REPRODUCTIVE BIOLOGY OF AGKISTRODON PISCIVORUS … · reproductive biology of agkistrodon piscivorus lace´pe`de (squamata, serpentes, viperidae, crotalinae) dustin s. siegel 1,4,david

REPRODUCTIVE BIOLOGY OF AGKISTRODON PISCIVORUSLACEPEDE (SQUAMATA, SERPENTES, VIPERIDAE, CROTALINAE)

DUSTIN S. SIEGEL1,4, DAVID M. SEVER2, JUSTIN L. RHEUBERT

2, AND KEVIN M. GRIBBINS3

1Department of Biology, Saint Louis University, St. Louis, MO 63103, USA2Department of Biological Sciences, Southeastern Louisiana University, Hammond, LA 70402, USA

3Department of Biology, Wittenberg University, Springfield, OH 45501, USA

ABSTRACT: Aspects of the reproductive biology of male and female Agkistrodon piscivorus are describedusing histological techniques, reviewed, and compared with historical data on A. piscivorus. These includeanatomical description at the macro and microscopic levels, and correlation of the male and femaleurogenital cycles to the reproductive life history of A. piscivorus. New anatomical descriptions anddiscussion of the efferent ducts, including the ductuli efferentes, proximal and distal ductuli epididymides,ductus deferens, and ampulla ductus deferens, are also presented here at the light and electron microscopylevels. Morphology of all distinct regions of the male and female urogenital systems are discussed andcompared with historical investigations on other squamates. In comparison to other snakes, A. piscivoruspossesses some unique reproductive characters, whereas others are more conserved. In terms of thereproductive cycle, the ability of males and females to store sperm allows the dissociation of reproductiveevent timing between the sexes. Thus, the only event that must be coordinated between the sexes iscopulation, which is proposed to occur in the fall and the spring in A. piscivorus. In females, the atrophyand activity of the reproductive organs (e.g., secretory activity) varies concurrently with vitellogenesis andthe mating seasons. In males, spermatogenesis peaks in the summer, independent of the mating season,except in Louisiana where a spring and fall peak of spermatogenesis occur, and where the spring and fallmating seasons overlap. Renal sexual segment hypertrophy in males peaks in the fall and spring in moresouthern populations (Alabama and Louisiana) concurrent with fall and spring mating seasons. In Georgia,only a fall peak is observed. Secretory activity of the male excurrent ducts also peaks during times of matingactivity in one population studied (Louisiana).

Key words: Agkistrodon; Histology; Morphology; Reproduction; Squamata; Ultrastructure; Viperidae.

RECENTLY, Agkistrodon piscivorus (Cotton-mouth), a semiaquatic, viviparous, and ven-omous snake common throughout the south-eastern United States (Conant and Collins,1998), has been recognized as a valuable toolfor studying certain aspects of the biology ofsnakes, and was the subject of a day-longsymposium at the Joint Meeting of Ichthyol-ogists and Herpetologists in Montreal, Canada(July, 2008). Unlike many other crotalines, A.piscivorus is abundant in areas of suitablehabitat, and for that reason alone represents amodel for studying the biology of crotalinesnakes with invasive techniques (e.g., histolo-gy). We feel it timely to discuss and reviewhistorical aspects of the reproductive biologyof A. piscivorus, particularly the reproductiveanatomy and reproductive cycles (male andfemale), with data from recent investigations(Graham et al., 2008; Gribbins et al., 2008;Gribbins et al., in press; Sever et al., 2008;Siegel and Sever, 2006, 2008a,b). This marks

the first review of the biology of A. piscivorusthat focuses completely on aspects of repro-ductive biology (a previous review focusedattention on the natural history of this species;see Burkett, 1966). The following summary ofhistorical and recent data carries great value inpinpointing future directions for studyingreproduction of A. piscivorus and othercrotalines.

Topics of this review include the gross andmicroscopic reproductive anatomy of maleand female Agkistrodon piscivorus, seasonaltiming of reproductive events, and variation intiming of reproductive events in differentgeographic populations of A. piscivorus. Interms of anatomy, comparisons are primarilywith other snakes, although other squamatesare considered where appropriate. Compari-sons of reproductive cycles generally arerestricted to different populations of A.piscivorus and to other North Americancrotaline snakes. Considering other investiga-tors are currently focusing attention towardthe reproductive ecology of A. piscivorus4 CORRESPONDENCE: e-mail, [email protected]

Herpetological Monographs, 23 2009, 74–107

E 2009 by The Herpetologists’ League, Inc.

74

(pers. comm. N. B. Ford; E. Menzel), weleave this facet of reproductive biology out ofthe following review.

BACKGROUND

Previous Anatomical Studies of Viperidae

Few comprehensive studies, incorporatingboth males and females, exist on the repro-ductive anatomy of squamates, and of thesestudies viperids have been utilized in anatom-ical description rarely. Volsoe (1944) de-scribed the male urogenital system in Viperaberus, and Saint Girons (1959, 1973) de-scribed the female in V. aspis and Cerastescerastes. Both of these in-depth tomes ofliterature were performed on Old Worldvipers from the Viperinae, the sister group tothe Crotalinae, which contains all New Worldvipers and a few Old World species (Castoeand Parkinson, 2006). Collectively, membersof the Crotalinae are termed the pitvipers.Ludwig and Rahn (1943) noted histologicalstructure of the posterior oviduct of acrotaline (Crotalus viridus) in an effort toinvestigate copulatory adjustment and spermtransport. Other studies have described ana-tomical characters in crotalines in an effort todetermine the timing of reproductive eventsby describing the testicular cycle and/or renalsexual segment (Rss) hypertrophy (e.g., Al-dridge et al., 2008; Schuett et al., 2002).However, prior to this analysis, in-depthanatomical descriptions have never beencompiled on all reproductive organs of acrotaline utilizing both males and females.Also, until now comparison of ultrastructuralvariation observed in reproductive organs tolife history traits (e.g., reproductive cycles) hasonly been accomplished in a non-viperid,Seminatrix pygaea of the Colubridae (Grib-bins et al., 2005; Sever et al., 2000; Sever etal., 2002; Sever and Ryan, 1999).

Reproductive Cycles of NorthAmerican Crotalines

The reproductive cycles of North Americancrotalines have been reviewed by Schuett(1992) and Aldridge and Duvall (2002). Thus,the following is for the most part a briefsynthesis of these massive works, updatedwith new information.

In general, female North American crota-lines (Agkistrodon, Crotalus, Sistrurus) beginvitellogenesis in the late summer to early fall(Schuett, 1992; for complete list of Crotalusand Agkistrodon species studied see Al-dridge and Duvall, 2002; Sistrurus, Aldridgeet al. 2008). Around this time matingcommences, which is evidenced by thepresence of sperm in the posterior oviduct(Gloyd and Conant, 1990; Rahn, 1942; Siegeland Sever, 2006). Subsequently, a period ofhibernation ensues, where the ovarian folli-cles of females arrest in development (Al-dridge and Duvall, 2002; Aldridge et al.,2008). After hibernation, females may entera second mating season (for a list of biannualbreeders see Aldridge and Duvall, 2002), atwhich time follicles continue developmentuntil ovulation in the late spring (Aldridgeand Duvall, 2002; Aldridge et al., 2008;Schuett, 1992). Thus, fertilization occurs inthe spring and utilizes sperm from theprevious fall mating season, a spring matingseason, or both. All North American crota-line snakes are viviparous and carry theiryoung until parturition in the fall (Aldridgeand Duvall, 2002).

Spermatogenesis in male North Americancrotalines begins in the late spring–earlysummer (aestival; Aldridge and Duvall, 2002;Saint-Girons, 1982), coincident with an in-crease of testosterone (Johnson et al., 1982;Schuett et al., 1997; Schuett et al., 2002), andends in the early fall coincident with matingand male agonistic behaviors (Schuett et al.,1997). However, secondary sexual characteractivity (i.e., Rss) can be hypertrophied in onlythe fall or in the fall and spring (Aldridge,2002; Johnson et al., 1982) and is alsoandrogen dependent (Prasad and Sanyal,1969). The hypertrophy of the Rss may belinked to the mating season of temperatepitvipers (Aldridge, 2002), however, it is clearthat spermatogenesis can be dissociated frommating activity due to the ability of males andfemales to store sperm for prolonged periodsof time (Aldridge and Duvall, 2002; Schuett,1992). This is not always the case, which canbe seen in Crotalus atrox and C. scutulatuswhere mating and spermatogenesis overlap(Goldberg, 2007; Schuett et al., 2002; Tayloret al., 2004).

2009] HERPETOLOGICAL MONOGRAPHS 75

Male crotalines that mate in the spring (fora list see Aldridge and Duvall, 2002) alsopossess an increase in testosterone levels atthis time (Schuett et al., 1997; Schuett et al.,2002; Schuett et al., 2005; Taylor et al., 2004),which likely stimulates reproductive behaviorssuch as mating and male–male combat(Schuett et al., 1997). However, this increasein testosterone is not as high compared to thesummer–fall peak, which cues both spermato-genesis and reproductive behaviors (Schuettet al., 1997).

MATERIALS AND METHODS

Specimens and Tissue Preparation

The following synthesis adds new anatom-ical descriptions of the efferent ducts of maleAgkistrodon piscivorus. Adult male specimenswere collected in southeast Louisiana fromthe Amite River Diversion Canal (North30.22616/West 090.68506), Livingston Parish(private property owned by Dr. CliffordFontenot; North 30.871/W090.36202), andPass Manchac (North 30.29426/West 090.35592). These were the same samples collect-ed for use in recent studies on A. piscivorustestes (Gribbins et al., 2008; Gribbins et al., inpress) and kidneys (Sever et al., 2008).Efferent ducts were dissected out of everymale and prepared in the exact same mannerfor light and electron microscopy analyses asSever et al. (2008). Sampling per month wasas follows: January, one; February, two;March, two; April, two; May, two; June, five;July, two; August, two; September, one;October, six; and November, two.

Microscopy Analysis

Representative light micrographs were ob-tained by a combination of a Leica DM2000microscope and Leica DFC420 digital camera(Leica Microsystems, Wetzlar, Germany).Transmission electron micrographs were ob-tained by a combination of a JEOL 100transmission electron microscope and L3CCCD digital camera system (Scientific Instru-ments and Applications, Duluth, GA). Scan-ning electron micrographs were taken with aPhilips XL-20 electron microscope (PhilipsElectronics N.V., Eindoven, Netherlands). Alldigital micrographs were uploaded directly

into Adobe Illustrator and Photoshop CS forediting (Adobe Systems, San Jose, CA).

COMPARATIVE MORPHOLOGY: MALES

Gross Morphology

Overview.—When discussing the reproduc-tive anatomy of male squamates, three mainareas of interest are important to consider: (1)the testes, (2) efferent ducts (including theepididymis and ductus deferens), and (3) thekidney (Fig. 1), which produces a sexualsecretion in squamates and possibly allLepidosauria (Sever et al., 2000). From theseminiferous tubules of the testis, sperm passsequentially through the ductuli efferentes,ductuli epididymides, ductus epididymis, duc-tus deferens, and the ampulla ductus defer-entis. The ampulla joins with the ureter in thecloacal wall to form a urogenital duct thatopens into the cloaca. The testis and the ductsassociated with it are suspended from thebody wall by a mesorchium, so the testis canbe easily lifted and moved from side to side.The rest of the urogenital system, i.e., ductusdeferens, kidney and ureter, are retroperito-neal until the ductus deferens and ureterseparately enter the cloacal wall.

Histology

Testis and germ cell morphology.—Thetestis of Agkistrodon piscivorus is morpholog-ically similar to that of other terrestrialamniotes (Gribbins et al., 2008). Seminiferoustubules are packaged within a thick connec-tive tissue capsule. The tubules in transversesection have the same organization as those ofother squamates (Fig. 2) and have verydistinct basement membranes and a perma-nent population of Sertoli cells that reside onthe basal lamina. Germ cells develop throughthe stages of spermatogenesis within cellularprocesses of the Sertoli cells, which composethe seminiferous epithelium of each tubule.Spermatogonia reside near the basementmembrane and divide by mitosis within thebasal compartment to give rise to the middlelayer of cells within the seminiferous epithe-lium, the spermatocytes. Spermatocytes un-dergo both divisions of meiosis and give rise tothe round spermatids of spermiogenesis. Thespermatids undergo spermiogenesis within

76 HERPETOLOGICAL MONOGRAPHS [No. 23

the apical portion of the seminiferous epithe-lium and are shed as mature spermatozoa intothe lumen of the seminiferous tubules at thecompletion of spermiogenesis.

The germ cell morphology is very similar tothat of the colubrid Seminatrix pygaea (Grib-

bins et al., 2005) and other reptiles studied todate (Gribbins and Gist, 2003; Gribbins et al,2003; Gribbins et al., 2006). The germ cellscan be described and categorized according tothe nomenclature established by Russell et al.(1990) for mammalian species.

FIG. 1.—(A) Gross morphology of the male Agkistrodon piscivorus urogenital system. (B) Close up of testisdemonstrating the association of efferent ducts to the testes. Ag, adrenal gland; Cl, cloaca; Dd, ductus deferens; Ep,epididymis; Hp, hemipene; Kd, kidney; Ts, testis; Ur, ureter.

FIG. 2.—Cross section of a May seminiferous tubule from Agkistrodon piscivorus, and cell types represented within.Labeled cell types (in order of development): SpA, type A spermatogonia; SpB, type B spermatogonia; LP, leptotenespermatocytes; S1, step 1 spermatid; S2, step 2 spermatid; S4, step 4 spermatid; S6, step 6 spermatid; S7, step 7spermatid, MS, Mature Spermatozoon.


The seminiferous epithelium contains twomorphological populations of pre-meiotic cells(Spermatogonia A and B; Fig. 3. SpA andSpB) during all months of the year. Thesecells are characterized by nuclei with randomclumps of heterochromatin. The major mor-phological differences between the two typesof spermatogonia are that the A type are ovoidin shape with one large nucleolus and B typeare more round in shape and usually lack aprominent nucleolus. Both spermatogonialtypes are generally found near the basementmembrane of the epithelium in associationwith Sertoli cells. During the spermatogeniccycle both types of spermatogonia undergomitosis to maintain the spermatogonial popu-lation and many of the B spermatogonia divideto form pre-leptotene spermatocytes that thenenter meiosis.

Meiotic cells are characterized by anincrease in nuclear size and condensation ofchromatin into chromosomes. Pre-leptotenethrough pachytene spermatocytes (Fig. 3. PL–PA) show a swift increase in nuclear size withcondensation of the chromatin into visualizedchromosomes.

Diplotene spermatocytes (Fig. 3. DI), meta-phase I cells (Fig. 3. M1), secondary spermato-cytes (Fig. 3. SS), and metaphase II cells(Fig. 3. M2) can be found together within theseminiferous epithelium throughout the ac-tive months of spermatogenesis. In diplotenespermatocytes, the nuclear membrane beginsto degenerate and the almost fully con-densed chromatin fibers form a tight circlejust under this degenerating membrane.Metaphase 1 cells have fully condensedchromosomes that aggregate on the meta-phase plate. The results of meiosis I are thesecondary spermatocytes and their chroma-tin is dispersed randomly throughout thenucleoplasm. During Metaphase 2, chromo-somes aggregate around the metaphase plateagain. The only differential factor betweenMetaphase 1 and Metaphase 2 is thatMetaphase 2 cells are slightly smaller andcontain about half the amount of chromatinthan Metaphase 1 cells.

Spermiogenesis can be divided into sevensteps within the Agkistrodon piscivorusgerminal epithelium based on the terminol-ogy of Russell et al. (1990) for mammalian

species. Steps of spermiogenesis are definedbased on acrosomal formation, nuclear elon-gation, and chromosomal condensation. Step1–4 spermatids (Fig. 3. S1–S4) represent theearly round spermatids. The major trendseen during the round spermatid stage is thedevelopment of the acrosome and acro-some granule on the apex of the spermatidnucleus.

Step 5 spermatids mark the transitionbetween round and elongating spermatids(Steps 5–7; Fig. 3. S5–S7). Elongation beginsat the opposite end of the nucleus away fromthe acrosome creating a nucleus that isstretched on its dorsoventral plane. As elon-gating spermatids undergo development, theyalso begin to accumulate near the apicalsurfaces of the Sertoli cells with their tailsstretching out into the lumen and theirnuclear heads facing the basement mem-brane. As elongation proceeds, spermatidsbecome cylindrical and roughly filiform inappearance. The width of these spermatidsdecreases as the DNA within the nucleuscondenses into a dark uniform mass within thenucleoplasm. Once spermiogenesis is com-plete the mature spermatozoa (Fig. 3. MS) arereleased into the lumen of the seminiferoustubules where they will be transported to theexcurrent ducts of the male reproductivesystem.

Germ cell development strategy.—Sper-matogenic events occur during two separateseasons (spring and late summer–early fall)throughout a single calendar year withinAgkistrodon piscivorus testes from the mostsoutherly population studied (Louisiana; Grib-bins et al., 2008). The beginning of the firstwave of spermatogenesis commences inMarch with the presence of meiotic cells.Spermatogenesis climaxes in June (Fig. 4A)with the late stages of spermiogenesis, heavyspermiation, and mature spermatozoa flood-ing the lumen of the seminiferous tubules.During this time (March–June) a gradualincrease in the seminiferous tubule diameter(STD; 126 mm, 131 mm, 150 mm, 181 mm) andmean germinal epithelial height (GET;28.2 mm, 29.2 mm, 38.1 mm, 43.8 mm; Fig. 5top and bottom) can be observed histologicallyand quantitatively (Fig. 5, superscripts STD:a–d; GET: a–c).


The testes enter a quiescent period duringJuly (Fig. 4B) in which the germinal epithe-lium only consists of spermatogonia A and Bcells. The epithelium becomes highly vacuo-lated and the lumen is void of mature

spermatozoa. A significant decrease in theseminiferous tubule diameter (92.7 mm) andthe epithelial height (20 mm; Fig. 5, super-scripts STD: a,b and GET: a) can be observedthrough statistical analyses.

FIG. 3.—Germ cell types found within the seminiferous epithelium of Agkistrodon piscivorous. SpA, type Aspermatogonia; SpB, type B spermatogonia; PL, pre-leptotene spermatocyte; LP, leptotene spermatocyte; ZY, zygotenespermatocyte; PA, pachytene spermatocyte; DI, diplotene spermatocyte; M1, meiosis I; SS, secondary spermatocyte;M2, meiosis II; S1, step 1 spermatid; S2, step 2 spermatid; S3, step 3 spermatid; S4, step 4 spermatid; S5, step 5spermatid; S6, step 6 spermatid; S7, step 7 spermatid; MS, mature spermatozoon.


Spermatogenesis commences again in latesummer–early fall (August) with the presenceof meiotic cells and early spermatids andcontinues through October. During themonth of October (Fig. 4C) all germ cellswithin the cell cycle are represented withinthe epithelium of the seminiferous tubules.The lumina of the tubules are again floodedwith mature spermatozoa similar to that ofJune. This increase in testicular activity isparalleled by an increase in seminiferoustubule diameter and epithelial height inOctober (STD: 215.4 mm and GET: 41.5 mm;Fig. 5, superscripts STD: d and GET: c).

November through February mark thesecond quiescent period of spermatogenesis.During these months the seminiferous tubulesdecrease (95 mm, 101.25 mm, 99.8 mm) in size,become highly vacuolated, and only spermato-gonia A and B cells are present. The lumen isonce again void of mature spermatozoa. Thedecrease in testicular activity is shown by thedecrease in tubular diameter and epithelialheight similar to that observed during July(Fig. 5, superscripts STD: a,b and GET: a,b).

From the above data, two distinct events ofspermatogenesis can be seen within the testesof the southeastern Louisiana population ofAgkistrodon piscivorus. The first event beginsin March with the onset of spermatogenesisand is completed in July with heavy spermi-ation. The two events are separated by aquiescent period in July before spermatogen-

esis commences again in August. This isevidence for the first description of biannualspermatogenesis within a North Americanspecies of Crotalinae.

Crotalines of North America typically havea single event of spermatogenesis that occursduring the late spring, peaks during summer,and ends in early fall (Aldridge, 1979;Aldridge, 1993; Aldridge, 2002; Aldridge andBrown, 1995; Aldridge et al., 2008; Goldberg,1999a,b; Goldberg, 2000a,b,c; Goldberg andHolycross, 1999; Goldberg and Rosen, 2000;Schuett et al., 2002). Johnson et al. (1982)described the spermatogenic events of anAlabama population of Agkistrodon piscivo-rous, and the spermatogenic cycle mirroredthat of a typical North American Crotalinae.Seasonal testicular histology from Graham etal. (2008) confirms this finding in a populationfrom Georgia. The only other viperid knownto exhibit biannual events in the spermato-genic cycle is Vipera berus (Saint-Girons,1982), however, this is due to the culminationof spermiogenic events in the spring, not asecond bout of spermatogenesis as describedin Louisiana A. piscivorus. In general, sper-matogenesis in the colubrids of North Amer-ica follows a very similar pattern to typicalNorth American crotalines with spermatogen-esis commencing in the spring, reaching apeak in the summer, and concluding in the fall(for review see Aldridge et al., in press).However, one species, Masticophis bilineatus,

FIG. 4.—(A) Cross sectional view of June seminiferous tubule with higher magnifications of present cell types. Celltypes present within the seminiferous epithelium include: SpA, type A spermatogonia; SpB, type B spermatogonia; PL,pre-leptotene spermatocyte; PA, pachytene spermatocyte; DI, diplotene spermatocyte; S2, step 2 spermatid; S6, step 6spermatid; MS, mature spermatozoa. (B) Cross sectional view of July seminiferous tubule with higher magnifications ofpresent cell types. Cell types present within the seminiferous epithelium include: SpA, type A spermatogonia; SpB, typeB spermatogonia. (C) Cross sectional view of October seminiferous tubule with higher magnifications of present celltypes. Cell types present within the seminiferous epithelium include: SpA, type A spermatogonia; SpB, type Bspermatogonia; PA, pachytene spermatocyte; M1, meiosis I; S2, step 2 spermatid; S3, step 3 spermatid; S4, step 4spermatid; S5, step 5 spermatid; S7, step 7 spermatid; MS, mature spermatozoa.


possesses a biannual spermatogenic cycle withdistinct spermatogenic cycles occurring inboth the fall and the spring (Goldberg,1998). Two peaks also seem to occur inTrimorphodon biscutatus (Goldberg, 1995),however, it has been hypothesized that thespring peak is just the conclusion of the fallspermatogenic cycle as described in V. berus.

Ultrastructure of spermiogenesis.—Sperma-tid morphology in Agkistrodon piscivorustestes (Gribbins et al., in press) is very similarto that observed within the colubrid Semina-trix pygaea (Gribbins, unpublished). A num-ber of recent studies have focused on someaspects of spermiogenesis (Al-Dokhi, 2004,2005; Al-Dokhi et al., 2004, 2005a,b; Dehlawiand Ismail, 1994; Hondo et al., 1994; Ismail etal., 1995) or ultrastructure of the spermato-

zoon (Al-Dokhi et al., 2007; Oliver et al., 1996;Tourmente et al. 2006, 2008) within Ser-pentes. To our knowledge, A. piscivorusrepresents the first snake species to have itsentire spermiogenic cycle described ultra-structurally. Furthermore, there are only twopapers dealing with ultrastructure within themale testis of crotalines and they primarilycover the morphology of the spermatozoon(Cunha et al., 2008; Tourmente et al., 2008),with no information on spermiogenesis.

Spermatid morphological changes duringspermiogenesis within Agkistrodon piscivorusare very similar to those seen within othersquamates studied to date, such as Scincellalateralis (Gribbins et al., 2007; Gribbins et al.,in press). However, subtle differences existbetween S. lateralis and other squamate testesduring the morphogenesis of spermatids whencompared to A. piscivorus. For brevity, wewill focus here on the more interestingdifferences observed in the ultrastructure ofspermatids and mature spermatozoa. Figure 6shows the round (A and B) and early elongatestages (C) of spermiogenesis in A. piscivorus.The developing acrosome and juxtapositionedGolgi apparatus are seen within A. piscivorusround spermatids similar to that described forother squamates. However, the developmentof the acrosome granule early in acrosomedevelopment in a centralized position withinthe acrosome vesicle, prior to contact with thenuclear apex, seems to be unique to A.piscivorus spermatids. In S. lateralis thegranule forms later (based on vesicle size),and the granule is located against the acro-some vesicle membrane. The later events ofacrosome development in round spermatidswithin A. piscivorus (Fig. 6C) are similar tothat of S. lateralis. One major differencebetween S. lateralis and A. piscivorus elongatestages is the large and prominent peripheralfibers associated with microtubule doublets 3and 8 (Fig. 7F,G) in cross sections of thedeveloping flagella within the midpiece of A.piscivorus round and elongating spermatids(Fig. 6C), which is a similar trait found inmature spermatozoa of Crotalus durissus(Cunha et al., 2008). These fibers have alsobeen described in other squamate spermato-zoa (Jamieson, 1999; Oliver et al., 1996;Tourmente et al., 2008). Unique to A.

FIG. 5.—(Top) Variation in seminiferous tubule diam-eter and (Bottom) variation in germinal epithelial heightduring the annual reproductive cycle male Agkistrodonpiscivorus from Louisiana. Values represented on thisgraph are means 6 1 standard error. Different super-scripts indicate significant differences (Dunn’s multiplerange test between months). Note that September (*)represents an aberration in the data. This specimen wascaught one week after hurricane Katrina and outside of itsnormal habitat. Thus, spermatogenesis was halted becauseof stress-induced hormones (see Gribbins et al. 2008).


piscivorus spermiogenesis is the large opennucleoplasmic vesicles on either side thedeveloping flagella in early elongating sper-matids (Fig. 6C, inset). Another major differ-ence between A. piscivorus elongating sper-matids and those of other squamates studiedto date is the accumulation of a thick band ofdense staining material inside of the outeracrosomal membrane (Figs. 6C, 7D) duringthe development of the late stage spermatids(Fig. 7A,B). This may be a synapomorphy ofCrotalinae; however, more species withinSerpentes need to be studied before thissuggestion can be tested robustly. The acro-some complex of late elongates is very similarin A. piscivorus to that described for othersquamates except for the absence of theperpendicular part of the manchette(Fig. 7E), which has also been described inS. lateralis (Gribbins et al., 2007). Densebodies within the midpiece of A. piscivorussperm are scattered and not as abundant asthose described in some lizards, a trait alsoobserved in another crotaline, C. durissus(Cunha et al., 2008).

Morphology of the efferent ducts.—Exceptin the region of the efferent ducts, the outerseminiferous tubules are bordered by a layerof collagen fibers and the thin, superficial

visceral pleuroperitoneum (Fig. 8A). In theregion of the excurrent ducts, the tunicapropria of the serosa splits, becomes thick-ened and invested with smooth muscle, bloodvessels and more fibrous connective tissue(Fig. 8A,B). Distal ends of seminiferous tu-bules as well as the ductuli efferents, ductuliepididymides, and the ductus epididymis areencased in this capsule, along with the adrenalgland (Fig. 8B).

The ductuli epididymides can be foundalong the entire medial side of the testis. Theductus epididymis and adrenal gland appearat the start of the posterior two-thirds of thetestis. Ductuli efferentes are short tubules thatbranch off the distal ends of seminiferoustubules that protrude into the epididymalsheath (Fig. 8C). The seminiferous tubulesnarrow abruptly into the ductuli efferentes,which are composed of simple cuboidal,basophilic epithelium and lack cilia or asecretory product (Fig. 8D).

The proximal and distal ductuli epididymi-des differ histologically from one another andfrom the epididymis (Fig. 9A,B). The distalductuli epididymides empty into the epididy-mis along its entire length.

The proximal ductuli epididymides (Pde)are up to 100 mm in diameter, somewhat

FIG. 6.—Round and early elongate spermatids undergoing acrosome development during the early stages ofspermiogenesis within the seminiferous epithelium of Agkistrodon piscivorus. (A) The acrosomal vesicle (whitearrowhead) is juxtapositioned to the apical portion of the nucleus (Nu). The vesicle is in the early phase of growth. Thecytoplasm of the spermatid has numerous mitochondria (white arrow) and many layers of endoplasmic reticula (blackarrow). Insert: Shows that the vesicle (AV) has not quite made contact with the nuclear membrane. (B) Very late-stageround spermatid exhibiting a deep indented acrosome (AV) and a prominent acrosomal granule (white arrowhead). Anaccumulation of dark staining proteins are lining the nuclear membrane side of the subacrosomal space (black arrow).The caudal portion of the nucleus has begun elongation and the proximal centriole (black arrowhead) can be seen insagittal section and the growing flagellum is shown in transverse section (white arrow) near the caudal end of thenucleus. Insert: Shows the proximal centrioles (white arrow) and transverse flagella in greater detail. The flagellum hastwo opposing protein plaques/dense bodies (black arrowhead) juxtapositioned to the conserved 9 + 2 microtubulecore. (C) The development of the flagella is prominent in sagittal sections of elongates showing the presence of both theproximal (Pc) and distal (Dc) centrioles arranged in a perpendicular fashion. The distal centriole begins to elongate toform the flagellum. Insert: Large open nucleoplasmic vesicles (black arrowheads) devoid of chromatin can be seenlateral to the developing flagellum.


FIG. 7.—(A) Sagittal section of Agkistrodon piscivorus elongate spermatid in the termination phase of spermiogenesiswith all three major parts of the spermatid visible (acrosome, nucleus, and flagella; white line through the middle of Adenotes that two separate micrographs were combined to obtain this image). Lines and represented letters showapproximately where transverse sections (CS) occurred within spermatids at or near the same stage of development as A inorder to obtain cross sections B–G. Note the well-developed parallel microtubules of the manchette (black arrow) runningalong side the nucleus and the perforatorium (white arrowhead). (B) CS through the cranial subacrosomal space andacrosomal vesicle. Granulated protein layer of subacrosomal space (white *), acrosomal vesicle (black *), proteinaccumulation on the inside of the outer acrosomal membrane (white arrowhead). (C) CS through the subacrosomal space,perforatorium, and acrosomal vesicle. The white arrowhead points to the perforatorium, which is surrounded by thegranulated proteins of the subascrosomal space. The white circular region around the subascrosomal space is the acrosomalvesicle, which again has protein accumulations under its outer membrane (black arrowhead). Also note the numerousSertoli cell membrane layers surrounding the entire acrosomal complex (white arrow). (D) CS through the nucleus andacrosome vesicle shoulders. The white * labels the subacrosomal space. Within the middle of this space is the conical pointof the nucleus in CS. Also present are the protein accumulation under the outer membrane of the acrosome (white arrow)and the Sertoli cell membrane layers (white arrowhead). (E) CS through nucleus proper. Nucleus (Nu), Manchette (*),inner single circular microtubule layer (white arrow). (F) CS through the proximal neck of the flagella. Axoneme is nicelyrepresented with 9 pairs of doublet microtubules and a microtubule triplet within the center of the axoneme. Attachedprotein blacks within the axoneme (white arrow). This CS represents a transition zone between the basal plate and theproximal neck. (G) CS through the midpiece. Dense fibrous sheath/ring (white arrowhead), mitochondria (white arrow).


FIG. 8.—Light micrographs of transverse sections through the testis and proximal efferent ducts of a 57.6 cm SVLmale Agkistrodon piscivorus collected 2 October. (A) Medial testis showing transition of superficial lining associatedwith the region of the efferent ducts. (B) Relationship between the seminiferous tubules, efferent ducts, and the adrenalgland. (C) Relation of seminiferous tubules, ductuli efferentes, and ductuli epididymides. (D) Transition betweenseminiferous tubules and ductuli efferentes. Adg, adrenal gland; Bv, blood vessel; Cf, collagen fibers; Dde, distal ductuliepididymides; Def, ductuli efferentes; Pde, proximal ductuli epididymides; St, seminiferous tubules; Tp, tunica propria;Vp, visceral pleuroperitoneum.


irregular in shape, and in March–April andAugust–November, the epithelium is simplecolumnar. The basal nuclei have their longaxes oriented with that of the cells, and the

apical cytoplasm is basophilic, although thebasophilia is not as intense as for nuclei(Fig. 9C). The epithelium is reduced tocuboidal in a specimen collected 23 February

FIG. 9.—Light micrographs of the proximal efferent ducts associated with the testis of a 46.1 cm SVL maleAgkistrodon piscivorus collected 13 March. (A) Overview stained with hematoxylin-eosin. (B) Overview stained withbromphenol blue. (C) Detail of histology of the ductuli epididymides stained with hematoxylin-eosin. (D) Distal ductuliepididymides stained with PAS and alcian blue. (E) Detail of ductus epididymis stained with hematoxylin-eosin. (F)Ductus epididymis stained with PAS. BB+, bromphenol blue positive; Dde, distal ductuli epididymides; Ep, epithelium;Lu, lumen; PAS+, periodic acid-Schiff’s reagent positive; Pde, proximal ductuli epididymides; SpSm, sperm insecretory material.


and one collected 27 June. Neither theepithelium or luminal material other thansperm react with periodic acid-Schiff’s (PAS),alcian blue, or bromphenol blue. Sperm froma spermiation event apparently pass throughthese tubules quickly as sperm occur only in aspecimen collected 14 June and one collected12 August.

The distal ductuli epididymides (Dde) aresmaller and more numerous than the Pde,with a diameter of 35–50 mm. The epitheliumis cubodial with the long axes of the basalnuclei oriented with the basement membrane,and the cytoplasm, like that of the Pde, isbasophilic (Fig. 9C). PAS+ granules (Fig. 9D)occur in specimens from every month exceptfor a specimen collected on 27 June, in whichthe Dde, like the Pde, is highly regressed.Bromphenol blue positive material is alsoscattered through the cytoplasm of the Dde.Fluocculent material, possibly in vacuoles,occurs in the cytoplasm of specimens collect-ed 30 May, 14 June, and 12 August. Sperm arefound in specimens collected 24 May, 17August, 17 October, and 22 November.

The ductus epididymis is the largest of theproximal efferent ducts, with a diameter of150–250 mm. The epithelium appears pseu-dostratified with columnar principle cells likethose of the Pde and scattered basal cells. Theapical cytoplasm of the basal cells is onceagain basophilic, but secretory material in thelumen is decidedly eosinophilic (Fig. 9E). Thesecretory material in the lumen is also stronglyPAS positive (Fig. 9F). Luminal secretorymaterial is found in specimens collected 13March, 28 March, 24 May, 30 May, 18 June,and 22 November, and sperm are associatedwith the secretory mass on 13 March, 28March, and 17 August (Fig. 9E). A scantamount of sperm is found in the epididymis ofa male collected 23 February and which lackssecretory material in the lumen. The cyto-plasm of glands with secretory material in thelumen did not respond generally to histo-chemical tests, although a scattered, diffusereaction to PAS, alcian blue, and bromphenolblue is detectable. The individual collected 27June has a completely regressed epididymis,but in the others, the epithelium generallyseems hypertrophied. Vacuoles of flocculentmaterial occur in the epithelium of a specimen

collected 12 August, which also possesses suchvacuoles in the Dde.

The ductus deferens is the continuation ofthe epididymis posterior to the testis, andpasses retroperitoneally with the ureter to-ward the cloaca. The ductus deferens isformed into short tight loops (not coils) as itpasses caudally. The demarcation betweenepididymis and ductus deferens is gradual,with the lumen of the ductus becoming wider,from 200–300 mm in diameter, and theepithelium becoming lower. Other studieshave reported that the epithelium of thesquamate ductus deferens is pseudostratifiedbut that is difficult to verify by light micros-copy (Fig. 10A,B). The epithelium of theprinciple cells usually is cuboidal with a scantamount of basophilic cytoplasm, althoughcolumnar principle cells are found in speci-mens collected 18 June, 27 June, 27 July, and17 October. Copious sperm occur in theductuli deferentia of every specimen exam-ined. Secretory material associated with thesperm mass is eosinophilic, bromphenol bluepositive, and PAS positive (Fig. 10B–D), andthese reactions are most intense in areasaround the border of the sperm mass thatlack sperm. In the latter areas the secretorymaterial is globular. The bromphenol bluereaction is less intense than for smooth muscleencasing the ductus deferens (Fig. 10C).Scattered reactions to alcian blue occuraround the sperm mass in specimens collected13 March, 30 May, 27 June, 27 July, and 17October. Numerous vacuoles are found be-tween the apical cytoplasm and the luminalsecretory material in a specimen collected 24May. The cytoplasm of the ductus deferensusually seems rather unreactive to histochem-ical tests (Fig. 10C–D), but in all seasons PASpositive, alcian blue positive, and bromphenolpositive material is located in the apicalcytoplasm. Transmission electron microscopy(TEM) reveals the presence in the apicalcytoplasm of secretory vacuoles containing adiffuse material, characteristic of carbohy-drates (Fig. 10E). Scanning electron micros-copy shows the tight packing of sperm(Fig. 10F).

The ampulla consists of the most distal 6–10 mm of the ductus deferens retroperitonealin the coelom, and another segment of similar


length that passes through the cloacal wallsbefore joining with the urethra. Externally,the coelomic ampulla is discerned by astraightening of the loops from the ductus

deferens. Internally, the ampulla is character-ized by irregular folded epithelial walls(Fig. 11A,B) until the final intramural one-fourth when the epithelium becomes regular

FIG. 10.—Micrographs of the ductus deferens of a 46.1 cm SVL male Agkistrodon piscivorus collected 13 March. (A)Light micrograph overview stained with hematoxylin-eosin. (B) Light micrograph detail of sperm in secretory materialstained with hematoxylin-eosin. (C) Light micrograph stained with bromphenol blue. (D) Light micrograph stained withPAS and alcian blue. (E) Transmission electron micrograph of luminal border. (F) Scanning electron micrographshowing sperm packed in the lumen. BB+, bromphenol blue positive; Ep, epithelium; Lu, lumen; Mu, muscularis;PAS+, periodic acid-Schiff’s reagent positive; SpSm, sperm in secretory material; Sn, sperm nucleus; Sp, sperm; Sr,serosa; Sv, secretory vacuoles.


again as the junction with the ureter isapproached. The cytoplasm of the anteriorthree-fourths of the ampulla is characterizedby numerous vacuoles, and the vacuoles alsooccur between the apices of the epithelial cellsand the secretory material encasing the spermmass (Fig. 11C,D). Sperm occur in theampulla throughout the year, although spermare scant in October and November speci-mens. The eosinophilic secretory material iselaborated from the apical cytoplasm(Fig. 11C). Throughout the year, luminalmaterial is intensely PAS positive, and smallPAS positive granules are scattered through-out the epithelium (Fig. 11D). Positive reac-tions with alcian blue and with bromphenolblue are also observed in the cytoplasm andluminal material, but these reactions are notas strong and pervasive as those for PAS.TEM confirms the presence of apical secre-tory vacuoles containing a diffuse substanceand a dense matrix containing sperm in thelumen (Fig. 11F). The uniformity of thesecretory matrix is revealed by scanningelectron microscopy (Fig. 11F).

Discussion of the efferent ducts.—Theproximal efferent ducts show a regression inactivity in June and July, although the ductusdeferens and ampulla show less seasonalvariation. The periods (spring and fall) whenthe proximal ducts are most highly hypertro-phied correspond to the fall and spring matingseasons of Agkistrodon piscivorus.

The two most comprehensive histologicalstudies on the efferent ducts of snakes were byVolsøe (1944) and Fox (1952). In Viperaberus, and apparently Natrix natrix, Volsøe(1944) indicated that the ductuli efferentesconsist of squamous epithelium and theproximal ductuli epididymides have columnarepithelium that has ‘‘irregular flames of longcilia.’’ At certain seasons the proximal ductuliepididymids were recorded to be filled witheosinophilic secretion granules. Volsøe (1944)also noted that proximal ductuli epididymidesnarrow without bifurcation into the distalductuli epididymides, which are longer, morecuboidal, generally ciliated, and lack secretorygranules. Fox (1952) reported similar resultsfor Thamnophis elegans.

These observations are at odds with ourobservations on Agkistrodon piscivorus. The

ductuli efferentes of A. piscivorus are clearlycuboidal, not squamous. Cilia appear to bepresent in the ductuli epididymides, but weawait the results from ultrastructural work toconfirm their presence and cytology. Thedistal segment is secretory, not the proximal.Finally, the distal portion clearly has moretubules than the proximal portion, indicatingdivision of the proximal ductuli epididymidesinto smaller units.

In lizards, the few studies that have beendone of the ductuli epididymides indicate thatthey arise from an extra-testicular rete testisand open into a large sinus inside theepididymis (Akbarsha et al., 2007; Averal etal., 1992; Fox, 1977). Production of secretorymaterial is not mentioned, but the efferentductules of Sitana ponticeriana are describedas spermiophagic.

Volsøe (1944) found that at ‘‘certain sea-sons’’ the epithelium of the epididymis ofVipera berus contains numerous small gran-ules which become blackened with osmiumtetroxide, which is an indicator of lipids. Fox(1952) stated the columnar epithelium of theepididymis of Thamnophis elegans is ‘‘proba-bly secretory’’ during certain seasons of theyear. Dufare and Saint Girons (1984) exam-ined the ductus epididymides of 89 species ofsquamates, including 35 species in sevenfamilies of snakes. They found five types ofsecretory activity in the main duct of theepididymis. Type 1 consists of large secretorygranules with a chromophilic central coresurrounded by a chromophobic vacuole, andthis type characterizes the Lacertidae. Types2–4 show decreasing size and density ofsecretory products, and Type 5 demonstratesno secretory activity. Type 5 characterized allspecies of snakes that were examined, includ-ing V. berus.

We found limited evidence of secretoryactivity in the ductus epididymis of Agkistro-don piscivorus. The scattered, diffuse reactionto carbohydrate and protein stains detected,however, does indicate that an ultrastructuralexamination might reveal product synthesis.

Neither Volsøe (1944) nor Fox (1952)reported secretory activity in the ductusdeferens. The ductus deferens is the storageorgan for sperm in male squamates, but theepithelium has been characterized as non-


FIG. 11.—Micrographs of the ampulla ductus deferentis of a 46.1 cm SVL male Agkistrodon piscivorus collected 13March (A,B,E,F), a 48.5 cm specimen collected 28 March (C), and 53.6 cm specimen collected 24 May (D). (A) Lightmicrograph overview stained with hematoxylin-eosin. (B) Light micrograph overview stained with bromphenol blue. (C)Light micrograph detail of sperm in secretory material stained with hematoxylin-eosin. (D) Light micrograph detailstained with PAS and alcian blue. (E) Transmission electron micrograph of luminal border. (F) Scanning electronmicrograph showing layer of secretory material in the lumen. Ep, epithelium; Lu, lumen; Mu, muscularis; Mv,microvilli; PAS+, periodic acid-Schiff’s reagent positive; Ppt, principle piece of the tail; SpSm, sperm in secretorymaterial; Sm, secretory material; Sn, sperm nucleus; Sp, sperm; Sr, serosa; Sv, secretory vacuoles; Va, vacuoles.


secretory (Fox, 1977). Our histochemical andultrastructural results on the ductus deferensof Agkistrodon piscivorus appear to be thefirst observations on secretory activity in thesquamate ductus deferens.

Volsøe (1944) noted that the ductus def-erens straightens before entering the cloacalwall, but stated that the ampulla ductusdeferentis occurs in the ureter. The areaindicated in his drawing is posterior to thejunction of the ureter and ductus deferens,which is actually the urogenital duct. This areais widened, and apparently was his criterionfor considering this area the ampulla. Theurogenital ducts on either side unite in Viperaberus, although their lumina remain separate.Fox (1952) followed the interpretation ofVolsøe (1944).

The occurrence of an ampulla ductusdeferentis consisting of the posterior end ofthe ductus deferens has been documentedhistologically in two species of lizards, Calotesversicolor (Akbarsha and Meeran, 1995) andSitana ponticeriana (Akbarsha et al., 2005),and one species of snake, Seminatrix pygaea(Sever, 2004). Our results clearly indicate thatan ampulla also occurs in the ductus deferensof Agkistrodon piscivorus. Sever (2004) foundno evidence of secretory activity in theampulla of S. pygaea, but Akbarsha et al.(2005) found that the ampulla of S. ponticeri-ana is divided into glandular and storageportions, and the storage portion was involvedin endocytosis and phagocytosis of deadsperm. We found no evidence of spermi-ophagy, but the anterior three-fourths of theampulla of A. piscivorus is involved inproduction of a carbohydrate secretion.

Renal sexual segment.—The renal sexualsegment (Rss) is an enlargement of the distalconvoluted tubule of the kidney in malesnakes and lizards, and may include thecollecting ducts and ureter as well. Thesecretions of the Rss may sustain and activatesperm (Bishop, 1959; Cuellar, 1966), providecourtship pheromones (Volse, 1944), formcopulatory plugs (Devine, 1975; Nilson andAndren, 1982; Ross and Crews, 1977), and/orhave other purposes generally associated withseminal fluid (Prasad and Reddy, 1972).Several female lizards are also known topossess a hypertrophy of the nephron similar

to males (Sever and Hopkins, 2005). Numer-ous histological studies on the Rss of squa-mates have been done since the first suchreport by Gampert (1866), and this literaturehas been reviewed by Sever et al. (2002, 2008)and Sever and Hopkins (2005). Amongreptiles, the presence of a Rss can beconsidered a synapomorphy at least forSquamata and probably for Lepidosauria.Sever et al. (2008) studied the histology andultrastructure of the Rss of Agkistrodonpiscivorus. This study was the first to describethe ultrastructure of the Rss in a family ofsnakes other than the Colubridae.

Seasonal variation occurs in Rss diameterand epithelial height, with hypertrophy inspring, a marked reduction in size Maythrough July, and hypertrophy again in latesummer and early fall (Fig. 12A). Portions ofthe Rss are easily distinguished from othertubules in the kidney by their relatively largersize and staining characteristics. When hyper-trophied, the epithelium is columnar, uni-formly eosinophilic, and possesses basal nuclei(Fig. 12B). In contrast, proximal convolutedtubules are cuboidal and basophilic. Rsstubules react strongly for proteins withbromphenol blue (Fig. 12C) and are largelyPAS positive (Fig. 12D) but areas of alcianblue activity also occur. Proximal convolutedtubules are alcian blue positive and reactweakly with bromphenol blue.

Transmission electron microscopy of theRss shows the cytoplasm filled with electron-dense secretory granules (,2 mm dia), largeempty vacuoles, and small vesicles containinga diffuse material (Fig. 12E). Secretory gran-ules as well as the small vesicles abut upon thelarge vacuoles in some areas, and the vacuoleshave connections to the intercellular canalic-uli. Rough endoplasmic reticulum and Golgicomplexes are found in the perinuclear areaand are associated with condensing vacuolesthat may be irregular in shape. Release of thesecretory products could involve both apo-crine and merocrine processes. Scanningelectron microscopy reveals how secretorygranules fill the cytoplasm of active Rss(Fig. 12F).

In June and July, the Rss go through aperiod where secretory activity is reduced andgland diameter and epithelial height is obvi-


FIG. 12.—Data and micrographs that concern the renal sexual segment (Rss) of male Agkistrodon piscivorus.Specimens in the micrographs include a 52.2 cm SVL male collected 23 February (B,C,E), a 46.1 SVL individualcollected 13 March (D), and a 56.1 cm specimen collected 22 November (F). (A) Seasonal variation in the diameter andepithelial height of the Rss; error bars represent standard variation. (B) Light micrograph overview of Rss stained withhematoxylin-eosin. (C) Light micrograph overview of Rss stained with bromphenol blue. (D) Light micrograph overviewstained with PAS and alcian blue. (E) Transmission electron micrograph of luminal border. (F) Scanning electronmicrograph showing a tubule packed with secretory granules. Ac, apical cytoplasm; BB+, bromphenol blue positive; Cv,condensing vacuole; Ep, epithelium; Ic, intercellular canaliculi; Lu, lumen; Nu, nucleus; PAS+, periodic acid-Schiff’sreagent; Pct, proximal convoluted tubules; Rss, renal sexual segment; Sg, secretory granules; Va, vacuoles.


ously decreased. The Rss epithelium isdecidedly basophilic, and the positive PASand bromphenol blue reactions are limited toscattered granules. TEM of the June and Julysamples reveals the loss of mature, electron-dense secretory granules and smaller secreto-ry vesicles, and the reduction in size andnumber of vacuoles.

Studies on lizards to date have reported thatthe Rss is hypertrophied only during periodsof sexual activity and cannot be distinguishedfrom adjacent tubular regions during sexualquiescence (Fox, 1977; Gabri, 1983; Sever andHopkins, 2005). In snakes, complete regres-sion has not been reported (Fox, 1977; Severet al., 2002). Johnson et al. (1982) reportedthat granules are present in the epithelial cellsof the Rss in Alabama Agkistrodon piscivorusin April and in the lumina from Marchthrough October. Although some PAS positiveand bromphenol blue positive granules werepresent in our specimens from June and July,secretory activity is dramatically reduced fromother times of the year.

COMPARATIVE MORPHOLOGY: FEMALES

Gross Morphology

The gross female reproductive morphologyof Serpentes was reviewed by Blackburn(1998). Figure 13 depicts the undissectedand dissected anatomy of female Agkistrodonpiscivorus. The oviduct is positioned ventral(when intestine is full with fecal waste) orlateral (when intestine is emptied) to theintestines within the body wall (Fig. 13A; aspecimen with a full intestine is depictedhere). The anterior regions are positioneddorsally to the ventral abdominal vein, whichis associated with fat at its periphery(Fig. 13A). After removal of the ventralabdominal vein and fat, the anterior portionsof the oviduct can be observed clearly(Fig. 13B), and four regions can be discerned(Fig. 13B): an enlarged posterior regiontermed the vaginal pouch (Ludwig and Rahn,1943), a narrow tubular region termed thenon-glandular uterus, a slightly enlarged andsometimes-folded region termed the glandu-lar uterus, and a highly folded anterior regiontermed the infundibulum. Each oviduct issuspended in the body cavity by the dorsal

mesentery. The ovaries, filled with small orenlarged macrolecithal follicles depending onreproductive condition (non-reproductive de-piction in Fig. 13B), lie lateral to theirrespective oviduct (Fig. 13B).

Histology (Reviewed from Siegel and Sever,2008a,b)

Infundibulum and sperm storage tubules.—The most anterior oviductal region in A.piscivorus is the infundibulum that opensanteriorly into the body cavity through theostium. The infundibulum possesses a thinmucosa with a simple squamous epithelium(Fig. 14A). Small invaginations into the laminapropria form simple tubular glands in theinfundibular region (Fig. 14A). Ultrastructur-ally these glands are undifferentiated from theepithelium lining the lumen. Ciliated cells areobserved sporadically throughout the epithe-lium with secretory cells that produce mainlylipoid materials. Ciliated cells possess basallypositioned nuclei and are packed with apicallypositioned mitochondria interacting with basalbodies anchoring the cilia. Lipoid materialproduced in the secretory cells is eitherdiffuse and unorganized or tightly packed intodenser lipid droplets. Smooth endoplasmicreticulum is abundant in the cytoplasm ofthese secretory epithelial cells. In Seminatrixpygaea, another snake investigated with ultra-structural analysis, only organized lipoidmaterial was produced by the infundibulumalong with protein packed secretory vacuoles(Sever et al., 2000). As with the epithelium ofthe entire oviduct, epithelial cells are adheredapically by tight junctions, followed basally bydesmosomes. Two thin muscle layers encom-pass the mucosa, an interior circular muscleand exterior longitudinal muscle. As in othersquamates investigated (for review see Black-burn, 1998) this muscularis externa is contin-uous throughout the whole length of theoviduct. The entire oviduct and its muscularlayers are encompassed by a layer of visceralpleuroperitoneum, which communicates withthe parietal pleuroperitoneum lining thedorsal body wall via the dorsal mesentery(for review on mesenteries in snakes seeBlackburn, 1998).

The most posterior tubular glands of theinfundibulum possess a wider terminal por-


tion and invade deeper into the lamina propria(Fig 14B). These glands are sperm storagetubules. Distally, the glands are comprised ofnumerous ciliated cells; terminally, however,these ciliated cells decrease in concentrationand secretory cells take over the majority ofthe epithelium (Fig. 14C). When sperm arepresent in sperm storage tubules they alignthemselves in parallel arrays with their nucleifacing the terminal ends of the glands(Fig. 14C). This condition is consistent withwhat is observed in the snakes Thamnophissirtalis (Halpert et al., 1982; Hoffman andWimsatt, 1972) and Diadophis punctatus(Perkins and Palmer, 1996), and the lizardsAcanthodactylus scutellatus (Bou-Resli et al.,1981) and Scincella lateralis (Sever andHopkins, 2004). However, one snake speciesinvestigated, S. pygaea, exhibits sperm withnuclei facing the openings of the spermstorage tubules (Sever and Ryan, 1999).

Infundibular sperm storage tubules werealso observed in the snakes Thamnophiselegans (Fox, 1956), Vipera aspis (Saint-Girons, 1957, 1959), Cerastes cerastes (Saint-Girons, 1962a,b), multiple members of thefamilies Typhlopidae and Leptotyphlopidae(Fox and Dessauer, 1962), and Tantillacoronata (Aldridge, 1992). In studies wherehistochemical analysis was conducted, thesecretory material produced by infundibularsperm storage tubules was always positive forneutral carbohydrates (Hoffman and Wimsatt,1972; Perkins and Palmer, 1996; Sever andRyan, 1999), including sperm storage tubulesof A. piscivorus. Thamnophis sirtalis parietalis

studied by Halpert et al. (1982) also possessedneutral carbohydrate secretions in the infun-dibular sperm storage tubules. However, it isunclear whether or not the epithelium of thetubules is synthesizing this material. Structureof sperm storage tubules was typically de-scribed as simple to complex tubular except inTantilla coronata where sperm storage tubulesare more alveolar (Aldridge, 1992).

Glandular uterus.—The glandular uterus ischaracterized by having a thicker muscularisand lamina propria than the infundibulum,and scattered invaginations of the epitheliumlining the lumen forming simple tubularglands (Fig. 14D,E). The simple cuboidalepithelium of the mucosa (Fig. 14E) secretesan acidic mucoid substance whereas theuterine glands secrete a protein along theirwhole length. Ultrastructurally, the epitheli-um bordering the lumen of the uterus iscomposed primarily of secretory cells thatproduce an electron lucent material releasedapocrinely. Ciliated cells are intermixed ran-domly with the secretory cells. The glands ofthe epithelium are composed strictly ofsecretory cells (Fig. 14F). Cells of this epi-thelium are packed with an electron densesecretory material (Fig. 14F) released by amerocrine type secretory mode. During syn-thesis and release of the protein materialproduced by these glands synthetic organelles,including mitochondria and rough endoplas-mic reticulum, are abundant throughout thecell and perinuclearly respectively.

The only other snake uterus that wasinvestigated thoroughly with ultrastructural

FIG. 13.—Gross morphology of the female Agkistrodon piscivorus genital system. (A) Undissected anatomy. (B)Dissected anatomy. Cl, cloaca; Ft, fat; Gu, glandular uterus; Inf, infundibulum; Int, intestine; Ngu, non-glandularuterus; Ov, ovary; Vp, vaginal pouch.


techniques is that of Seminatrix pygaea (Severet al., 2000). The uterine glands in this snakewere basically undifferentiated invaginationsof the epithelium lining the lumen (Sever et

al., 2000), whereas in Agkistrodon piscivorusthe uterine glands are highly differentiatedfrom the luminal border. With the shift fromoviparity to viviparity, the activity and number

FIG. 14.—Histology of female Agkistrodon piscivorus genital system. (A) Anterior infundibulum of a specimen fromNovember. (B) Sperm storage region of the posterior infundibulum of an November specimen. (C) TEM micrograph ofsperm at the terminal end of a sperm storage tubule in a May female. (D) Overview of the glandular uterus in aNovember specimen. (E) High magnification of D showing structural components of the glandular uterus. (F) TEMmicrograph of cellular components in the epithelium of a uterine gland from an April female. (G) Histological overviewof the non-glandular uterus from a June specimen. (H) Sperm in the lumen of the non-glandular uterus of a May female.(I) Cellular components of the epithelium lining the lumen of the non-glandular uterus in a May female. (J) Histologicaloverview of the vaginal pouch in an April specimen. (K) Sperm associated with eosinophilic secretory material in thevaginal pouch of a July female. (L) Cellular components of the epithelium lining the lumen of the vaginal pouch in aMarch female. Ab, apocrine bleb; Bb, basal bodies; Bv, blood vessel; Ep, epithelium; Ig, infundibular gland; Lp, laminapropria; Lu, lumen; Mpt, mid principle piece of the sperm tail; Ms, muscularis; Mt, mitochondria; No, nucleolus; Nu,nucleus; Ru, rugae; Sm, secretory material; Sn, sperm nucleus; Sp, sperm; Ssts, sperm storage tubules; Sv, secretoryvacuole; Ug, uterine gland.


of uterine glands is hypothesized to decrease(Blackburn, 1998). Considering uterine glandsare responsible for eggshell formation (Palmeret al., 1993), the act of not producing aneggshell in viviparous species would tend tosupport this hypothesis. However, it is clearthat in some snake species, including A.piscivorus, uterine glands may still play amajor roll in reproduction. Because viviparityhas evolved multiple times independentlyin squamates (Blackburn, 1985; Blackburn,1999; Shine, 1985), uterine gland morphologyand activity would vary between taxa that haveindependently evolved live-bearing capabili-ties. Uterine gland variation, and associationwith parity mode has currently not beeninvestigated in snakes, however, Girling et al.(1998) demonstrated ultrastructural variationin lizards that possessed differing modes ofparity. Interestingly, uterine glands of thelizard Saltuarius wyberba are most similar interms of ultrastructure to uterine glands of A.piscivorus (Girling et al., 1998). However, thislizard is oviparous and produces a parchment-type eggshell (Girling et al., 1998). Otherstudies on squamate uteri that demonstrateunique uterine morphology due to paritymode shift include those of Girling (2002)and Thompson et al. (2006).

Unfortunately, in our two-year collectingefforts, pregnant Agkistrodon piscivorus werenot obtained. Thus, discussion of the pregnantuterus is limited to light microscopy from amuseum specimen and from other historicalstudies. It appears that in late gestationuterine glands are still actively synthesizingand releasing secretory material in A. pisci-vorus. This is evidenced by the presence ofintensely staining protein granules in theglandular uterus in late August. Thus, uterinegland activity most likely continues through-out gestation. As reviewed by Blackburn(1998) in a variety of squamates, the pregnantuterus of A. piscivorus is highly reduced interms of the oviducal lining (e.g., muscularisthickness, lamina propria thickness, andheight of epithelial cells). Similar to a colubrid(Thamnophis sirtalis; Hoffman, 1970) andanother crotaline (Crotalus viridus; Rahn,1942), the uterus of A. piscivorus becomeshighly vascularized during pregnancy. How-ever, no statistical analyses were conducted to

assess this quantitatively, and as Blackburn(1998) states, the quantification of vasculari-zation in the uterus from histological sectionsis difficult.

Microscopy of fetal membranes and uterus/extra-embryonic membrane association werenot conducted in our investigations on repro-ductive morphology in Agkistrodon piscivorusdue to the lack of pregnant specimens.However, recent literature is available on thestructure and evolution of placentation/vivi-parity in snakes, particularly thamnophiines.For information regarding these topics werefer readers to Blackburn (1998) and Black-burn and Flemming (2008).

Non-glandular uterus.—The non-glandularuterus, sometimes called the vaginal tube orsegment (Saint-Girons, 1973), anterior vagina(Hoffman and Wimsatt, 1972), or furrowedportion of the uterus (Aldridge, 1992; Hal-pert et al., 1982), was previously described inthe Colubridae and Viperidae. We term thisregion the non-glandular uterus because likethe glandular uterus, embryos invade thisregion during embryogenesis. However, un-like the glandular uterus, no tubular glandscan be observed in the lamina propria of thisregion (Fig. 14G). In Agkistrodon piscivorusthis oviducal segment possesses a very thickmuscularis, a thin lamina propria comparedto the glandular uterus (Fig. 14G,H), and asimple cuboidal epithelium that becomescolumnar when moving posteriorly downthe oviduct (Fig. 14H). The thick exteriorlayer of longitudinal muscle creates thelongitudinal folds that are observed grossly(Fig. 14G). The majority of the epithelium iscomposed of ciliated cells, and secretory cellscan be observed scattered throughout(Fig. 14I) with acidic mucous secretions inthe apical portion of the cells. Ultrastructur-ally, the secretions are packed in electronlucent vacuoles and are released via anapocrine type secretion (Fig. 14I). Duringtimes of high secretory activity, mitochondriaand Golgi bodies are observed apical to thenucleus in cells that have recently secretedtheir product. Nuclei of ciliated and secreto-ry epithelial cells are located in a basalposition (Fig. 14I). This region of the oviductcontains numerous mast cells in the laminapropria.


To our knowledge, the only other snakewhere a similar region has been describedultrastructurally has been Seminatrix pygaea(Sever et al., 2000). In that study, this regionwas termed the vagina. However, after a closelook at the similarities of these regions, it isprobable that the non-glandular uterus inAgkistrodon piscivorus and vagina in S.pygaea are in fact homologous. As mentionedabove, Halpert et al. (1982) also described thisregion of the oviduct using basic histology inThamnophis sirtalis. They believed that se-cretions from the non-glandular uterus inter-acted with sperm during mating and facilitat-ed transport of sperm anteriorly up theoviduct. This is not likely in A. piscivorusbecause sperm were never associated withacidic mucoid secretions like those from theepithelium of the non-glandular uterus whilein the female reproductive tract. The secre-tions associated with sperm observed in thefemale oviduct were more similar to thoseobserved in the secretory epithelium of theRss in males (Sever et al., 2008). We questionwhether the secretory epithelium of the non-glandular uterus in T. sirtalis could really beproducing a carrier matrix for sperm.

Vagina.—The vagina in Agkistrodon pisciv-orus, and all viperids, is an interestingoviductal region. Histologically, no compara-ble region has been described. This region ofthe oviduct possesses a single layer of highcolumnar epithelial cells that are primarilysecretory in function (Fig. 14J,K,L). Ciliatedcells are occasionally interspersed with themajority secretory cells that produce anelectron lucent material with an electrondense core. This material is composed ofacidic and neutral carbohydrates and coats theapical surface of the epithelium after secretionin a merocrine manner. Around the basallypositioned nuclei (Fig. 14L) there exists largeelectron dense lipid droplets. Between adja-cent epithelial cells, and the epithelium andthe basal lamina, exist large vacuolization’s ofthe intercellular canaliculi. Plasma-like cellsare often found inhabiting these vacuoles.These plasma-like cells are not the onlyimmune cells inhabiting this region of theoviduct, for like the non-glandular uterus,mast cells are found in very high concentra-tions in the wide lamina propria (Fig. 14J,K)

of the vaginal pouch. The muscularis of thevaginal pouches is thick compared to allregions of the oviduct except the non-glandu-lar uterus (Fig. 14J).

It is clear through literature searches (Gabeand Saint-Girons, 1965; Sanchez-Martınez,2007), personal investigation by D. S. Siegelon other snake families, and conversationswith R. D. Aldridge, that the vaginal pouch ofvipers is actually what is traditionally calledthe urodeum in other colubroid snakes. Thus,the vaginal pouches are actually the anterior-most portion of the cloaca. Anatomically, thisis an interesting discovery because like othervertebrates that possess a true vagina, what istraditionally called the vagina of vipers isderived from cloacal tissue, instead of coelo-mic mesothelial tissues, which gives rise to theinfundibulum and uterus (Raynaud and Pieau,1985). Therefore, in a sense, vipers have a‘‘true’’ vagina. Functionally, the purpose of ahighly extended urodeum like that seen invipers is unknown. However, electron andlight micrographs of the vaginal pouch epi-thelium reveals similarities to the vaginalepithelium of some mammals (e.g., themicrovillus, columnar, mucin secretory epi-thelium of rodents; Lamb et al., 1978). It isunclear if the squamate urodeum is capable oftactile function, or if the hemipenis of vipersprotrude into the pouches; however, Pope(1941) describes the male hemipene reachingthe opening of the oviduct (i.e., protrudingthrough the urodeum) in a colubrid and Cope(1898) actually termed the cloaca the vaginabecause of its function in accepting a copula-tory organ. Like Ludwig and Rahn (1943), wehypothesize that the hemipenes of male vipersactually extend into the vaginal pouches oftheir conspecific females. Thus, vipers mayactually possess a correctly defined anatomicaland functional vagina.

Oviductal regions.—From an extensivereview of literature and histological examina-tion of snakes and lizards Blackburn (1998)simplified the squamate oviduct into threeregions: (1) a posterior vagina, (2) middleuterus, and (3) anterior infundibulum. Theregion termed the non-glandular portion ofthe uterus in Agkistrodon piscivorus (Siegeland Sever 2008a) is undoubtedly homologouswith the region Blackburn (1998) terms the


vagina, as he discusses in his lengthy review.This was based on similar histology andposition of this oviducal region. Thus, theterminology for the oviduct in Viperidae variesfrom Blackburn (1998) because of the pres-ence of an extension of the cloaca (vaginalpouches; see above), which may be a synapo-morphy of Viperidae. In conclusion, theoviducts of squamates all appear to have threehomologous regions (posterior, middle, andanterior), however, variation within and be-tween these regions occurs (e.g., uterine glandmorphology, Girling et al., 1998; spermstorage receptacle morphology and location,Eckstut et al., 2009).

REPRODUCTIVE CYCLE

In reviewing the reproductive cycle ofAgkistrodon piscivorus it is necessary toconsider males and females individually be-cause of the dissociation of the timing ofreproductive events. This dissociation createsa model in which the only event that has to becoordinated between males and females iscopulation. Although copulation is obviouslyan integral part of the reproductive cycle,published accounts of actual copulation eventsare lacking. Beyer (1893) provides evidence ofa male and female in coitus in the spring;however, only the female was captured.Positive identification of the conspecific as amale was not achieved. Wharton (1966) notesbisexual pairing of A. piscivorus in almost allmonths of the year and suggests A. piscivoruscould be continually mating throughout theactive seasons. Siegel and Sever (2006, 2008a)note sperm in the posterior portion of thefemale reproductive tract (vaginal pouch andnon-glandular uterus) in the late summer–falland spring before ovulation has occurred.These data suggest a biannual mating season,at least in Louisiana populations of A.piscivorus, and are consistent with datapresented for A. piscivorus reviewed byAldridge and Duvall (2002).

Males

To this date, four studies have trackedreproductive events in male Agkistrodonpiscivorus by utilizing histology of the testisand Rss (Graham et al., 2008; Gribbins et al.,2008; Johnson et al., 1982; Sever et al., 2008).

Sever et al. (2008) and Gribbins et al. (2008)studied specimens from the most southerndistribution, using snakes from southeasternLouisiana. The most northern study wasconducted by Graham et al. (2008) at theconfluence of Morning Creek with the FlintRiver in Georgia, while Johnson et al. (1982)investigated a slightly more southern popula-tion in Perry County, Alabama.

Testicular cycle.—In the population of maleAgkistrodon piscivorus investigated in Louisi-ana, development of seminiferous tubuleepithelia can be observed early in the springafter hibernation and continues until July(Figs. 4,5). During this time period, allmeiotic stages of spermatocyte advancement,spermiogenesis, and spermiation can be ob-served in the testes. Subsequent to a quies-cent period in mid-summer (July; Figs. 4,5),the development of the seminiferous tubulesreoccurs and a second term of sperm forma-tion can be observed (Figs. 4,5; see Germ celldevelopment strategy for in-depth review).The histology and data analysis robustlysupports that male A. piscivorus in Louisianapossess a biannual spermatogenic cycle. Toour knowledge, this is the first time this hasbeen reported in a North American crotaline.

Johnson et al. (1982) and Graham (2008)reported that male Agkistrodon piscivorus areonly spermatogenic in the summer–fall timeperiod in Alabama and Georgia respectively.This fits what has also been described for allother temperate pitvipers in North America(for review see Aldridge and Duvall, 2002).

Rss cycle.—Similar to the spermatogeniccycle of male Agkistrodon piscivorus fromLouisiana, two peaks of major secretorymaterial synthesis and release can be observedin the Rss; one in the spring and one in thelater summer–fall (Sever et al., 2008; seeabove). A spring and late summer–fall peak ofRss activity was also described by Johnson etal. (1982) from male A. piscivorus in Alabama.In Georgia a significant peak of activity, basedon Rss diameter was not observed (Graham etal., 2008). However, a noticeable non-signif-icant peak was observed in the late summer–fall (Graham et al., 2008). We believe that thisrepresents one peak of major Rss activity inmale A. piscivorus from Georgia, and thatstatistical significance would be revealed if


epithelial height was used for monthly com-parison instead of tubular diameter. Tubulardiameter varies less from month to monthwhereas epithelial height is the more impor-tant indicator of Rss activity (e.g., directlymeasuring the amount of secretory materialproduced in the Rss).

Geographic variation in reproductive events.—From the above analysis, a trend can beobserved while moving northeast from popu-lations of Agkistrodon piscivorus in southeast-ern Louisiana. In northerly populations pri-mary and secondary sexual character activityin males is pushed to one end of the activeseason, in this case, summer and fall respec-tively. Agkistrodon piscivorus males fromsoutheastern Louisiana exhibit biannual re-productive events, the most northern popula-tion studied to date in Georgia has annualreproductive events, and the population inAlabama between these two extremes istransitory (Fig. 15). Increases in the produc-tion of material in the excurrent ducts (seeabove) also coincide with the mating seasons(fall and spring) in Louisiana A. piscivorus.However, no data is available from theAlabama and Georgia populations for com-parison.

In general, spermatogenesis and Rss hyper-trophy are androgen dependent. Thus, varia-tion in the timing of male reproductive eventspresents two very interesting questions:

(1) Are two peaks of testosterone necessaryfor biannual spermatogenesis observed inthe Louisiana population of Agkistrodonpiscivorus? Fall and spring peaks oftestosterone have been recorded in otherNorth American crotalines, Crotalusatrox (Schuett et al., 2005; Taylor et al.,2004) and C. scutulatus (Schuett et al.,2002), that do not have biannual sper-matogenesis. Typically, spring testoster-one peaks have been interpreted asincreases in testosterone to benefit onlymating activity (Schuett et al., 2005;Taylor et al., 2004), and thus, thetestosterone peaks in the spring are notas high as those observed in the summer–fall (Schuett et al., 1997; Taylor et al.,2004) when spermatogenesis is occur-ring. However, in Louisiana A. piscivorusspring testosterone peaks (if observed)may also be involved in stimulatingspermatogenesis and therefore may beas high as those observed in the fall.

FIG. 15.—Geographical variation in male secondary sexual characteristics and sperm formation in Agkistrodonpiscivorus from Louisiana (A), Alabama (B), and Georgia (C). White rectangles represent seasonality ofspermatogenesis. Black rectangles represent seasonality of major secretory activity in the renal sexual segment.


Interestingly, A. piscivorus, which hasbeen recorded as a biannual breeder(although published accounts of matingactivity are lacking; see Graham et al.,2008, and above), does not have a springpeak of testosterone in the Alabama andGeorgia populations studied (Graham etal., 2008; Johnson et al., 1982). Thissupports the hypothesis that A. piscivorusmay not be a biannual breeder in allportions of its range (Graham et al., 2008).

(2) Given that the Rss is androgen depen-dent (Prasad and Sanyal, 1969), how isspring hypertrophy maintained withoutan increase in testosterone in the Ala-bama population of Agkistrodon pisciv-orus and many other North Americancrotalines that exhibit biannual Rss hy-pertrophy (Aldridge and Duvall, 2002)?A hypothesis for this is that testosteroneis effective in stimulating Rss secretionsynthesis, however, it is not necessary formaintenance after initial stimulation (Al-dridge, personal communication). Thisalso could account for the lack of atestosterone peak in the spring in bian-nually reproducing populations.

Traditionally the Agkistrodon piscivoruscomplex is made up of three subspecies(Gloyd and Conant, 1990): an eastern variety(A. p. piscivorus), a western variety (A. p.leucostoma), and a Florida variety (A. p.conanti). However recent molecular data andphylogeographic analysis of the A. piscivoruscomplex reveal that the only independentlyevolving lineages in the complex are thecontinental members (A. p. leucostoma andA. p. piscivorus) and the Florida member (A.p. conanti; Guiher and Burbrink, 2008;Douglas et al., 2009). Considering the varia-tion in reproductive timing reviewed above(from the southwest to the northeast) is withinthe genetically similar continental members, itis hypothesized that variation in reproductiveattributes is due to local adaptation, and notevolutionary history. Out of the four repro-ductive cycle patterns described for NorthAmerican crotalines (Aldridge and Duvall,2002), it appears that males from the A.piscivorus complex are utilizing the ‘‘Temper-ate Zone Primitive Pattern’’ in Louisiana

(Spring and fall activity), ‘‘Temperate ZoneDerived Pattern’’ in Georgia (summer–fallactivity), and a pattern that falls somewhere inbetween the former two in Alabama (Fig. 15).However, no previous patterns described abiannual spermatogenic cycle as seen inLouisiana A. piscivorus in North Americancrotalines. Similar results have been present-ed on another North American crotaline,Crotalus oreganus, where the timing ofreproductive events is variable across therange of genetically similar populations (Al-dridge, 2002; Ashton and de Queiroz, 2001;Pook et al., 2000).

Females

Vitellogenesis.—As in most North Americancrotaline snakes (for review see Aldridge andDuvall, 2002), vitellogenesis commences inthe later summer–early fall in Agkistrodonpiscivorus (Siegel and Sever, 2008b). Thispattern is observed by the noticeable increasein follicular size beginning around October inLouisiana A. piscivorus. Follicles cease fur-ther enlargement during winter hibernation inLouisiana females, which lasts from approxi-mately late November to late February.Follicular growth recommences in spring afteremergences from over-winter hibernacula.This is consistent with observations from tenspecies of North American crotalines (forreview see Aldridge and Duvall, 2002).However, Taylor et al. (2004) and Schuett etal. (2004) note a caveat to this pattern withdata that they propose support vitellogenesiscommencing in the spring in a population ofCrotalus atrox in Arizona. Aldridge et al.(2008) review this controversy and proposethat too few specimens may have beenexamined from the fall in the C. atrox studiedto make the above conclusions.

Although data are lacking for hormonalcontrol of vitellogenesis in Agkistrodon pisciv-orus, Bonnet et al. (1994; 2001), Schuett et al.(2004), and Taylor et al. (2004) provide someinsight from studies on Vipera aspis (Viper-inae) and Crotalus atrox respectively. Thesestudies found the progress of vitellogenesis isconcurrent with elevated estradiol levels.Vitellogenesis terminates with a decrease inestradiol and a subsequent increase in pro-gesterone.


Oviducal cycle.—The glandular uterus andsperm storage tubules are the only two regionsof the oviduct that exhibit dramatic seasonalchanges in female Agkistrodon piscivorus(Siegel and Sever, 2008b). This is similar todata published for a colubrid, Seminatrixpygaea (Sever et al., 2000). The vaginal andanterior infundibular portion of the A. pisciv-orus oviduct exhibits no seasonal variation.Figure 16 graphically depicts the entire ovi-ducal cycle of A. piscivorus females.

Initial follicular growth in Agkistrodonpiscivorus occurs in the fall (approximatelymid-September to October), coincident withthe fall mating season (evidenced by spermpresence in the posterior oviduct of females).During this time secretory activity increases inthe epithelium of the non-glandular uterusand the epithelium lining the lumen of theglandular uterus. Concurrent to the increaseof secretory activity of the epithelium liningthe lumen of the mid-oviduct, simple tubularglands of the glandular uterus and posteriorinfundibulum (sperm storage tubules) beginsynthesizing secretory material that was pre-viously not observed. While the epitheliumlining the lumen of the uterine regionsincreases the synthesis and release of mucous(Fig. 14I), tubular glands of the glandularuterus produce an electron dense protein(Fig. 14F). Synthetic organelles (rough endo-plasmic reticulum, Golgi bodies) also increasein abundance in the epithelial cells at this time

(Fig. 14F,I). Sperm storage tubules synthesizean electron lucent neutral carbohydrate dur-ing this stage of the oviducal cycle. Theseobservations are consistent with what occursin the infundibulum and uterine regions ofSeminatrix pygaea, except that uterine glandsof the S. pygaea uterus exhibit minor seasonalchanges in terms of secretory activity (Sever etal., 2000). Secretory material synthesis andrelease continues in the oviduct and varieslittle until the spring. After hibernation thematerial synthesized by sperm storage tubulesbecomes electron dense (Fig. 14C), an occur-rence that has not been noted in any otherinvestigations on snakes.

The next dramatic change in the oviducalcycle does not occur until ovulation, after finalfollicular growth and the spring matingseason. At this time the synthesis of secretionsin the sperm storage tubules ceases. In othersnakes investigated (Thamnophis sirtalis andSeminatrix pygaea), the lack of secretoryactivity coincides with the disappearance ofsperm (e.g., Halpert et al., 1982; Sever andRyan, 1999). However, residual amounts ofsperm remain in Agkistrodon piscivorus (seenext section).

Uterine glands continue to remain active inAgkistrodon piscivorus until parturition. Atthis time they regress in activity and size incomparison to their previous state prior tovitellogenesis. Parturition, which occurs in thelater summer–early fall in A. piscivorus (Allen

FIG. 16.—The oviducal cycle of female Agkistrodon piscivorus and its relation to the vitellogenic cycle. Measurementsare equal to the greatest follicular diameter during different stages of the vitellogenic cycle. Pvod, sperm present in theposterior oviduct; SST sperm, sperm present in the sperm storage tubules; SST activity, secretory activity of spermstorage tubules; Utg, secretory activity of uterine glands.


and Swindell, 1948; Beyer, 1893; Burkett,1966; Conant, 1933; Ford, 2002; Ford et al.,2004; Funk, 1964; Gloyd and Conant, 1990;Wharton, 1960), marks the end of the oviducalcycle in a reproductive year. Offspring num-ber ranges from 5–20 (Dundee and Rossman,1989). Female A. piscivorus is hypothesized tobreed less than annually (Wharton, 1966), as ishypothesized for many temperate crotalinesnakes (Klauber, 1972), and thus activity inthe oviduct does not increase again until atleast the fall of the subsequent year. Collec-tion data on Louisiana A. piscivorus isconsistent with this hypothesis with approxi-mately half of the females being reproductivein a given year.

Vipera aspis (Bonnet et al., 1994, 2001) andCrotalus atrox (Schuett et al., 2004; Taylor etal. 2004) provide insight on the hormonalcontrol of oviducal activity and parturition inviviparous vipers. In both viperid speciesgestation is concurrent with high levels ofprogesterone, which is undoubtedly the prox-imate stimulus for oviducal development to-wards gestation. At the timing of parturition,progesterone levels decrease, however, Bonnetet al. (2001) notes that exogenous supplemen-tation of progesterone before parturition doesnot delay birth. Thus, the decrease of proges-terone alone is not the proximate stimulus toparturition (Bonnet et al., 2001).

Sperm transport and storage.—The occur-rence of sperm storage in North Americancrotalines, and the potential costs and benefitswas reviewed extensively by Schuett (1992).Sperm transport and storage in colubrids wasdescribed by Fox (1956) and Halpert et al.(1982), and was reviewed in viperines bySaint-Girons (1973). In those studies and therecent investigations on sperm transport andstorage in Agkistrodon piscivorus (Siegel andSever, 2008a), sperm were found to be locatedin the posterior uterus shortly after copula-tion. However, histology of the mid andanterior oviduct revealed that sperm weremigrating from the posterior uterus to infun-dibular sperm storage tubules. Sperm storagein the posterior uterus is thus hypothesized tobe an artifact of mating activity. Becausesperm seem to remain in the posterior uterusfor an extended period of time (up to the timeof ovulation), two sites of sperm storage in

snakes could be proposed (Eckstut et al.,2009). In A. piscivorus, however, sperm in theposterior uterus were undergoing degradationshortly before ovulation, which supports thehypothesis of sperm presence because ofmating activity. Since the posterior uterushas such a small luminal diameter comparedto the rest of the oviduct (and the highlyenlarged vaginal pouch), it may be possiblethat a morphological bottleneck causes somesperm to get stuck in this area. We alsosuggest that some sperm are inherently poorand cannot migrate anteriorly. Nilson andAndren (1982) reported in Vipera berus thatthat the posterior uterus forms a non-gelati-nous copulatory plug preventing the migrationof sperm from subsequent matings to infun-dibular sperm storage tubules by uterinecontraction shortly after copulation. Theyproposed and provided evidence, that secre-tions from the renal sexual segment causedthis uterine contraction.

As mentioned above, in Thamnophis sirtalisand Seminatrix pygaea sperm are evacuatedfrom sperm storage tubules after ovulation.However, in Agkistrodon piscivorus femalesthis is not the case (Siegel and Sever, 2008a).Sperm remain in the sperm storage tubulesthrough pregnancy, the inactive season, and allthe way to the month before mating activity inthe subsequent mating season. At this point thesperm disappear. No evidence for how thisoccurs was observed. This presents an inter-esting question of the function of secretionsfrom sperm storage tubules. Because secretoryactivity in sperm storage tubules only occursduring vitellogenesis, and thus the matingseasons, we believe that this fact providesevidence that sperm storage tubule secretionsmay function in only attracting sperm. How-ever, evidence that supports the continuedviability of sperm in sperm storage tubules afterovulation and cessation of secretory activity isneeded to support this hypothesis.

Almeida-Santos and Salomao (1997, 2002)proposed that contractions of the uterus trapsperm in the posterior region of the oviductafter mating in all temperate pitvipers. How-ever, from the above information, it is clearthat sperm in all snakes investigated, includingAgkistrodon piscivorus, which was included inthe work by Almeida-Santos and Salomao


(2002), migrate anteriorly shortly after mating.Therefore, we believe this phenomenon ob-served by Almeida-Santos and Salomao (1997,2002) serves another function and could besimilar to the non-gelatinous copulatory plugreported by Andren and Nilson (1987) inVipera berus.

One study suggested that secretions fromthe posterior uterus interacted with spermand may facilitate sperm migration anteriorlyto infundibular sperm storage tubules (Hal-pert et al., 1982). However, in Agkistrodonpiscivorus secretions from the epitheliumlining the lumen of the posterior uterusproduced only mucous, no proteins (Siegeland Sever, 2008a). Sperm in this region wereassociated with glycoproteins similar in ap-pearance to those synthesized in the renalsexual segment of the male kidney in A.piscivorus and other snakes investigated(Sever et al., 2002; Sever et al., 2008).Considering the paucity of secretory materialproduced in the efferent ducts of male A.piscivorus, we believe the majority of seminalfluid associated with sperm in the femaleoviduct comes from the Rss of males.

Overlap of Male and FemaleReproductive Cycles

As stated above, the reproductive cycles ofmale and female North American crotalinesare clearer when thought of individuallybecause of the dissociation of reproductiveevents that can occur because of the ability ofmales and females to store sperm (Schuett,1992). However, in Fig. 17 (corroboration ofFig. 15 and 16) we present a graph thatdepicts the association of the male and femalereproductive cycles in Louisiana A. piscivorusby qualifying the activity of major organsinvolved in reproduction. Because of spermstorage, the only reproductive event that mustbe coordinated between males and females ismating activity, however, as Fig. 17 demon-strates, mating activity is also associated withsynthesis of abundant secretory material in themale excurrent ducts (Ed) and Rss. Theseresults add to the increasing data that the Rssis highly important in reproduction in snakes,and also implicate other potential areas (e.g.,distal ductuli epididymids) for synthesis ofseminal fluids in snakes.

The fall wave of spermatogenesis alsooverlaps the mating season in Agkistrodonpiscivorus from Louisiana. This is evidencedby sperm presence in the posterior oviduct offemales in August while spermatogenesis isstill occurring in males (Fig. 17). Additionally,the spring wave of spermatogenesis observedin Louisiana A. piscivorus is coordinated withmating activity. However, the ductus defer-entia of male A. piscivorus males were nevervoid of sperm before, during, or afterspermatogenesis. Therefore, an interestingquestion arises from these data: When aresperm from spermatogenic events actuallyutilized by males in a mating? It would seemthat since sperm are never absent in theductus deferentia, new sperm producedwould be aggregated at the most posteriorportion of the ductus deferentia and wouldnot be used in the first subsequent matingevent to spermatogenesis. However, there areno data available that determine how manymating events must occur to utilize spermfrom a specific spermatogenic event.

FUTURE DIRECTIONS

Obviously gaps exist in the present analysesof the reproductive cycles in Agkistrodon

FIG. 17.—Combined male and female reproductivecycles derived from Louisiana snakes (note: spermato-genic and renal sexual segment cycles would vary indifferent geographic locations). Ed activity, secretoryactivity of the excurrent ducts; Rss hypertrophy, secretoryactivity of the renal sexual segment; SST activity, secretoryactivity of sperm storage tubules; Utg, secretory activity ofuterine glands.


piscivorus. At this time, the only informationon the female oviducal cycle and seasonalvariation of spermatogenesis and secondarysexual characteristics in males is from a smallportion of the entire A. piscivorus range. Malehormonal data from southeastern Louisiana tocompare with the work by Johnson et al.(1982) and Graham et al. (2008) are desired.Any data on the hormonal control of repro-duction in females of this species are needed,and are lacking in North American crotalinesin general. While gathering these data,attention must be paid to the extremities ofthe A. piscivorus range (e.g., Texas, NorthCarolina, Kentucky, etc.), and Florida where adivergent mtDNA lineage resides (A. p.conanti; Guiher and Burbrink, 2008). Grossmorphology and histological analysis of muse-um specimens from different geographicregions could help supply some of the desireddata, and this seems feasible with the sub-stantial number of A. piscivorus present inmuseum collections. Other than an extensivemuseum survey, the only way to accomplishthese goals is to further collect wild specimensthroughout the reproductive and non-repro-ductive seasons.

The lack of hormonal data suggests morefield studies are necessary to gather bloodsamples from Agkistrodon piscivorus in theirnatural environment. More time spent to-wards field studies in different geographicareas may also add to the missing data of whenA. piscivorus are actually mating throughouttheir large geographic range. However, moretime spent in the field does not always resultin witnessing of reproductive behaviors, espe-cially in a rather shy snake in terms of sexualactivity like A. piscivorus (S. Graham, person-al communication). Radio-telemetry is withouta doubt an avenue that could help with thesemissing data and many projects utilizing thistechnique have recently been undertaken (E.Menzel, personal communication). Schuett andmany collaborators have been successful in thepast in utilizing captive colonies to collecthormonal data, and these types of studies aredesirable, especially in females where precisereproductive condition is important in evaluat-ing the hormonal control of reproduction.

We feel one of the more interestingphenomenon observed in Agkistrodon pisciv-

orus (and other squamates) is the ability tostore sperm for very long times. However,currently no data are available that track thelifespan of one cohort of sperm from onespermatogenic event, to its utilization inmating, and finally to its utilization infertilization. Specific chemicals like tritiatedthymidine that can label sperm from onecohort in a spermatogenic event may be veryuseful in determining the entire lifespan ofsperm, and the wealth of reproductive data onA. piscivorus may indicate its use in flagshipinvestigations that employ these types oflabeling techniques.

Whatever the direction of research onAgkistrodon piscivorus takes, we feel thatthe recent explosion of baseline reproductivedata on A. piscivorus makes this snake a primecandidate for future investigation. Utilizationof this species will not only benefit ourunderstanding of reproduction in viperids,but overall, also add to our knowledge of theevolution of reproductive biology in snakes.

Acknowledgments.—We wish to thank numerous indi-viduals who aided in the completion of this project: D.Bloom, R. E. Chabarria, and T. A. Schriever for help incollecting Cottonmouths; L. Alexander, A. Bagwill, A.Camus, M. Eckstut, and C. Morgan for help in processingmale kidney and duct tissues; C. and K. Fontenot foraccess to their property for specimen collection; E.Lemmons and E. Poldemann for help in processingtesticular tissues; B. Crother for discussion on thephylogenetics of snakes; M. Collier for help in statisticalanalysis; R. Aldridge for discussion on the reproductivebiology of snakes; Four anonymous reviewers who gaveextremely helpful comments on earlier versions of thismanuscript; Saint Louis University, Southeastern Louisi-ana University, and Wittenberg University for continualsupport during this project. D. M. S. acknowledgessupport from National Science Foundation grant DEB-0809831.

LITERATURE CITED

AKBARSHA, M. A., AND M. M. MEERAN. 1995. Occurrenceof ampulla in the ductus deferens of the Indian gardenlizard Calotes versicolor Daudin. Journal of Morphol-ogy 225:261–268.

AKBARSHA, M. A., V. TAMILARASAN, B. KADALMANI, AND P.DAISY. 2005. Ultrastructural evidence for secretion fromthe epithelium of ampulla ductus deferentis of the fan-throated lizard Sitana ponticeriana Cuvier. Journal ofMorphology 266:94–111.

AKBARSHA, M. A., B. KADALMANI, AND V. TAMILARASAN.2007. Efferent ductules of the fan-throated lizardSitana ponticeriana Cuvier: light and transmissionelectron microscopy study. Acta Zoologica 88:265–274.


AL-DOKHI, O. A. 2004. Electron microscopic study ofsperm head differentiation in the Arabian horned viperCerastes cerastes (Squamata, Reptilia). Journal ofBiological Sciences (Pakistan) 4:111–116.

AL-DOKHI, O. A. 2005. Ultrastructure of sperm taildifferentiation in the Arabian horned viper Cerastescerastes (Squamata, Reptilia). Saudi Journal of Biolog-ical Sciences 12:49–55.

AL-DOKHI, O. A., Y. AL-ONAZEE, AND M. MUBARAK. 2004.Light and electron microscopy of the testicular tissue ofthe snake Eryx jayakari (Squamata, Reptilia) with areference to the dividing germ cells. Journal ofBiological Sciences 4:345–351.

AL-DOKHI, O. A., Y. AL-ENEZY, AND M. MUBARAK. 2005a.Ultrastructure of spermiogenesis of the snake Eryxjayakari (Squamata, Reptilia): II. Sperm tail differen-tiation. Indian Journal of Applied and Pure Biology20:107–114.

AL-DOKHI, O. A., Y. AL-ENEZY, AND M. MUBARAK. 2005b.Ultrastructure study of the spermiogenesis of the snakeEryx jayakari (Squamata, Reptilia): II. Sperm headdifferentiation. Ultra Scientist of Physical Sciences17:1–10.

AL-DOKHI, O. A., Y. AL-ONAZEE, AND M. MUBARAK. 2007.Fine structure of the epididymal sperm of the snakeEryx jayakari (Squamata, Reptilia). International Jour-nal of Zoological Research 3:1–13.

ALDRIDGE, R. D. 1979. Seasonal spermatogenesis insympatric Crotalus viridis and Arizona elegans inNew Mexico. Journal of Herpetology 13:187–192.

ALDRIDGE, R. D. 1992. Oviductal anatomy and seasonalsperm storage in the Southeastern Crowned Snake(Tantilla coronata). Copeia 1992:1103–1106.

ALDRIDGE, R. D. 1993. Male reproductive anatomy andseasonal occurrence of mating and combat behavior ofthe rattlesnake Crotalus v. viridis. Journal of Herpe-tology 27:481–484.

ALDRIDGE, R. D. 2002. The link between mating seasonand male reproductive anatomy in the rattlesnakesCrotalus viridis oreganos and Crotalus viridis helleri.Journal of Herpetology 36:295–300.

ALDRIDGE, R. D., AND W. S. BROWN. 1995. Malereproductive cycle, age at maturity, and cost ofreproduction in the rattlesnake (Crotalus horridus).Journal of Herpetology 29:399–407.

ALDRIDGE, R. D., AND D. DUVALL. 2002. Evolution of themating season in the pitvipers of North America.Herpetological Monographs 2002:1–25.

ALDRIDGE, R. D., S. R. GOLDBERG, S. S. WISNIEWSKI, A. P.BUFALINO, AND C. B. DILLMAN. In press. The reproduc-tive cycle and estrus in the colubrid snakes of temperateNorth America. Contemporary Herpetology.

ALDRIDGE, R. D., B. C. JELLEN, M. C. ALLENDER, M. J.DRESLIK, D. B. SHEPARD, J. M. COX, AND C. A. PHILLIPS.2008. Reproductive biology of the Massasauga (Sistru-rus catenatus) from South-central Illinois. Pp. 403–412.In W. K. Hayes, K. R. Beaman, M. D. Cardwell, andS. P. Bush (Eds.), Biology of the Rattlesnake. LomaLinda University Press, Loma Linda, California, USA.

ALLEN, R. E., AND D. SWINDELL. 1948. Cottonmouthmoccasin of Florida. Herpetologica 4:2–16.

ALMEIDA-SANTOS, S. M., AND M. G. SALOMAO. 1997. Long-term sperm storage in the female neotropical rattle-

snake Crotalus durissus terrificus (Viperidae: Crotali-nae). Japanese Journal of Herpetology 17:46–52.

ALMEIDA-SANTOS, S. M., AND M. G. SALOMAO. 2002.Reproduction in neotropical pitvipers, with emphasison species of the genus Bothrops. Pp. 445–462. InG. W. Schuett, M. Hoggren, M. E. Douglas, and H. W.Greene (Eds.), Biology of the Vipers. Eagle MountainPublishing, Eagle Mountain, Utah, USA.

ANDREN, C., AND G. NILSON. 1987. The copulatory plug ofthe adder, Vipera berus: Does it keep sperm in or out?Oikos 49:230–232.

ASHTON, K. G., AND A. DE QUEIROZ. 2001. Molecularsystematics of the western rattlesnake, Crotalus viridis(Viperidae), with comments on the utility of the D-Loop in phylogenetic studies of snakes. MolecularPhylogenetics and Evolution 21:176–189.

AVERAL, H. I., M. MANIMEKALAI, AND M. A. AKBARSHA.1992. Differentiation along the ductus epididymis ofthe Indian garden lizard Calotes versicolor (Daudin).Biological Structures and Morphogenesis 4:53–57.

BEYER, G. E. 1893. Contributions on the life histories ofcertain snakes. American Naturalist XXXII, 17–24.

BISHOP, J. E. 1959. A histological and histochemical studyof the kidney tubules of the common garter snake,Thamnophis sirtalis, with special reference to the sexualsegment in the male. Journal of Morphology 104:307–357.

BLACKBURN, D. G. 1985. Evolutionary origins of viviparityin the Reptilia. II. Serpentes, Amphisbaenia, andEchthyosauria. Amphibia-Reptilia 5:259–291.

BLACKBURN, D. G. 1998. Structure, function, and evolu-tion of the oviducts of squamate reptiles, with specialreference to viviparity and placentation. Journal ofExperimental Zoology 282:560–617.

BLACKBURN, D. G. 1999. Viviparity and oviparity: evolutionand reproductive strategies. Pp. 994–1003. In T. E.Knobil and J. D. Neill (Eds.), Encyclopedia ofReproduction. Academic Press, New York, New York,USA.

BLACKBURN, D. G., AND A. F. FLEMMING. 2008. Morphol-ogy, development, and evolution of fetal membranesand placentation in squamate reptiles. Journal ofExperimental Zoology B (Molecular and Developmen-tal Evolution) 310B:1–11.

BONNET, X., G. NAULLEAU, AND R. MAUGET. 1994. Theinfluence of body condition on 17-b oestradiol levels inrelation to vitellogenesis in female Vipera aspis(Reptilia, Viperidae). General and Comparative Endo-crinology 93:424–437.

BONNET, X., G. NAULLEAU, D. BRADSHAW, AND R. SHINE.2001. Changes in plasma progesterone in relation tovitellogenesis and gestation in the viviparous snakeVipera aspis. General and Comparative Endocrinology121:84–94.

BOU-RESLI, M. N., L. F. BISHAY, AND N. S. AL-ZAID. 1981.Observations on the fine structure of the sperm storagecrypts in the lizard Acanthodactylus scutellatus hardyi.Archives de Biologie (Bruxelles) 92:287–298.

BURKETT, R. D. 1966. Natural history of the cottonmouthmoccasin, Agkistrodon piscivorus (Reptilia). Universityof Kansas Publications, Museum of Natural History1966:435–491.

CASTOE, T. A., AND C. L. PARKINSON. 2006. Bayesian mixedmodels and the phylogeny of pitvipers (Viperidae:


Serpentes). Molecular Phylogenetics and Evolution39:91–110.

CONANT, R. 1933. Three generations of cottonmouths,Agkistrodon piscivorus (Lacepede). HerpetologicalReview 1:43.

CONANT, R., AND J. T. COLLINS. 1998. A Field Guide toReptiles and Amphibians Eastern/Central North Amer-ica. Houghton and Mifflin Company, New York, NewYork, USA.

COPE, E. D. 1898. The crocodilians, lizards and snakes inNorth America. Report of the U.S. National Museum1890:155–1270.

CUELLAR, O. 1966. Oviducal anatomy and sperm storagestructures in lizards. Journal of Morphology 119:7–20.

CUNHA, L. D., L. TAVARES-BASTOS, AND S. N. BAO. 2008.Ultrastructural description and cytochemical study ofthe spermatozoon of Crotalus durissus (Squamata,Serpentes). Micron 39:915–925.

DEHLAWI, G. Y., AND F. M. ISMAIL. 1994. Studies of theultrastructure on the spermiogenesis of Saudian rep-tiles. 13-The sperm differentiation in the snake.Proceedings of the Zoological Society of the ArabRepublic of Egypt 25:95–106.

DEVINE, M. C. 1975. Copulatory plugs: Enforced chastity.Science 187:844–845.

DOUGLAS, M. E., M. R. DOUGLAS, G. W. SCHUETT, AND

L. W. PORRAS. 2009. Climate change and evolution ofthe new world pitviper genus Agkistrodon (Viperidae).Journal of Biogeography 36:1164–1180.

DUFARE, J. P., AND H. SAINT-GIRONS. 1984. Histologiecomparee de l’epididyme et de ses secretions chez lesreptiles (lizards et serpents). Archives d’AnatomieMicroscopique 73:15–26.

DUNDEE, H. A., AND D. A. ROSSMAN. 1989. TheAmphibians and Reptiles of Louisiana. Louisiana StateUniversity Press, Baton Rouge, Louisiana, USA.

ECKSTUT, M. E., D. M. SEVER, M. E. WHITE, AND B. I.CROTHER. 2009. Phylogenetic analysis of sperm storagein female squamates. Pp. 1–34. In L. T. Dahnof (Ed.),Animal Reproduction: New Research Developments.Nova Science Publishers, New York, New York, USA.

FORD, N. B. 2002. Ecology of the western cottonmouth(Agkistrodon piscivorus leucostoma) in NortheasternTexas. Pp. 167–177. In G. W. Schuett, M. Hoggren,M. E. Douglas, and H. W. Greene (Eds.), Biology ofthe Vipers. Eagle Mountain Publishing, Eagle Moun-tain, Utah, USA.

FORD, N. B., F. BRISCHOUX, AND D. LANCASTER. 2004.Reproduction in the western cottonmouth, Agkistrodonpiscivorus leucostoma, in a floodplain forest. SouthwestNaturalist 49:465–471.

FOX, W. 1952. Seasonal variation in the male reproductivesystem of Pacific coast garter snakes. Journal ofMorphology 90:481–553.

FOX, W. 1956. Seminal receptacles of snakes. AnatomicalRecord 124:519–539.

FOX, H. 1977. The urogenital system of reptiles.Pp. 1–157. In C. Gans and T. S. Parsons (Eds.),Biology of the Reptilia, Vol. 6. Morphology E. Aca-demic Press, New York, New York, USA.

FOX, W., AND H. C. DESSAUER. 1962. The single rightoviduct and other urogenital structures of femaleTyphlops and Leptotyphlops. Copeia 1962:590–597.

FUNK, R. S. 1964. Birth of a brood of westerncottonmouths, Agkistrodon piscivorus leucostoma.Transactions of the Kansas Academy of Science 67:199.

GABE, M., AND H. SAINT-GIRONS. 1965. Contribution a lamorphologie comparee du cloaque et des glandesepidermoıdes de la region cloacale chez les lepidosaur-iens. Memoires du Museum National d’HistoireNaturelle XXXII, 149–292.

GABRI, M. S. 1983. Seasonal changes in the ultrastructureof the kidney collecting tubules in the lizard Podacris(5Lacerta) taurica. Journal of Morphology 175:143–151.

GAMPERT, O. 1866. Ueber die niere von Tropidonotusnatrix und der cyprinoiden. Zeitschrift fur Wissench-aftliche Zoologie 16:369–373.

GIRLING, J. E. 2002. The reptilian oviduct: A review ofstructure and function and directions for futureresearch. Journal of Experimental Zoology 293:141–170.

GIRLING, J. E., A. CREE, AND L. J. GUILLETTE, JR. 1998.Oviductal structure in four species of gekkonid lizarddiffering in parity mode and eggshell structure.Reproduction, Fertility and Development 10:139–154.

GLOYD, H. K., AND R. CONANT. 1990. Snakes of theAgkistrodon Complex: A Monographic Review. Societyfor the Study of Amphibians and Reptiles (SSAR),Contributions to Herpetology 6. SSAR, Oxford, Ohio,USA.

GOLDBERG, S. R. 1995. Reproduction in the Lyre Snake,Trimorphodon biscutatus (Colubridae), from Arizona.Southwest Naturalist 40:334–335.

GOLDBERG, S. R. 1998. Reproduction in the sonoranwhipsnake, Masticophis bilineatus (Serpentes: Colubri-dae). Southwest Naturalist 43:412–414.

GOLDBERG, S. R. 1999a. Reproduction in the blacktailrattlesnake, Crotalus molossus (Serpentes: Viperidae).Texas Journal of Science 51:323–328.

GOLDBERG, S. R. 1999b. Reproduction in the red diamondrattlesnake in California. California Fish and Game85:177–180.

GOLDBERG, S. R. 2000a. Reproduction in the twin-spottedRattlesnake, Crotalus pricei (Serpentes: Viperidae).Western North America Naturalist 60:98–100.

GOLDBERG, S. R. 2000b. Reproduction in the rockrattlesnake, Crotalus lepidus (Serpentes: Viperidae).Herpetological Natural History 7:83–86.

GOLDBERG, S. R. 2000c. Reproduction in the tigerrattlesnake, Crotalus tigris (Serpentes: Viperidae).Texas Journal of Science 51:31–36.

GOLDBERG, S. R. 2007. Testicular cycle of the westerndiamondback rattlesnake, Crotalus atrox (Serpentes:Viperidae), from Arizona. Bulletin of the MarylandHerpetological Society 43:103–107.

GOLDBERG, S. R., AND T. HOLYCROSS. 1999. Reproductionin the desert massasauga, Sistrurus catenatus edwardsi,in Arizona and Colorado. Southwestern Naturalist44:531–535.

GOLDBERG, S. R., AND P. C. ROSEN. 2000. Reproduction inthe mojave rattlesnake, Crotalus scutulatus (Serpentes:Viperidae). Texas Journal of Science 52:101–109.

GRAHAM, S. P., R. L. EARLEY, S. K. HOSS, G. W. SCHUETT,AND M. S. GROBER. 2008. The reproductive biology ofmale cottonmouths (Agkistrodon piscivorus): Do plas-


ma steroid hormones predict the mating season?General and Comparative Endocrinology 159:226–235.

GRIBBINS, K. M., AND D. H. GIST. 2003. Cytologicalevaluation of spermatogenesis within the germinalepithelium of the male European wall lizard, Podarcismuralis. Journal of Morphology 258:296–306.

GRIBBINS, K., D. H. GIST, AND J. CONGDON. 2003.Cytological evaluation of spermatogenesis in the red-eared slider, Trachemys scripta. Journal of Morphology255:337–346.

GRIBBINS, K. M., C. S. HAPP, AND D. M. SEVER. 2005.Ultrastructure of the reproductive system of the blackswamp snake (Seminatrix pygaea). V. The temporalgerm cell development strategy of the testis. ActaZoologica 86:223–230.

GRIBBINS, K. M., R. M. ELSEY, AND D. H. GIST. 2006.Cytological evaluation of the germ cell developmentstrategy within the testis of the American alligator,Alligator mississippiensis. Acta Zoologica 87:59–69.

GRIBBINS, K. M., E. M. MILLS, AND D. M. SEVER. 2007.Ultrastructural examination of spermiogenesis withinthe testis of the ground skink, Scincella laterale(Squamata, Sauria, Scincidae). Journal of Morphology268:181–192.

GRIBBINS, K. M., J. L. RHEUBERT, M. ANZALONE, D. S.SIEGEL, AND D. M. SEVER. In press. Ultrastructure ofspermiogenesis in the cottonmouth, Agkistrodon pisciv-orus (Squamata: Viperidae: Crotalinae). Journal ofMorphology.

GRIBBINS, K. M., J. L. RHEUBERT, M. H. COLLIER, D. S.SIEGEL, AND D. M. SEVER. 2008. Histological analysis ofspermatogenesis and the germ cell developmentstrategy within the testis of the male western cotton-mouth snake, Agkistrodon piscivorus leucostoma. An-nals of Anatomy 190:461–476.

GUIHER, T. J., AND F. T. BURBRINK. 2008. Demographicand phylogeographic histories of two venomous NorthAmerican snakes of the genus Agkistrodon. MolecularPhylogenetics and Evolution 48:543–553.

HALPERT, A. P., W. R. GARSTKA, AND D. CREWS. 1982.Sperm transport and storage and its relation to theannual cycle of the female red-sided garter snake,Thamnophis sirtalis parietalis. Journal of Morphology174:149–159.

HOFFMAN, L. H. 1970. Placentation in the garter snake,Thamnophis sirtalis. Journal of Morphology 131:57–88.

HOFFMAN, L. H., AND W. A. WIMSATT. 1972. Histochemicaland electron microscopic observations on the spermreceptacles in the garter snake oviduct. AmericanJournal of Anatomy 134:71–96.

HONDO, E., M. KUROHMARU, M. TORIBA, AND Y. HAYASHI.1994. Seasonal changes in spermatogenesis and ultra-structure of developing spermatids in the Japanese ratsnake, Elaphe climacophora. Journal of VeterinaryMedical Science 56:835–840.

ISMAIL, F. M., G. Y. DEHLAWI, AND O. S. SARHAN. 1995.Ultrastructure of the spermiogenesis of Saudian rep-tiles. 10-The sperm differentiation in Echis carinatus(Serpentes, Viperidae). Proceedings of the ZoologicalSociety of the Arab Republic of Egypt 26:56–65.

JAMIESON, B. G. M. 1999. Spermatozoal phylogeny of theVertebrata. Pp. 303–331. In C. Gagnon (Ed.), TheMale Gamete: From Basic Science to Clinical Applica-tions. Cache River Press, Vienna, Illinois, USA.

JOHNSON, L. F., J. S. JACOB, AND P. TORRANCE. 1982.Annual testicular and androgenic cycles of the cotton-mouth (Agkistrodon piscivorus) in Alabama. Herpeto-logica 38:16–25.

KLAUBER, L. M. 1972. Rattlesnakes: Their Habits, LifeHistories, and Influence on Mankind. University ofCalifornia Press, Berkeley, California, USA.

LAMB, J. C. IV, R. R. NEWBOLD, W. E. STUMPF, AND J. A.MCLACHLAN. 1978. Transitional changes in the surfaceepithelium of the cycling mouse vagina, cervix anduterus: scanning electron microscopic studies. Biologyof Reproduction 19:701–711.

LUDWIG, M., AND H. RAHN. 1943. Sperm storage andcopulatory adjustment in the prairie rattlesnake. Copeia1943:15–18.

NILSON, G., AND C. ANDREN. 1982. Function of the renalsex secretion and male hierarchy in the adder, Viperaberus, during reproduction. Hormones and Behavior16:404–413.

OLIVER, S. C., B. G. M. JAMIESON, AND D. M. SCHELTINGA.1996. The ultrastructure of spermatozoa of SquamataII. Agamidae, Varanidae, Colubridae, Elapidae, andBoidae (Reptilia). Herpetologica 52:216–241.

PALMER, B. D., V. G. DEMARCO, AND J. R. GUILLETTE. 1993.Oviductal morphology and eggshell formation in thelizard, Sceloporus woodi. Journal of Morphology217:205–217.

PERKINS, J. M., AND B. D. PALMER. 1996. Histology andfunctional morphology of the oviduct of an oviparoussnake, Diadophis punctatus. Journal of Morphology277:67–79.

POOK, C. E., W. WUSTER, AND R. S. THORPE. 2000.Historical biogeography of the western rattlesnake(Serpentes: Viperidae: Crotalus viridus), inferred frommitochondrial DNA sequence information. MolecularPhylogenetics and Evolution 15:269–282.

POPE, C. H. 1941. Copulatory adjustment in snakes.Zoological Series of Field Museum of Natural History24:249–252.

PRASAD, M. R. N., AND P. R. K. REDDY. 1972. Physiology ofthe sexual segment of the kidney in reptiles. Generaland Comparative Endocrinology Supplements 3:649–662.

PRASAD, M. R. N., AND M. K. SANYAL. 1969. Effect of sexhormones on the sexual segment of kidney and otheraccessory reproductive organs of the Indian HouseLizard Hemidactylus flaviviridis Ruppell. General andComparative Endocrinology 12:110–118.

RAHN, H. 1942. The reproductive cycle of the prairierattlesnake. Copeia 1942:233–240.

RAYNAUD, A., AND C. PIEAU. 1985. Embryonic develop-ment of the genital system. Pp. 149–300. In C. Gansand F. Billett (Eds.), Biology of the Reptilia: Volume 15Development B. John Wiley & Sons, Inc., New York,New York, USA.

ROSS, P., AND D. CREWS. 1977. Influence of the seminalplug on mating behavior in the garter snake. Nature(London) 267:344–345.

RUSSELL, L. D., S. A. P. HIKIM, R. A. ETTLIN, AND E. D.LEGG. 1990. Histological and histopathological evaluationof the testis. Cache River Press, Vienna, Illinois, USA.

SAINT-GIRONS, H. 1957. Le cycle sexuel chez Vipera aspis(L.) dans l’ouest de la france. Bulletin Biologique de laFrance et la Belgique 91:284–350.


SAINT-GIRONS, H. 1959. Donnees histochimiques sur lesglucides de l’appareil genital chez les vipers, au coursdu cycle reproducteur. Annales de Histochimie4:235–243.

SAINT-GIRONS, H. 1962a. Le cycle reproducteur de lavipere a cornes, Cerastes cerastes (L.), dans la nature eten captivite. Bulletin de la Societe Zoologique deFrance 87:41–51.

SAINT-GIRONS, H. 1962b. Presence de receptacles semi-naux chez les cameleons. Beaufortia 9:165–172.

SAINT-GIRONS, H. 1973. Sperm survival and transport inthe female genital tract of reptiles. Pp. 105–113. InE. S. E. Hafez and C. G. Thibault (Eds.), The Biologyof Spermatozoa. Karger, Basel, Switzerland.

Saint-Girons, H. 1982. Reproductive cycles of male snakesand their relationships with climate and femalereproductive cycles. Herpetologica 38:5–16.

SANCHEZ-MARTINEZ, P. M., M. P. RAMIREZ-PINILLA, AND

D. R. MIRANDA-ESQUIVEL. 2007. Comparative histologyof the vaginal-cloacal region in Squamata and itsphylogenetic implications. Acta Zoologica 88:289–307.

SCHUETT, G. W. 1992. Is long-term sperm storage animportant component of the reproductive biology oftemperate pitvipers. Pp. 169–184. In J. A. Campbelland E. D. Brodie, Jr (Eds.), Biology of the Pitvipers.Selva, Tyler, Texas, USA.

SCHUETT, G. W., S. L. CARLISLE, A. T. HOLYCROSS, J. K.O’LEILE, D. L. HARDY, SR., E. A. VAN KIRK, AND W. J.MURDOCH. 2002. Mating system of male mojaverattlesnakes (Crotalus scutulatus): Seasonal timing ofmating, agonistic behavior, spermatogenesis, sexualsegment of the kidney, and plasma sex steroids.Pp. 515–532. In G. W. Schuett, M. Hoggren, M. E.Douglas, and H. W. Greene (Eds.), Biology of theVipers. Eagle Mountain Publishing, Eagle Mountain,Utah, USA.

SCHUETT, G. W., M. S. GROBER, E. A. VAN KIRK, AND W. J.MURDOCH. 2004. Long-term sperm storage and plasmasteroid profile of pregnancy in a western diamond-backed rattlesnake (Crotalus atrox). HerpetologicalReview 35:328–333.

SCHUETT, G. W., D. L. HARDY, H. W. GREENE, R. L.EARLEY, M. S. GROBER, E. A. VAN KIRK, AND W. J.MURDOCH. 2005. Sympatric rattlesnakes with contrastingmating systems show differences in seasonal patterns ofplasma sex steroids. Animal Behavior 70:257–266.

SCHUETT, G. W., H. J. HARLOW, J. D. ROSE, E. A. VAN

KIRK, AND W. J. MURDOCH. 1997. Annual cycle of plasmatestosterone in male copperheads, Agkistrodon contor-trix (Serpentes, Viperidae): relationship to timing ofspermatogenesis, mating, and agonistic behavior. Gen-eral and Comparative Endocrinology 105:417–424.

SEVER, D. M. 2004. Ultrastructure of the reproductivesystem of the black swamp snake (Seminatrix pygaea).IV. Occurrence of an ampulla ductus deferentis.Journal of Morphology 262:714–730.

SEVER, D. M., AND W. A. HOPKINS. 2004. Oviductal spermstorage in the ground skink Scincella laterale Holbrook(Reptilia: Scincidae). Journal of Experimental Zoology301A:599–611.

SEVER, D. M., AND W. A. HOPKINS. 2005. Renal sexualsegment of the ground skink, Scincella laterale (Reptil-ia, Squamata, Scincidae). Journal of Morphology266:46–59.

SEVER, D. M., T. J. RYAN, T. MORRIS, D. PATTON, AND S.SWAFFORD. 2000. Ultrastructure of the reproductivesystem of the black swamp snake (Seminatrix pygaea).II. Annual oviducal cycle. Journal of Morphology245:146–160.

SEVER, D. M., AND T. J. RYAN. 1999. Ultrastructure of thereproductive system of the Black Swamp Snake(Seminatrix pygaea). I. Evidence for oviducal spermstorage. Journal of Morphology 241:1–18.

SEVER, D. M., T. J. RYAN, R. STEPHENS, AND W. C.HAMLETT. 2002. Ultrastructure of the reproductivesystem of the black swamp snake (Seminatrix pygaea).III. Rss of the male kidney. Journal of Morphology252:238–254.

SEVER, D. M., D. S. SIEGEL, A. BAGWILL, M. E. ECKSTUT,L. ALEXANDER, A. CAMUS, AND C. MORGAN. 2008. Renalsexual segment of the cottonmouth snake, Agkistrodonpiscivorous (Reptilia, Squamata, Viperidae). Journal ofMorphology 269:640–653.

SHINE, R. 1985. The evolution of viviparity in reptiles: Anecological analysis. Pp. 605–694. In C. Gans and F.Billet (Eds.), Biology of the Reptilia, Vol. 15. Wiley,New York, New York, USA.

SIEGEL, D. S., AND D. M. SEVER. 2006. Utero-musculartwisting and sperm storage in viperids. HerpetologicalConservation and Biology 1:87–92.

SIEGEL, D. S., AND D. M. SEVER. 2008a. Spermaggregations in female Agkistrodon piscivorus (Reptil-ia:Squamata): a histological and ultrastructural investi-gation. Journal of Morphology 269:189–206.

SIEGEL, D. S., AND D. M. SEVER. 2008b. Seasonal variationin the oviduct of female Agkistrodon piscivorus(Reptilia:Squamata): an ultrastructural investigation.Journal of Morphology 269:980–997.

TAYLOR, E. N., D. F. DENARDO, AND D. H. JENNINGS. 2004.Seasonal steroid hormone levels and their relationto reproduction in the western diamond-backed rattle-snake, Crotalus atrox (Serpentes: Viperidae). Generaland Comparative Endocrinology 136:328–337.

THOMPSON, M. B., J. B. BIAZIK, S. LUI, S. M. ADAMS, AND

C. R. MURPHY. 2006. Morphological and functionalchanges to the uterus of lizards with different placentalcomplexities. Herpetological Monographs 2006:178–185.

TOURMENTE, M., G. CARDOZO, M. BERTONS, A. GUIDO-

BALDI, L. GIOJALAS, AND M. CHIARAVIGLIO. 2006. Theultrastructure of the spermatozoa of Boa constrictoroccidentalis, with considerations on its mating systemand sperm competition theories. Acta Zoologica87:25–32.

TOURMENTE, M., L. GIOJALAS, AND M. CHIARAVIGLIO. 2008.Sperm ultrastructure of Bothrops alternatus andBothrops diporus (Viperidae, Serpentes), and itspossible relation to the reproductive features of thespecies. Zoomorphology 4:241–248.

VOLSØE, H. 1944. Structure and Seasonal Variation of theMale Reproductive Organs of Vipera berus (L.). SpoliaZoologica Musei Hauniensis V. Skrifter, UniversitetetsZoologiske Museum, København, Denmark.

WHARTON, C. H. 1960. Birth and behavior of a brood ofcottonmouths, Agkistrodon piscivorus with notes on tailluring. Herpetologica 16:125–129.

WHARTON, C. H. 1966. Reproduction and growth inthe cottonmouths, Agkistrodon piscivorus Lacepede, ofCedar Keys, Florida. Copeia 1966:149–161.


Documents

REPRODUCTIVE BIOLOGY OF AGKISTRODON PISCIVORUS … · reproductive biology of agkistrodon piscivorus lace´pe`de (squamata, serpentes, viperidae, crotalinae) dustin s. siegel 1,4,david