Ultrastructure of the Annual Cycle of Female SpermStorage in Spermathecae of the Torrent Salamander,Rhyacotriton variegatus (Amphibia: Rhyacotritonidae)David M. Sever,1* Cynthia K. Tait,2 Lowell V. Diller,3 and Laura Burkholder3

1Department of Biology, Saint Mary’s College, Notre Dame, Indiana 465562Bureau of Land Management, Vale, Oregon 979183Simpson Timber Company, Arcata, California 95521

ABSTRACT This study is the first report on the ultra-structure of the sperm storage glands (spermathecae) inthe salamander Rhyacotriton variegatus. The populationstudied is associated with cold-water, rocky streams of theredwood (Sequoia) zone in northern California. Males pos-sess sperm in their vasa deferentia and undergo spermi-ation throughout the year, but mating is seasonal. Mostfemales with large, vitellogenic follicles (2.0–3.9 mmmean dia.) collected from February–June contain sperm intheir spermathecae, although some females with largefollicles lack sperm. Other mature-size females collectedduring this period have small ovarian follicles (0.9–1.2mm mean dia.) and lack stored sperm. All females col-lected from September–November have small follicles(0.6–1.6 mm mean dia.) and lack sperm, except in oneinstance in which a female collected in November had asmall amount of degraded sperm, apparently retainedfrom the previous breeding season. The spermathecae con-sist of simple tubulo-alveolar glands in which the necktubules produce a mucoid secretory product, and the distalbulbs, where sperm are stored, contain secretory vacuolesof uniform density that stain positively for glycosamino-glycans. In specimens containing sperm, some bulbs haveabundant sperm and others lack sperm, but the ultra-structure is similar in both conditions. The acini containcolumnar epithelial cells with wide intercellular canalic-uli, and a merocrine process releases the secretion. Sper-miophagy occurs. In specimens from spring and summerwith small ovarian follicles, the neck tubules are similarto those of breeding females, but the distal bulbs arereduced to cords of cells lacking a discernible lumen. Se-cretory activity in the distal bulbs is initiated in the fall.Spermathecae of R. variegatus are most similar to those ofa stream-dwelling plethodontid, Eurycea cirrigera. J. Mor-phol. 261:1–17, 2004. © 2004 Wiley-Liss, Inc.

KEY WORDS: Amphibia; Urodela; Rhyacotritonidae;Rhyacotriton; reproduction; sperm storage; spermathecae;ultrastructure

Sperm storage occurs in all females of the sevenfamilies of salamanders comprising the suborderSalamandroidea (Sever, 1991a, 1994). Instead ofoviductal sperm storage, as known in other femalevertebrates that store sperm, sperm storage in fe-male salamanders occurs in cloacal glands called

spermathecae. The ancestral condition forsalamanders is lack of sperm storage glands, a con-dition found in three families with external fertili-zation, Sirenidae, Hynobiidae, and Cryptobranchi-dae (Sever, 1991b).

The anatomy of sperm storage glands has beenextensively studied in salamanders and was mostrecently reviewed by Sever (2003). Much work hasbeen done on spermathecae at the light microscopylevel, but ultrastructural studies using transmissionelectron microscopy are the best means for studyingcytology of sperm/epithelial interactions and the se-cretory cycle of spermathecal epithelium.

In this article, we present the first ultrastructuralobservations on sperm storage in females of a spe-cies of the Rhyacotritonidae, using Rhyacotriton var-iegatus (Fig. 1A). Rhyacotriton variegatus is one offour species in the family Rhyacotritonidae, which isrestricted to the Pacific Northwest region of theUnited States (Good and Wake, 1992). Sister-grouprelationships of Rhyacotritonidae to othersalamander families are obscure (Larson et al.,2003).

Members of this family occupy springs, seeps, andthe edges of swift-flowing, permanent streams withcoarse substrates (Petranka, 1998). In Oregon, ovi-position can occur at almost any time, althoughthere may be a peak of courtship and egg-layingactivity in spring or early summer (Nussbaum andTait, 1977). The few nests reported for any of thespecies were found submerged in rock crevices orunder gravel in first order streams and springs. Inthese nests, eggs were apparently laid in summer orfall (Nussbaum, 1969; Karraker, 1999; Russell et al.,2002). A description of the cloaca and spermathecae

*Correspondence to: David M. Sever, Department of Biology, SaintMary’s College, Notre Dame, IN 46556.E-mail: dsever@saintmarys.edu

Published online inWiley InterScience (www.interscience.wiley.com)DOI: 10.1002/jmor.10149

JOURNAL OF MORPHOLOGY 261:1–17 (2004)

© 2004 WILEY-LISS, INC.

of Rhyacotriton variegatus at the light microscopylevel was provided by Sever (1992b).

The purpose of this study is three-fold. First, weprovide descriptions of the seasonal variation in ul-trastructure of the spermathecae of Rhyacotritonvariegatus. Second, we compare our findings on R.variegatus to those on ultrastructural characteris-tics of sperm storage in other salamanders. Finally,we provide preliminary observations on other as-pects of the male and female reproductive cycles ofR. variegatus in northern California. This study ispart of a more extensive project coordinated by LVDon the reproductive biology of R. variegatus.

MATERIALS AND METHODS

The specimens of Rhyacotriton variegatus used in this studycame from the North Fork Mad River watershed on lands ownedby the Simpson Timber Company in western Humboldt County,California, in the north coast redwood (Sequoia sempervirens)zone. Rhyacotriton variegatus is considered a “species of concern”by the state of California, and permits for the collection of alimited sample were granted to Lowell V. Diller from the Califor-nia Department of Fish and Game. Collection dates and otherdata on the 25 female specimens used in this study are shown inTable 1. Animals generally were sacrificed and preserved within24 h of collection. Collections were made in February 2002, April2000, 2001, June 1999, 2001, September 2001, and November1999, 2001. The spermathecae of these specimens were examinedby light (LM) and transmission electron microscopy (TEM). A

sample from March 2001 was used in a separate study involvingscanning electron microscopy (SEM) of sperm packing in thespermathecae (Sever et al., in prep.).

Specimens were killed by immersion in 10% aqueous solution ofMS-222 (3-aminobenzoic acid ethyl ester). The Animal Care andUse Committee of Saint Mary’s College approved this procedure.After death, snout–vent length (SVL) was measured from the tipof the snout to the posterior end of the cloacal orifice. Vitellogenicfollicles were counted and their diameters measured to the near-est 0.1 mm. Based upon the size of the follicles and appearance ofthe oviducts, females were separated into three categories: 1)Gravid—females with large yolked follicles (�2.0 mm mean dia.)and thick, convoluted oviducts (Fig. 1); 2) Spent—females withsmaller follicles but flaccid, thick, convoluted oviducts indicativeof recent stretching; and 3) First yolkers—females with smallfollicles and thin, straight or wavy oviducts. First yolkers arefemales reaching sexual maturity, beginning to yolk eggs for thefirst time, but are not yet ready to breed.

The cloacal area and an associated oviduct were removed fromeach specimen. The spermathecal area is pigmented (Fig. 1B),and this region was bisected. One-half was prepared for TEM andthe other half for LM. For TEM, a 1-mm block of pigmented tissuewas removed and fixed in a 1:1 solution of 2.5% glutaraldehydeand 3.7% formaldehyde in cacodylate buffer at pH 7.2. Afterinitial fixation, tissues were rinsed in distilled-deionized water,postfixed in 2% osmium tetroxide, dehydrated through a gradedseries of ethanol, cleared in propylene oxide, and polymerized inan epoxy resin (Embed 812, Electron Microscopy Sciences, PortWashington, PA). Plastic sections were cut with an RMC MT7ultramicrotome (Research and Manufacturing Co., Tucson, AZ)and DiATOME (Biel, Switzerland) diamond knives. Semithinsections (0.5–1 �m) for LM were placed on microscope slides andstained with toluidine blue. Ultrathin sections (70 nm) for TEM

Fig. 1. Rhyacotriton variega-tus. A: Dorsal view of a 60.8 mmSVL female collected 10 Aprilwith ovarian follicles 3.5 mmmean dia. B: Ventral view, dis-sected to show reproductivetract. Od, oviduct; Ov, ovary; St,spermathecae.

2 D.M. SEVER ET AL.

were collected on uncoated copper grids and stained with solu-tions of uranyl acetate and lead citrate. Ultrathin sections wereviewed with a Hitachi H-300 TEM (Nissei Sangyo America,Mountain View, CA).

Spermathecal tissue prepared for LM was processed with tes-tes and vasa deferentia removed from males. Previous studies onthe male reproductive cycle of Rhyacotriton variegatus and otherspecies in the genus found that males possess sperm in their vasadeferentia throughout the year (Nussbaum and Tait, 1977;Zalisko and Larsen, 1988). We removed testes and vasa deferen-tia from males collected with females throughout the year, in-cluding two males (52.6, 53.0 mm SVL) collected 1 April, threemales (51.1, 53.9, and 56.4 mm SVL) collected 10 April, threemales collected 18 June (50.8, 51.9, and 53.6 mm SVL), and fourmales (44.9, 45.7, 47.7, and 48.9 mm SVL) collected 1 November.

Tissues prepared for LM were fixed in 10% neutral bufferedformalin (NBF), rinsed in water, dehydrated in ethanol, clearedin toluene, and embedded in paraffin. Paraffin sections (10 �m)were cut with a rotary microtome and affixed to albuminizedslides. Slides were stained with hematoxylin-eosin. In addition,alternate slides from a 54.2 mm SVL female collected 10 Febru-ary and a 55 mm SVL female collected 1 April were stained withAlcian blue 8GX at pH 2.5 (AB, primarily for carboxylated gly-cosaminoglycans) followed by the periodic acid-Schiff’s procedure(PAS, for neutral carbohydrates and sialic acids), or stained withbromophenol blue (BB, for proteins). Procedures followed Dawes(1979), Humason (1979), and Kiernan (1990).

RESULTSMale Reproductive Cycle

Males from all samples possess testes in whichspermiation is actively occurring (Fig. 2A,C,E) andhave vasa deferentia containing sperm (Fig.2B,D,F). Spermiogenic activity in the testes seemed

equally intense in all samples, but did not occuruniformly throughout the testes. Each specimencontained testicular lobules in which little or nospermatogenesis was occurring. The testes are sim-ple, not multiply lobed, as known in somesalamander families. A caudo-cephalic wave of sper-matogenic activity characteristic of simple testes inmany salamanders (e.g., Uribe Aranzabal, 2003) isnot apparent in those of Rhyacotriton variegatus.Sperm seem most numerous in the vasa deferentiaof males from the February–April samples (Fig. 2A).In other months, sperm are often associated withcolloidal material in the lumina of the vasa deferen-tia (Fig. 2D,F). However, the presence of sperm inthe vas deferens in all samples indicates that malespossibly could engage in mating activity throughoutthe year.

Female Reproductive Cycle

The largest ovarian follicles (2.0–3.9 mm meandia.) occur in females from the February–June sam-ples, although some females from these samples pos-sess much smaller follicles (0.9–1.2 mm mean dia.;Table 1). The females with small follicles could beimmature, although the smallest of these (45.1 mmSVL) is within the range of mature body size re-ported by Nussbaum and Tait (1977). Females fromthe September–November samples all have smallfollicles (0.6–1.6 mm mean dia.). The females with

TABLE 1. Female Rhyacotriton variegatus used in this study1

Date SVL

Follicles

Cond2 SpermN Range Mean SE

10 Feb 52.2 8 1.5–2.5 2.0 0.11 GD 010 Feb 54.2 10 1.6–3.0 2.4 0.11 GD �10 Feb 54.2 6 2.1–3.1 2.8 0.16 GD �10 Feb 54.4 7 1.8–2.6 2.1 0.11 GD �10 Feb 56.8 11 1.9–2.7 2.3 0.08 GD �1 April 45.3 10 0.8–1.1 0.9 0.03 FY 01 April 55.0 9 3.0–3.5 3.2 0.06 GD �10 April 48.6 8 0.8–1.6 1.1 0.05 FY 010 April 60.8 12 3.2–3.9 3.5 0.07 GD �18 June 45.1 7 0.6–1.2 1.0 0.06 FY 018 June 48.1 7 1.0–1.5 1.2 0.07 FY 018 June 50.3 6 1.4–3.3 2.8 0.16 GD �18 June 53.0 9 3.0–4.4 3.5 0.16 GD �23 Jun 45.8 eggs not yolked; count unreliable FY 023 Jun 52.1 7 3.8–4.0 3.9 0.03 GD 023 Jun 57.4 8 2.7–3.7 3.1 0.15 GD �3

30 Jun 50.6 6 2.1–3.2 2.7 0.17 GD 07 Sep 52.6 11 1.0–1.4 1.2 0.04 ? 01 Nov 43.4 5 0.5–0.7 0.6 0.03 FY 01 Nov 44.3 5 0.7–1.0 0.9 0.04 FY 01 Nov 47.9 16 0.6–1.0 0.8 0.04 FY 01 Nov 52.2 9 1.2–1.9 1.6 0.08 ST 05 Nov 52.6 14 1.1–1.8 1.4 0.04 ST �5 Nov 53.7 10 1.0–1.5 1.3 0.06 ST 05 Nov 55.0 12 1.1–1.7 1.4 0.07 ST 0

1Measurements for SVL and the range and mean of follicle dia. are in mm.2Condition � Gr, gravid; FY, first yolking; ST, spent.3Three eggs were inside the oviduct and too distorted to measure, so n � 5 for calculation of mean and SE.

3TORRENT SALAMANDER SPERM STORAGE

Fig. 2. Rhyacotriton variegatus. Sagittal paraffin sections stained with hematoxylin-eosin. Scale bar in lower right corner is thesame for all micrographs. Testis (A) and vas deferens (B) of a male collected 10 April. Testis (C) and vas deferens (D) of a male collected18 June. Testis (E) and vas deferens (F) of a male collected 1 November. Co, colloidal material; Sg, spermatogonia; Sp, sperm; Ss,secondary spermatocytes; St, spermatids.

larger follicles (1.2–1.6 mm mean dia.) may havebeen yolking follicles for the breeding season thefollowing spring. A more detailed analysis of thereproductive cycle, based upon a much larger sam-ple (161 specimens), is in progress (Tait et al., inprep.).

Females collected February–June with mean fol-licle dia. �2.0 mm possessed sperm in their sper-mathecae except for one specimen collected in Feb-ruary and two specimens from June (Table 1). Thespecimen from February has a spermatophore cap inits cloaca and sperm apparently had not yet mi-grated into the spermathecae. Females with meanfollicle dia. �1.6 mm lack sperm in their spermathe-cae, except for a spent specimen collected 11 Novem-ber with a mean follicle size of 1.4 mm, which has afew sperm in some tubules, perhaps remnants fromthe previous breeding season (see below).

Spermathecae

The spermathecae consist of simple tubulo-alveolar glands that branch from an invaginationinto the roof of the anterior, tubular portion of thecloaca (Fig. 3A). This invagination was called the“dorsal tube” by Sever (1992b). The dorsal tube issimilar to the common tube characteristic of thePlethodontidae, except that the dorsal tube is not asnarrow and spermathecae evaginate from the entireupper half of the dorsal tube, whereas in plethodon-tids narrow neck tubules extend only from the apexof the common tube (Sever, 1992a).

The spermathecae extend into the loose connec-tive tissue tunica propria of the cloacal sheath (Fig.3B). In individuals in which sperm occur, some sper-mathecae contain many sperm, whereas others con-tain few or no sperm (Fig. 3B,C). Occasionally, distalportions of a spermathecal tubule have areas of des-quamated epithelium, and this condition apparentlyis not artifactual because adjacent tubules are nor-mal (Fig. 3B).

The most interesting aspect of the spermathecaeand of their relationship to the rest of the cloaca atthe LM level is the variation in the epithelium. Thecloacal cavity is lined with stratified, keratinizedepithelium continuous with the epidermis of the sur-rounding skin (Fig. 3A). As the dorsal tube invagi-nates, this lining changes into simple columnar mu-cinogenic cells, and this lining continues into theproximal neck tubules of the spermathecae (Fig.3C). The apical mucous product reacts PAS� forneutral carbohydrates and stains AB� for glycos-aminoglycans. This mucoid secretion, however, issomewhat different in the distal bulbs of the sper-mathecae, which are simple columnar or cuboidal,depending on their activity, as described below. Theapical cytoplasm of active distal bulbs is also AB�but is PAS–. All secretory regions of the spermathe-cal epithelium are BB– for proteins.

Ultrastructure of the spermathecae in Feb-ruary and April. With one exception, gravid fe-males collected February–April with ovarian folli-cles 2.0–3.9 mm mean dia. contain sperm in theirspermathecae, but some distal bulbs lack sperm(Fig. 4A), whereas others are packed with sperm(Fig. 4B). The one exception, mentioned previously,has a spermatophore cap in her cloaca, indicating arecent mating. Sperm in the spermathecae are notaligned in any orderly fashion. The distal bulbs aresimilar in cytology, however, whether sperm arepresent or not.

The epithelium is simple columnar with large het-erochromatic nuclei aligned with the long axis of thecell and occupying the basal halves of the cells (Fig.4A,B). Myoepithelial cells occur between the base ofthe epithelial cells and the basal lamina (Fig. 3A).The apical halves of the epithelial cells are filledwith irregularly shaped secretory vacuoles of mod-erate but uniform electron density. The product con-tained in the vacuoles is released into the lumen bya merocrine process (Fig. 4C). Golgi complexes andrough endoplasmic reticulum are found in the pe-rinuclear areas (Fig. 4D). Intercellular canaliculinarrow to tight junctions at the luminal border (Fig.4C), but in the basal half of the cell wide intercellu-lar spaces occur. In the widened areas of the inter-cellular canaliculi the plasma membranes from ad-jacent cells interdigitate (Fig. 4D).

Distal bulbs contain areas where the epithelium isdesquamated, so that complete or nearly completegaps exist between the lumen and the stroma of thesuperficial tunica propria (Fig. 5A). Sperm occurboth in the lumen and in the tunica propria, andsperm appear normal in cytology in both areas (Fig.5B,C). In a few instances, portions of sperm arefound embedded in the cytoplasm or in betweenepithelial cells in the intercellular canaliculi (Fig.5D). This is considered evidence of spermiophagy,because sperm trapped in these areas are unlikely tobe subsequently released (Sever, 1992a; Sever andHamlett, 1998).

Females collected in February and April withsmall ovarian follicles 0.9–1.1 mm mean dia. wereundergoing their first yolking and lack sperm intheir spermathecae. The distal bulbs at the LM levelappeared as cords of cells with no discernible lumen(Fig. 6A). Ultrastructural observations on the distalbulbs reveal that the cells contain large, uniformlydense nuclei that fill nearly the entire cell (Fig. 6B).Intercellular canaliculi are wide, with interdigitat-ing plasma membranes.

Neck tubules, however, do not vary much in ap-pearance between the mated females that possesslarge ovarian follicles and the unmated females thathave small ovarian follicles. Under both conditionsthe neck tubules contain numerous secretory vacu-oles. A gradation in quantity of secretory vacuolesoccurs, with fewest in the area adjoining the distalbulb (Fig. 6C) and the most in the region gradating

5TORRENT SALAMANDER SPERM STORAGE

Figure 3

6 D.M. SEVER ET AL.

into the dorsal tube (Fig. 6D). The secretory vacuolesof neck tubules are overall lighter in density thanthose of sperm-containing distal bulbs, and the se-cretory vacuoles in the neck tubules often contain aneccentric denser particle (Fig. 6D). Little secretoryproduct is apparent in the lumina of the neck tu-bules.

Ultrastructure of the spermathecae in June.The females collected in June with large ovarianfollicles have spermathecae similar in cytology tothe gravid females from February and April. Some ofthe neck tubules of spermathecae containing sperm,however, have fewer secretory vacuoles in the apicalcytoplasm and have abundant secretory product oc-cluding the lumina (Fig. 7A). Once again, in speci-mens containing sperm, distal portions of some sper-mathecae are empty, whereas others are packedwith irregularly aligned sperm (Fig. 7B).

Apical areas of some of the epithelial cells of thedistal bulbs are crowded with large, irregular secre-tory vacuoles, many of which contain central densematerial not noted in the April specimens (Fig. 7C),although others lack secretory vacuoles (Fig. 7D). Inthe cells that lack secretory vacuoles, numeroussmall vesicles are present and Golgi bodies andrough endoplasmic reticulum are often observed(Fig. 7D).

The specimens collected in June that lack spermand contain small ovarian follicles show somewhatdissimilar cytologies (Fig. 8). A female undergoingfirst yolking with ovarian follicles 1.2 mm mean dia.has constricted neck tubules and expanded distalbulbs (Fig. 8A). The epithelium varies in thicknessand nuclei are heterochromatic, basal, and havetheir long axes parallel to the basal lamina (Fig. 8B).Secretory vacuoles are numerous in the apical re-gions of some epithelial cells (Fig. 8B) and condens-ing vacuoles and organelles involved in synthesis ofsecretory products occur in the supranuclear areas(Fig. 8C). Thus, although this female is not in breed-ing condition for the current season, cytological ev-idence suggests an initiation of secretory activity sothat the glands would be ready for the next breedingseason.

Two other non-breeding females examined fromJune, one with ovarian follicles 1.0 mm mean dia.and the other with numerous nonyolked eggs (Table

1), may have been immature. The epithelial cells ofthe distal bulbs are nearly squamous and no indica-tion of secretory activity is apparent (Fig. 8D).

Ultrastructure of the spermathecae in Sep-tember and November. Only one of the femalesexamined from the fall had sperm in its spermathe-cae. This individual and other females collected inNovember with ovarian follicles 1.3–1.6 mm meandia. and thickened, flaccid oviducts are consideredspent individuals that oviposited the previousspring or summer and would likely have reachedbreeding condition again the following spring. In thedistal bulbs of these specimens the epithelial cellsare columnar, the nuclei are basal and parallel tothe long axes of the cells, and the apical halves of thecells are filled with secretory vacuoles (Fig. 9A).Some secretory product is found in the lumen (Fig.9A) and Golgi complexes occur in the supranuclearcytoplasm (9B). The apical portions of the neck tu-bules of these individuals are filled with large, irreg-ular secretory vacuoles that are once again lighter inoverall electron density than those of the distalbulbs (Fig. 9C). Golgi complexes and smooth endo-plasmic reticulum occur in the supranuclear areas ofneck tubules (Fig. 9D).

Finally, one specimen collected in November with1.4 mm mean follicle dia. possessed a small numberof sperm in several distal bulbs, although mostglands are devoid of sperm (Fig. 10A). The sper-mathecae are similar to those of other specimensfrom November except that the intercellular canal-iculi appear wider and membranous structures oc-cur in the lumen (Fig. 10C,D). The sperm do notappear normal. They are uniformly electron-denseand show no evidence of plasma membranes, theaxonemes, or mitochondrial rings (Fig. 10D). Thesesperm are considered degrading sperm that remainfrom a mating in the previous spring or summer.

Females with ovarian follicles 0.6–0.8 mm meandia. are considered first yolkers and they could havereached breeding condition the following spring orsummer. Spermathecae of these individuals are sim-ilar in cytology to those of females from February–June with follicles 0.9–1.1 mm mean dia.

DISCUSSION

Nussbaum and Tait (1977) reported that males ofRhyacotriton variegatus in Oregon are capable ofproducing spermatophores throughout the year, butthat males have reduced supplies of mature spermin their vasa deferentia from May to August. Zaliskoand Larsen (1988) reported that the vasa deferentiaof R. olympicus from Skamania County, Washing-ton, exhibit only minor seasonal variation and thatsperm typically can be found in the vasa deferentiaof mature males throughout the year. Neither Nuss-baum and Tait (1977) nor Zalisko and Larsen (1988)examined the spermatogenic cycle of the testes. Wealso found that sperm occur in the vasa deferentia of

Fig. 3. Rhyacotriton variegatus. A: Midsagittal paraffin sec-tion through the cloacal region of a female, stained withhematoxylin-eosin, and modified from Sever (1992a). B: Semithinepoxy section through the distal bulbs of a 60.8 mm SVL speci-men collected 10 April, stained with toluidine blue. C: Semithinepoxy section through the neck tubules and distal bulbs of a 53.0mm SVL specimen collected 18 June, stained with toluidine blue.Cc, cloacal chamber; Ct, cloacal tube; Db, distal bulb; De, desqua-mated epithelium; Dt, dorsal tube; Ep, epidermis; Mc, mucouscells; Nsl, no sperm in the lumen; Nt, neck tubules; Pi, posteriorintestine; Sc, spermatophore cap; Spl, sperm in the lumen; Tp,tunica propria.

7TORRENT SALAMANDER SPERM STORAGE

Fig. 4. Rhyacotriton variegatus. TEM through the spermathecae of a 60.8 mm SVL female collected 10 April that contained spermand had ovarian follicles 3.5 mm mean dia. A: Portion of a distal bulb lacking sperm in the lumen. B: Distal bulb containing luminalsperm. C: Apical cytoplasm showing a stage in release of secretory product. D: Perinuclear cytoplasm showing synthetic organelles.Go, Golgi apparatus; Ic, intercellular canaliculi; In, interdigitating plasma membranes; Lu, lumen; Me, melanin granules; My,myoepithelial cell; Nsl, no sperm in lumen; Nu, nucleus; Rer, rough endoplasmic reticulum; Se, secretory product; Spl, sperm in thelumen; Sv, secretory vacuoles; Tj, tight junction; Tp, tunica propria.

Fig. 5. Rhyacotriton variegatus. TEM through the spermathecae of a 60.8 mm SVL female collected 10 April that contained spermand had ovarian follicles 3.5 mm mean dia. A: Desquamated area of a spermathecal tubule. B: Luminal border. C: Stromal border.D: Sperm embedded in the epithelium. Bl, basal lamina; Cf, collagen fibers; Ic, intercellular canaliculi; Lu, lumen; Me, melaningranules; Mi, mitochondria; Mpt, middle piece of the tail; Nu, nucleus; Ppt, principle piece of the tail; Sn, sperm nucleus; Sp, sperm;Sv, secretory vacuoles; Tj, tight junction; Tp, tunica propria.

Fig. 6. Rhyacotriton variegatus. Semithin epoxy section (A) and TEMs (B–D) through the spermathecae of a 48.6 mm SVLspecimen collected 10 April lacking sperm and possessing ovarian follicles 0.05 mm mean dia. A: Overview of the spermathecal region,showing areas (B–D) illustrated in the electron micrographs. B: Distal bulb. C: Neck tubule. D: Junction between proximal portion ofneck tubule and dorsal tube of the cloaca. Ic, intercellular canaliculi; Lu, lumen; Nu, nucleus; Se, secretory product; Sv, secretoryvacuoles; Tp, tunica propria.

Fig. 7. Rhyacotriton variegatus. TEM through the spermathecae of a 53.0 mm SVL female collected 18 June that contained spermand had ovarian follicles 3.5 mm mean dia. A: Neck tubule. B: Distal bulb. C: Luminal border of distal bulb in an area containingnumerous secretory vacuoles. D: Luminal border of distal bulb in an area lacking secretory vacuoles. Go, Golgi apparatus; Ic,intercellular canaliculi; Lu, lumen; Mf, microfilaments; Mi, mitochondria; Mpt, middle piece of the tail; Nu, nucleus; Ppt, principlepiece of the tail; Rer, rough endoplasmic reticulum; Se, secretory product; Sn, sperm nucleus; Spl, sperm in the lumen; Ve, vesicles.

Fig. 8. Rhyacotriton variegatus. Light (A) and TEM (B–D) through the spermathecae of females that were collected 18 June andlacked sperm. One specimen (A–C) was 48.1 mm SVL and had ovarian follicles 1.2 mm mean dia., and the other (C,D) was 45.1 mmSVL and had ovarian follicles 1.0 mm mean dia. A: Several spermathecae and adjacent mucous cells (Mc) of the dorsal tube. B: Distalbulb epithelium. C: Supranuclear cytoplasm of a distal bulb epithelial cell. D: Epithelium of a distal bulb. Cv, condensing vacuoles;Db, distal bulb; Go, Golgi apparatus; Ic, intercellular canaliculi; Lu, lumen; Mc, mucous cells; Mi, mitochondria; Mv, microvilli; Nsl,no sperm in the lumen; Nt, neck tubule; Nu, nucleus; Rer, rough endoplasmic reticulum; Sv, secretory vacuoles; Tp, tunica propria.

Fig. 9. Rhyacotriton variegatus. TEM through the spermathecae of a 52.2 mm SVL female collected 1 November that lacked spermand had ovarian follicles 1.6 mm mean dia. A: Lumen and apical cytoplasm of a distal bulb. B: Supranuclear cytoplasm of a distal bulbepithelial cell. C: Neck tubule. D: Supranuclear cytoplasm of a neck tubule epithelial cell. Go, Golgi apparatus; Ic, intercellularcanaliculi; Lu, lumen; Mi, mitochondria; My, myoepithelial cells; Nu, nucleus; Se, secretory product; Ser, smooth endoplasmicreticulum; Sv, secretory vacuoles; Ve, vesicles.

Fig. 10. Rhyacotriton variegatus. Semithin epoxy section (A) and TEMs (B–D) through the spermathecae of a 52.2 mm SVL femalecollected 11 November that contained sperm and had ovarian follicles 1.0 mm mean dia. A: Overview of spermathecal distal bulbs,showing a few sperm in one tubule. B–C: Apical cytoplasm and adjacent lumen of a distal bulb containing sperm. D: Detail of spermin the lumen. Ic, intercellular canaliculi; Lu, lumen; Mb, membranous structures; Mv, microvilli; Nsl, no sperm in lumen; Se, secretoryproduct; Sn, sperm nucleus; Sp, sperm; Spl, sperm in lumen; Sv, secretory vacuoles; Tp, tunica propria; Ve, vesicles.

our specimens at all times and, furthermore, thiscondition results from active spermatogenic activitythroughout the year.

Males, therefore, seem capable of delivering ma-ture sperm throughout the year, and females pos-sess spermathecae for sperm storage that presum-ably could allow mating to occur long beforeoviposition. In Oregon, Nussbaum and Tait (1977)reported finding spermatophores in the cloacae offemales in every month except August, September,December, and January. In the current study, how-ever, the only mated females were those with largevitellogenic follicles in the February, April, and Junesamples, with the exception of one individual fromNovember that apparently retained some residualsperm from an earlier mating. Our results indicatean annual egg-laying cycle, because we found nomature, large-sized females in the spring with smalleggs, which we would expect if females needed 2years to reyolk after oviposition. In western Oregon,Nussbaum and Tait (1977) also found evidence thatmost females oviposit in the spring and early sum-mer on an annual cycle.

More is known about the ultrastructure of spermstorage in salamanders than in any other vertebrategroup. Nevertheless, the ultrastructure of femalesperm storage in salamanders has been studied inonly 2% of the known species of salamanders (Sever,2002). Besides the Rhyacotritonidae, the annual cycleof sperm storage has been studied at the ultrastruc-tural level in representatives of five of the remainingsix salamander families in which female sperm stor-age occurs. These studies involve two species of Pleth-odontidae (Sever, 1991c, 1992a, 1997; Sever and Bru-nette, 1993), three Salamandridae (Brizzi et al., 1995;Sever et al., 1996a, 1999, 2001), two Ambystomatidae(Sever, 1995; Sever and Kloepfer, 1993; Sever et al.,1995), one Amphiumidae (Sever et al., 1996b), and oneProteidae (Sever and Bart, 1996). Dicamptodontidae isthe only family in which the spermathecal ultrastruc-ture of a representative species has not been studied.Much diversity exists in reproductive habits andsperm storage characters among the species that havebeen studied.

Sever and Brizzi (1998) made an initial attempt tofind phyletic trends in sperm storage characters.They mapped 14 characters involved with sper-mathecae and sperm storage on a phylogeny ofsalamander families taken from Larson and Dim-mick (1993). The only character with definite phyl-etic value is whether sperm storage occurs in a “com-plex spermatheca” composed of a single compoundtubulo-alveolar gland (Plethodontidae) or in “simplespermathecae” consisting of numerous simple tubu-lar glands (other families). The variation in complexspermathecae of plethodontids was described bySever (2000).

Sever and Brizzi (1998) concluded that: 1) spermstorage is an ancient trait in salamanders, evolvingin the common ancestor of all the extant families in

the Salamandroidea; 2) some of the differences ob-served among taxa in spermathecal characters maynot be phyletically informative but related to otherspecies-specific reproductive adaptations; 3) spermstorage is apparently obligatory prior to fertilizationin salamandroids so that the duration of effectivesperm storage must be considered in any study onthe reproduction of these taxa; and 4) storage ofsperm facilitates multiple matings and provides theconditions for sperm competition within the sper-mathecae of salamanders (Halliday, 1998).

Sever (2002) performed a phenetic analysis of thesame dataset used by Sever and Brizzi (1998) withthe inclusion of data on Rhyacotriton variegatus.The simple matching coefficient used by Sever(2002) gives a different perspective than a cladisticanalysis. As mentioned previously, in the cladisticanalysis of Sever and Brizzi (1998) the only charac-ter found to have phyletic value was complex (pleth-odontids) vs. simple (other salamanders with spermstorage) spermathecae. In the phenetic analysis,other characters dealing with the types and distri-bution of epithelial cells and their relationship tospermiophagy assume more importance (Sever,2002).

Three groups were found in the phenetic analysis,and one of these groups was a clustering of Euryceacirrigera and Rhyacotriton variegatus. This cluster-ing is most interesting, because these are thespermathecae-bearing taxa in this sample thatbreed in rocky, swift-flowing streams (Nussbaum,1969; Nussbaum and Tait, 1977; Sever, 1999). Rhya-cotriton variegatus has a limited range in the up-lands of northwestern California and southwesternOregon (Petranka, 1998). Eurycea cirrigera occurswidely in the eastern United States and Canada andhas somewhat broader habitat tolerances, but thepopulation in which sperm storage has been studiedoccurs in rocky, wooded streams in the hilly, ungla-ciated Ohio River valley of Indiana (Sever, 1991c,1992a; Sever and Brunette, 1993). Rhyacotriton var-iegatus is in the family Rhyacotritonidae and E.cirrigera is in the Plethodontidae, two families thatlack a close sister-group relationship (Larson andDimmick, 1993).

These two species share these spermathecal sim-ilarities: secretory vacuoles of uniform density (oth-erwise known only in Notophthalmus viridescens);regionalized differences in secretory activity; differ-ent types of epithelial cells proximally and distally;regionalized spermiophagy; and desquamation ofepithelial cells into the stroma. The evolution ofsome of these characters may be linked within ataxon. In both species, the proximal ends of thespermathecal tubules have greater secretory activ-ity than the distal portions, and spermiophagy oc-curs only in the distal portions. Why these charac-ters have evolved convergently in two species withsimilar breeding habits and habitat begs even thewildest speculation at this time. In other words, do

15TORRENT SALAMANDER SPERM STORAGE

these characters have some adaptive value tosalamanders breeding in swift-flowing, rockystreams, and if so, why are these traits favored? Weare far from understanding the significance of thevariation of spermathecal characters in the repro-ductive activities of Rhyacotriton variegatus or anyother urodele.

Indeed, we still need to understand why femaleRhyacotriton variegatus store sperm. The matingseason is long (at least February–June), and thespecies aggregates in breeding sites along streams,so access to mates does not seem limited. Perhapsthe possibilities of sperm competition arising frommultiple matings are enhanced through sperm stor-age, as discussed by Sever et al. (2001) and Sever(2002) for the salamandrid Triturus vulgaris. Onemating is surely sufficient to fertilize the smallclutches of eggs produced by R. variegatus. However,we know nothing about the mating dynamics of R.variegatus. Of especial interest would be knowledgeof the duration of sperm storage between matingand oviposition. Sever (1996) proposed that short-term sperm storage (2 days or less) as found in manyAmbystoma is plesiomorphic relative to long-termsperm storage (weeks or months) documented forsome species. Perhaps sperm storage in R. variega-tus simply provides a time period for the female tomove from the area of mating activity (unknown, butpresumably the stream bank) to the place of eggdeposition (cracks in submerged rocks, under gravelor boulders in first order streams; Nussbaum, 1969;Karraker 1999; Russell et al., 2002). Or perhapssperm storage is an atavism of little significance,retained from the ancient origins of the Rhyacotri-tonidae. Further exploration of these topics will oc-cur in subsequent studies involving SEM of spermstorage (Sever et al., in prep.) and a more detailedanalysis of the reproductive cycle of R. variegatus(Tait et al., in prep.).

ACKNOWLEDGMENTS

We thank Elizabeth Ryder and Joel Thompson foraid in the collections, and Adrian Kirby and EmilyMoriarty for help in the laboratory. This is publica-tion number 25 from the Saint Mary’s College Elec-tron Microscopy Facility.

LITERATURE CITED

Brizzi R, Delfino G, Selmi MG, Sever DM. 1995. The spermathe-cae of Salamandrina terdigitata (Amphibia: Salamandridae):patterns of sperm storage and degradation. J Morphol 223:21–33.

Dawes C. 1979. Biological techniques for transmission and scan-ning electron microscopy. Burlington, VT: Ladd Research In-dustries.

Good DA, Wake DB. 1992. Geographic variation and speciation inthe torrent salamanders of the genus Rhyacotriton (Caudata:Rhyacotritonidae). U Calif Publ Zool 126:1–91.

Halliday T. 1998. Sperm competition in amphibians. In: BirkheadTR, Møller AP, editors. Sperm competition and sexual selection.New York: Academic Press. p 465–502.

Humason GL. 1979. Animal tissue techniques, 4th ed. San Fran-cisco: WH Freeman.

Karraker NE. 1999. Rhyacotriton variegatus (southern torrentsalamander) nest site. Herpetol Rev 30:160–161.

Kiernan JA. 1990. Histological and histochemical methods: the-ory and practice. New York: Pergamon Press.

Larson A, Dimmick WW. 1993. Phylogenetic relationships of thesalamander families: an analysis of congruence among morpho-logical and molecular characters. Herpetol Monogr 7:77–94.

Larson A, Weisrock DW, Kozak KH. 2003. Phylogenetic system-atics of salamanders (Amphibia, Urodela), a review. In: SeverDM, editor. Reproductive biology and phylogeny of Urodela(Amphibia). Enfield, NH: Science Publishers. p 31–108.

Nussbaum RA. 1969. A nest site of the Olympic salamander,Rhyacotrition olympicus (Gaige). Herpetologica 25:277–278.

Nussbaum RA, Tait CK. 1977. Aspects of the life history andecology of the Olympic salamander, Rhyacotriton olympicus(Gaige). Am Midl Nat 98:176–199.

Petranka JW. 1998. Salamanders of the United States and Can-ada. Washington: Smithsonian Inst Press.

Russell KR, Gonyaw AA, Strom JD, Diemer KE, Murk KC. 2002.Three new nests of the Columbia torrent salamander, Rhyaco-triton kezeri, in Oregon with observations of nesting behavior.Northw Nat 83:19–22.

Sever DM. 1991a. Comparative anatomy and phylogeny of thecloacae of salamanders (Amphibia: Caudata). I. Evolution atthe family level. Herpetologica 47:165–193.

Sever DM. 1991b. Comparative anatomy and phylogeny of thecloacae of salamanders (Amphibia: Caudata). II. Crypto-branchidae, Hynobiidae and Sirenidae. J Morphol 207:283–301.

Sever DM. 1991c. Sperm storage and degradation in the sper-mathecae of the salamander Eurycea cirrigera (Green). J Mor-phol 210:71–84.

Sever DM. 1992a. Spermiophagy by the spermathecal epitheliumof the salamander Eurycea cirrigera. J Morphol 212:281–290.

Sever DM. 1992b. Comparative anatomy and phylogeny of thecloacae of salamanders (Amphibia: Caudata). VI. Ambystoma-tidae and Dicamptodontidae. J Morphol 212:305–322.

Sever DM. 1994. Observations on regionalization of secretoryactivity in the spermathecae of salamanders and comments onphylogeny of sperm storage in female salamanders. Herpeto-logica 50:383–397.

Sever DM. 1995. Spermathecae of Ambystoma tigrinum (Am-phibia: Caudata): development and a role for the secretions.J Herpetol 29:243–255.

Sever DM. 1997. Sperm storage in the spermatheca of the red-back salamander, Plethodon cinereus (Amphibia: Plethodonti-dae). J Morphol 234:131–146.

Sever DM. 1999. Eurycea cirrigera. Cat Am Amphib Rept 648:1–6.

Sever DM. 2000. Sperm storage in female plethodontids withespecial reference to the Desmognathinae. In: Bruce RC, JaegerR, Houck, LC, editors. The biology of plethodontid salamanders.New York: Kluwer Academic/Plenum. p 345–369.

Sever DM. 2002. Female sperm storage in amphibians. J Exp Zool292:165–179.

Sever DM. 2003. Courtship and mating glands. In: Sever DM,editor. Reproductive biology and phylogeny of Urodela (Am-phibia). Enfield, NH: Science Publishers. p 323–381.

Sever DM, Bart HL. 1996. The ultrastructure of the spermathe-cae of Necturus beyeri (Amphiba: Proteidae) in relation to itsbreeding season. Copeia 1996:927–937.

Sever DM, Brizzi R. 1998. Comparative biology of sperm storagein female salamanders. J Exp Zool 282:460–476.

Sever DM, Brunette NS. 1993. Regionalization of eccrine andspermiophagic activity in the spermathecae of the salamanderEurycea cirrigera (Amphibia: Plethodontidae). J Morphol 217:161–170.

16 D.M. SEVER ET AL.

Sever DM, Hamlett WC. 1998. Sperm aggregations in the sper-mathecae of female desmognathine salamanders. J Morphol238:143–155.

Sever DM, Kloepfer NM. 1993. Spermathecal cytology of Ambys-toma opacum (Amphibia: Ambystomatidae) and the phylogenyof sperm storage in female salamanders. J Morphol 217:115–127.

Sever DM, Krenz JD, Johnson KM, Rania LC. 1995. Morphologyand evolutionary implications of the annual cycle of secretion andsperm storage in spermathecae of the salamander Ambystomaopacum (Amphibia: Ambystomatidae). J Morphol 223:35–46.

Sever DM, Rania LC, Krenz JD. 1996a. The annual cycle of spermstorage in the spermathecae of the red-spotted newt, Notoph-thalmus viridescens (Amphibia: Salamandridae). J Morphol227:155–170.

Sever DM, Doody JS, Reddish CA, Wenner MM, Church DR.1996b. Sperm storage in spermathecae of the great lamper eel,

Amphiuma tridactylum (Caudata: Amphiumidae). J Morphol230:79–97.

Sever DM, Halliday T, Waights V, Brown J, Davies HA, MoriartyEC. 1999. Sperm storage in females of the smooth newt (Tritu-rus v. vulgaris L.). I. Ultrastructure of the spermathecae duringthe breeding season. J Exp Zool 283:51–70.

Sever DM, Halliday T, Moriarty EC, Arano B. 2001. Sperm stor-age in females of the smooth newt (Triturus v. vulgaris L.). II.Ultrastructure of the spermathecae after the breeding season.Acta Zool 82:49–56.

Uribe Aranzabal MC. 2003. The testes, spermatogenesis andmale reproductive ducts. In: Sever DM, editor. Reproductivebiology and phylogeny of Urodela (Amphibia). Enfield, NH:Science Publishers. p 183–202.

Zalisko EJ, Larsen JH Jr. 1988. Ultrastructure and histochemis-try of the vas deferens of the salamander Rhyacotriton olympi-cus: adaptations for sperm storage. Scan Micros 2:1089–1095.

17TORRENT SALAMANDER SPERM STORAGE