand, since the matrix consisted primarily of carbonate micrite, it also amplified the suitability for fringe-cementation by enlarging the availability of precipitation nuclei and therefore created preferential sites of fringe-cementation. The increased amount of matrix in nodules (8.0_+ 7.8%; N = 34) apparently did not inhibit fringe-cementation. Therefore, the

possible lower permeability was compensated for by the larger amount of nuclei, which seems to be a more important factor in controlling fringe-cementation. The reduction of the primary porosity with 8% is balanced by the initial higher permeability of nodules.

Variations in the content of organic material

Organic material might locally have accumulated due to depositional processes and/or bio- logical activity, such as suspension feeding activity. Degradation of organic material can cause increased CO 2 concentrations (Berner, 1981). Bio- logical reactions, which use organic matter for their metabolism, are important pH-modifying agents. It is, however, questionable if, under nor- mal marine conditions and concentrations of organic matter, dissolution or precipitation of CaCO 3 can be induced. Early carbonate cement may possibly precipitate locally due to oxidation of organic matter (e.g., Kocurko, 1986) or to anaerobic methane oxidation (e.g. Roberts and Whelan, 1975; Raiswell, 1987). Usually, the latter process is accompanied in the marine environment by the authigenesis of pyrite due to sulphate reduction (Berner, 1981; Raiswell, 1987). Pyrite has not been observed in the Roda Sandstone Mem- ber. However, oxidizing conditions are consistent with the high primary permeability of these sands, excluding any authigenesis of pyrite.

If any carbon derived from fractionated organic material is incorporated into a carbonate cement, then this cement should have a light carbon isotopic composition. Carbon isotopic compositions of cement and total carbonate fraction of host rocks and nodules are not significantly different in the Roda Sandstone Member (Table 2). However, present isotope compositions are not a good mea- sure of primary composition because of the almost total replacement of primary and early diagenetic carbonate components by a later diagenetic low- magnesian calcite. The replacement was pervasive and probably destroyed any evidence of initial anomalous carbon isotope composition (Molenaar, 1990).


Origin of the nodules

Many nodules in the Roda Sandstone Member occur within sub-hardgrounds or hardgrounds and all are associated with the penecontemporaneous processes which were active converted the sediment below abandonment surfaces into hardgrounds (Table 3). Nodules resulted from relatively high amounts of fringe-cement that pre- vented compaction, and, as a consequence, the large amount of second generation calcite cement. Nodules were thus initiated during very early diagenesis. Initial primary differences in the sediment have been accentuated during later diagenesis, by relatively less compaction and greater secondary cementation than the host rocks, and during recent weathering.

Beginning stage of hardground formation. Fringe- cementation began at separate sites, representing possible future nodules, depending on the availability of suitable nucleation sites for precipitation and/or on local differences in porosity and permeability. The latter governed the flushing of the sediment with water and thus regulated the supply of chemicals for fringe-cementation. Nucleation of fringe-cement usually could only take place where carbonate nucleation centres were present. Nod- ules contain fringe-cement, and some contain infiltrated matrix. Matrix infiltration and fringe-cementation were earlier processes than massive bioturbation. After a short phase of matrix infiltration, fringe-cementation began and blocked the pores for further infiltration. The contrast between nodules and host sandstone is at a maxi- m u m .

Intermediate stage. When a pause in sedimentation prolonged, individual nodules aggraded into continuous layers (i.e. sub-hardgrounds). Layers developed through continuing fringe-cementation and burrowing. The layers contain fringe-cement and have a slightly increased content of bioclasts with respect to the original sand and contain some infiltrated matrix. This is the sub-hardground stage. Nodules occasionally may still remain as discrete entities within these layers.

Final stage. If a pause in sedimentation was long enough, the barren surface became inhabited by a fauna that burrowed the surface layer, especially where fringe-cementation was weak or when fringe-cement was still absent. Continuing bioturbation totally disrupted the original texture and structure of the sediment, which became increas- ingly enriched in bioclasts and matrix. Where cementation was locally more abundant, this re- suited in laterally continuous nodular hardgrounds. The differentiation between nodules and host sandstone is slight.

Present situation. The intensity of deformation of non-hardgrounds is different from that of sub- hardgrounds and hardgrounds. The framework of the latter was at most slightly elastically deformed. In contrast, deformation of the non-hardgrounds was accomplished by plastic deformation (squeez- ing) of the ductile clasts, such as carbonate intra- clasts, during burial. The second phase of cementation was completed first in the nodules, followed slightly later by cementation of the surrounding sub-hardgrounds or hardgrounds. Non-hardgrounds were cemented later. This caused differences in the amount of plastic and elastic deformation. As a result of recent decompaction, dilatation cracks developed between the different types of sandstone. The various parameters determining the kind and amount of deformation during burial compaction vary little and are laterally gradual within any given sandstone horizon. At a certain point, critical with respect to elastic behaviour, dilatation cracks developed along planes separat- ing well-lithified sub-hardgrounds and hardground intervals and compacted non-hardgrounds. The sharp boundaries of nodules and layers are accentuated by recent weathering, since dilatation cracks are preferential infiltration paths for meteoric water. The rounded forms of nodules are due to enhanced weathering along intersections between cracks parallel and perpendicular to abandonment surfaces.

Furthermore, differential behaviour during weathering is also caused by slight differences in the crystal size of the second generation of cement. In sub-hardgrounds, hardgrounds and nodules the cement has a smaller crystal size and form


a firm interlocking lithifying agent. However, the sub-poikilotopic cement is not very coherent in the non-hardgrounds. In addition, present-day contact with meteoric water causes renewed swell- ing of the smectite clay.

Summary

Nodular features originated through slight differences in sorting of the sediments and especially through differences in the amount of carbonate nuclei associated with abandonment surfaces. Subtle variations in the timing and degree of early fringe-cementation caused differences in the amount of mechanical compaction and, therefore, in the amount of post-compactional sparry calcite cement. The nodules can be easily recognized as a result of the variable susceptibility to present-day weathering.

Acknowledgements

P.L. de Boer, S.D. Nio and H. Macdonald are thanked for their stimulating reviews of the manuscript. J.M.M. Reith and T. Zalm (Labora- tory of the Sedimentology Department) are acknowledged for performing the grain-size analyses, using equipment made available by "Stichting voor Technische Wetenschappen", and the clay mineral analyses, respectively. The Ko- ninklijke/Shell Exploration and Production Laboratory (Rijswijk) is kindly thanked for fi- nancing part of the sampling in 1986.

References

Beard, D.C. and Weyl, P.K., 1973. Influence of texture on porosity and permeability of unconsolidated sand. Bull. Am. Assoc. Pet. Geol., 14: 1-24.

Berg, R.R., 1970. Method for determining permeability from reservoir rock properties. Trans., Gulf Coast Assoc. Geol. Soc., XX: 303-317.

Berner, R.A., 1981. Authigenic mineral formation resulting from organic matter decomposition in modern sediments. Fortschr. Mineral., 59: 117-135.

Bjedykke, K., 1973. Origin of limestone nodules in the lower Palaeozoic of the Oslo Region. Nor. Geol. Tidsskr., 53: 419-431.

Bryant, I.D., Kantorowicz, J.D. and Love, C.F., 1988. The origin and recognition of laterally continuous carbonate-cemented horizons in the Upper Lias Sands of southern England. Mar. Pet. Geol., 5: 108-133.

Chafetz, H.S., 1979. Petrology of carbonate nodules from a Cambrian tidal inlet accumulation, Central Texas. J. Sedi- ment. Petrol., 49: 215-222.

Curtis, C.D. and Coleman, M.L., 1986. Controls on the precipitation of early diagenetic calcite, dolomite and siderite concretions in complex depositional sequences. In: D.L. Gautier (Editor), Roles of Organic Matter in Sediment Diagenesis. Soc. Econ. Paleontol. Mineral., Spec. Publ., 38: 23-33.

Folk, R.L., 1968. Petrology of Sedimentary Rocks. Hemphilrs, Austin, Texas, 170 pp.

Folk, R.L. and Ward, W.C., 1957. Brazos River bar: a study in the significance of grain-size parameters. J. Sediment. Pet- rol., 27: 3-27.

Garrison, R.E., Luternauer, J.L., Grill, E.V., Macdonald, R.D. and Murray, J.W., 1969. Early diagenetic cementation of recent sands, Fraser River delta, British Columbia. Sedi- mentology, 12: 27-46.

Hudson, J.D. and Andrews, J.E., 1987. The diagenesis of the Great Estuarine Group, Middle Jurassic, Inner Hebrides, Scotland. In: J.D. Marshall (Editor), Diagenesis of Sedi- mentary Sequences. Geol. Soc. London, Spec. Publ., 36: 259-276.

James, W.C. 1985. Early diagenesis, Atherton Formation (Quaternary): a guide for understanding early cement distribution and grain modifications in nonmarine deposits. J. Sediment. Petrol., 55: 135-146.

Jenkyns, H.C., 1974. Origin of red nodular limestones (Am- monitico Rosso, Knollenkalke) in the Mediterranean Jurassic: a diagenetic model. In: K.J. Hsii and H.C. Jenkyns (Editors), Pelagic Sediments: On Land and under the Sea. Int. Assoc. Sedimentol., Spec. Publ., 1: 249-271.

Kantorowicz, J.D., Bryant, I.D. and Dawans, J.M., 1987. Con- trois on the geometry and distribution of carbonate cements in Jurassic sandstones: Bridport Sands, southern England and Viking Group, Troll Field, Norway. In: J.D. Marshall (Editor), Diagenesis of Sedimentary Sequences. Geol. Soc. London, Spec. Publ., 36: 103-118.

Kocurko, M.J., 1986. Interaction of organic matter and crys- tallization of high-magnesian calcite, South Louisiana. In: D.L. Gautier (Editor), Roles of Organic Matter in Sedi- ment Diagenesis. Soc. Econ. Paleontol. Mineral., Spec. Publ., 38: 13-21.

Krumbein, W.C., 1936. The use of quartile measures in de- scribing and comparing sediments. Am. J. Sci., 32: 98-111.

Leeder, F. and Park, W.C., 1986. Porosity reduction in sandstone by quartz overgrowth. Bull. Am. Assoc. Pet. Geol., 70: 1713-1728.

McBride, E.F., 1988. Contrasting diagenetic histories of concretions and host rock, Lion Mountain Sandstone (Cam- brian), Texas. Geol. Soc. Am. Bull., 100: 1803-1810.

Molenaar, N., 1990. Calcite cementation in shallow-marine Eocene sandstones. J. Geol. Soc. London (in press).

Molenaar, N., van de Bilt, G.P., van den Hock Ostende, E.R. and Nio, S.D., 1988. Early diagenetic alteration of shallow- marine mixed sandstones: an example from the Lower Eocene Roda Sandstone Member, Tremp-Graus Basin, Spain. Sediment. Geol., 55: 295-318.


MOiler, N.K. and Kvingan, K., 1988. The genesis of nodular limestones in the Ordovician and Silurian of the Oslo Region. Sedimentology, 35: 405-420.

Mullins, H.T., Neumann, A.C., Wilber, R.J. and Boardman, M.R., 1980. Nodular submarine cementation on Bahamian slope--a possible precursor for the origin of some nodular limestones. J. Sediment. Petrol., 50: 117-131.

Nijman, W. and Nio, S.D., 1975. The Eocene Monta~ana delta (Tremp-G-raus Basin, provinces of Lerida and Huesca, southern Pyrenees, N. Spain). Int. Assoc. Sedimentol., 9th Int. Congr., Nice, Part B, 20 + XXXVI pp.

Nio, S.D. and Yang, C.S., 1983. Dynamics, geometry and sequential upbuilding of large subtidal bedforms. Proc. Int. Symp. Sedimentation on the Continental Shelf, With Spe- cial Reference to the East China Sea, Hangzou, China, 1: 20-36.

Ori, G.G. and Friend, P.F., 1984. Sedimentary basins formed and carried piggyback on active thrust sheets. Geology, 12: 475-478.

Pirrie, D., 1987. Orientated calcareous concretions from James Ross Island, Antarctica. Br. Antarct. Surv. Bull., 75: 41-50.

Pryor, W.A., 1972. Reservoir inhomogeneties of some recent sand bodies. Soc. Pet. Eng. J., 12: 229-245.

Pryor, W.A., 1973. Permeabihty-porosity patterns and variations in some Holocene sand bodies. Bull. Am. Assoc. Pet. Geol., 57: 162-189.

Rahmani, R.A., 1970. Carbonate concretions in the Upper

Cretaceous Cedar District Fol~nation, British Columbia. Bull. Can. Pet. Geol., 18: 282-288.

Raiswell, R., 1971. The growth of Cambrian and Liassic concretions. Sedimentology, 17: 147-171.

Raiswell, R., 1987. Non-steady state microbiological diagenesis and the origin of concretions and nodular limestones. In: J.D. Marshall (Editor), Diagenesis of Sedimentary Se- quences. Geol. Soc. London, Spec. Publ., 36: 41-54.

Roberts, H.H. and Whelan, T., 1975. Methane-derived carbonate cements in barrier and beach sands of a sub- tropical delta complex. Geochim. Cosmochim. Acta, 39: 1085-1089.

Trask, P.D., 1930. Mechanical analysis of sediment by centri- fuge. Econ. Geol., 25: 581-599.

van Baaren, J., 1979. Quick-look permeability estimates using sidewall samples and porosity logs. Trans. 6th. Europ. Formation Evaluation Symp., Soc. Prof. Well Log Analysts, 11 pp.

Weeks, L.G., 1957. Origin of carbonate concretions in shales, Magdalena Valley, Colombia. Geol. Soc. Am. Bull., 68: 95-102.

Williams, G.D., 1985. Thrust tectonics in the south central Pyrenees. J. Struct. Geol., 7: 11-17.

Yang, C.S. and Nio, S.D., 1985. The estimation of palaeohy- drodynamic processes from subtidal deposits using time series analysis methods. Sedimentology, 32: 41-57.

Documents

Origin of nodules in mixed siliciclastic-carbonate sandstones, the Lower Eocene Roda Sandstone Member, southern Pyrenees, Spain