INTRODUCTION
Chemosynthesis-based communities supported by reducing environments are distributed widely throughout the World Ocean. Despite their broad occurrence, these communities are scattered across the ‘regular’ deep-sea floor, interspersed among areas that are not suitable for hydrothermal vent and seep animals. In this aspect, reproductive biology of the organisms inhabiting chemosynthesis-based communities is particularly important, as it may improve our understanding of the mechanisms involved in their dispersal among remote habitats and recruitment into reducing communities.
Although a considerable body of data has been collected on the life histories of the main groups of chemosynthesis-based communities, the information on reproductive biology remains fragmented (see Tyler & Young, Reference Tyler and Young1999 for review). Among bivalves, some reproduction data has been collected for Acharax alinae (Solemyidae) (Beninger & Le Pennec, Reference Beninger and Le Pennec1997), Bathypecten vulcani (Pectinidae) (Le Pennec et al., Reference Le Pennec, Le Pennec, Beninger and Dufour2002), some species of Bathymodiolus (Mytilidae) (Le Pennec & Beninger, Reference Le Pennec and Beninger1997; Eckelbarger & Young, Reference Eckelbarger and Young1999; Healy et al., Reference Healy, Buckland-Nicks and Barrie2001; Colaço et al., Reference Colaço, Martins, Laranjo, Pires, Prieto, Costa, Lopes, Rosa, Dando and Serrão-Santos2006; Dixon et al., Reference Dixon, Lowe, Miller, Villemin, Colaco, Serrao-Santos and Dixon2006; Tyler et al., Reference Tyler, Young, Dolan, Arellano, Brooke and Baker2007), and some representatives of vesicomyids (Berg, Reference Berg1985; Endow & Ohta, Reference Endow and Ohta1990; Cary & Giovanni, Reference Cary and Giovannoni1993; Lisin et al., Reference Lisin, Hannan, Kochevar, Harrold and Barry1997; Fujiwara et al., Reference Fujiwara, Tsukahara, Hashimoto and Fuikura1998; Heyl et al., Reference Heyl, Gilhooly, Chambers, Gilchrist, Macko, Ruppel and Van Dover2007). It has been shown that bivalves from chemosynthesis-based communities do not have a common reproductive strategy (Tyler & Young, Reference Tyler and Young1999), show different types of development (planktotrophic or lecithotrophic) and can have either the presence or absence of reproductive seasonality. Furthermore, reproduction patterns are conservative and considerably phylogenetically-constrained (Eckelbarger & Watling, Reference Eckelbarger and Watling1995).
Vesicomyid bivalves are consistently one of the dominant components of reducing communities. This family is morphologically disparate and comprises about a hundred described living species. These highly specialized molluscs live in symbiosis with sulphide-oxidizing bacteria allocated in their gills.
Information regarding the reproductive patterns of vesicomyids is rather limited. Further study of these patterns would help in the understanding of the mechanisms of establishment and maintenance of the broad, though disconnected, areas characteristically inhabited by some vesicomyids (Kojima et al., Reference Kojima, Fujikura and Okutani2004; Krylova & Janssen, Reference Krylova and Janssen2006; Krylova & Sahling, Reference Krylova and Sahling2006). Moreover, comparing the reproductive patterns of vesicomyids with other chemosymbiotic bivalves found in reducing communities could help differentiate adaptive strategies specific to the family. Additionally, the morphology of the reproductive system, particularly the structure of spermatozoids, is helpful for the study of phylogenetic relationships and taxonomy (Popham, Reference Popham1979; Healy, Reference Healy and Taylor1996).
Chemosynthetic-based communities associated with methane seepage have been recently discovered in the bathyal zone off central Chile (Concepción Methane Seep area (CMSA)) (Sellanes et al., Reference Sellanes, Quiroga and Gallardo2004). A diverse assemblage of chemosymbiotic bivalves (including some species of vesicomyids, a lucinid, two thyasirids and a solemyid) has been reported for this region (Holmes et al., Reference Holmes, Oliver and Sellanes2005; Oliver & Sellanes, Reference Oliver and Sellanes2005; Sellanes & Krylova, Reference Sellanes and Krylova2005), with the most abundant species being the vesicomyid Calyptogena gallardoi (Sellanes & Krylova, Reference Sellanes and Krylova2005). The genus Calyptogena is widely distributed throughout the Pacific Ocean, occurring exceptionally along continental and island margins. However, in the south-eastern Pacific, C. gallardoi remains the only species of the genus described so far.
This paper aims to present data on the reproduction patterns of C. gallardoi. The gonadal structure and morphology of gametes were studied using a qualitative analysis made through the observation of serial histological sections and scanning electronic microscopy (SEM). We also report on the external morphological aspects of the shell which give evidence for the presence of sexual dimorphism, a trait uncommon among bivalves, though characteristic for the genus Calyptogena (Coan et al., Reference Coan, Scott and Bernard2000; Krylova & Sahling, Reference Krylova and Sahling2006).
MATERIALS AND METHODS
The study site is located 72 km north-west off Concepción Bay, Chile (36° 27.87 S 73° 43.25 W), in the upper slope zone at a water depth of 740 m to 870 m (Figure 1). General sampling was conducted onboard AGOR ‘Vidal Gormáz’ of the Chilean Navy during October 2004 (VG-04 cruise) and September 2006 (SeepOx cruise). The biological samples were collected by an Agassiz trawl (AGT; mouth opening 1.5 × 0.5 m, mesh size 10 × 10 mm at the cod end), in 20 minute hauls. Animals were sorted from the non-biological material and preserved onboard. A total of 21 specimens of C. gallardoi were allocated for this study. Seven specimens (5 females and 2 males) were collected during the VG-04 cruise and 14 (12 females and 2 males) during the SeepOx cruise.
Fig. 1. Study area off Concepción Bay, central Chile. The triangle indicates the area where most of the trawlings that retrieved living specimens of Calyptogena gallardoi were performed. The star indicates the position in which the presence of shallow sub-surface gas hydrates has been observed.
Samples for SEM were fixed with 3% glutaraldehyde in a 0.025 M solution of cacodylate in micro-filtered seawater (pH: 7.2–7.4). The samples were then rinsed in cacodylate buffer, dehydrated in an ascending ethanol gradient from 20% to absolute ethanol, critical-point dried, mounted on tape, coated under vacuum with gold and examined with a JEOL T-300 scanning electron microscope.
Samples for histological analysis were preserved in Davidson's fixative for 24 hours, then rinsed in micro-filtered seawater, dehydrated from 30% to absolute ethanol and included in paraffin blocks. The 5 µm slices were dyed using haematoxylin–eosin according to standard technique and mounted for observation under light microscopy. The gonad morphology was studied using serial cuts from one specimen of each gender. The diameter of at least 100 oocytes was measured in three females in order to construct size–frequency histograms. Measurements were limited to oocytes with visible nucleoli.
A morphometric analysis of 14 specimens (only those with unbroken valves) was performed using digital images, which were analysed using the software Image-Pro Plus. Different measurements, including the total area of the valves, were obtained for each specimen. As an indicator of sexual dimorphism, the relation of total area over the length to height ratio was used to compare the valves of males and females. This relation should be higher for rounded shells and smaller for more elongated shells, which have a larger length to height ratio. The Student's statistical test (two way t-test; Zar, Reference Zar1998) was used to asses morphological differences between sexes.
RESULTS
Calyptogena gallardoi is a gonochoric species. In both sexes the large gonads are embedded in the posterior–dorsal part of the visceral mass, behind the digestive gland, and surrounding the reduced digestive tube (Figure 2A). Slit-like genital apertures are located near the base of the posterior pedal retractor. The gonad has a lobular structure with many divisions that contain central tubes branching into smaller, tubular acini.
Fig. 2. Male gonad and lining epithelium. (A) General view of the gonad; (B) lining tissue. ac, acini with spermatozoids; cse, cubic simple epithelium; ct, connective tissue, i, intestine; lm, longitudinal musculature.
Female gonad
The internal surfaces of the evacuator conduits found in the acini are lined by cilia that form a bulky, or ‘paddle’, structure in the distal portion (Figure 3A, B); the length of the cilia is 31.6±4.5 µm (mean±1 SD).
Fig. 3. Section of the ovary. (A) Light microscopy view of the evacuator conduit; (B) scanning electronic microscopy view of the evacuator conduit with the ciliae. ecl, evacuator conduit lumen; ecw, evacuator conduit wall; pc, paddle cilia.
Acini walls are lined by a thin, germinal epithelium where it is possible to distinguish oogonia with an average diameter of 11.6±2.5 µm (Figure 4A). The oogonia originate oocytes I of an average diameter of 17.7±2.4 µm (Figure 4A). Vitellogenesis appears to start once oocyte diameters reach 93.1±23.2 µm (Figure 4B, C) expanding to 190.1± 27.0 µm (Figure 4B). The previtellogenic oocytes attach to the acini walls through strong pedicels that get longer as the development advances (Figure 4B). The mature oocytes measured 273.8±23.1 µm in diameter, and generally they are localized in the lumen of the acini. They group in large-sized platelets and are characterized by a great quantity of yolk. The germinal vesicle (Figure 4D) represents ~32% of the diameter of the mature oocytes and ~50% of the diameter of the previtellogenic oocytes. Within each gonad, and even within individual acini, sexual cells of all different developmental stages can be found (Figure 4D). Oocyte size–frequency analysis of 3 representative mature females demonstrated that, in spite of the predominance in number of previtellogenic oocytes, there is also a significant share of mature oocytes (Figure 5).
Fig. 4. Oocyte development stages within the ovary of Calyptogena gallardoi. (A) Early stages; (B) advanced stages; (C) previtellogenic oocytes; (D) mature stage. aw, acini wall; gv, germinal vesicle; mo, mature oocyte; nppo, non-pedunculated previtellogenic oocyte; oo, oogonia; p, pedicel; ppo, pedunculated previtellogenic oocyte; vo, vitellogenic oocyte; yp, yolk platelets.
Fig. 5. Oocyte size–frequency distribution for three representative mature specimens of Calyptogena gallardoi, and image that exemplifies the presence of all oocyte development stages within the same gonad. mo, mature oocyte; nppo, non-pedunculated previtellogenic oocyte; oo, oogonia; ppo, pedunculated previtellogenic oocyte; vo, vitellogenic oocyte.
Male gonad
Tubular acini come together toward bigger acini. The parts of the wall of some larger acini located close to the genital aperture contain only mature sperm and are constituted of cubical ciliate epithelium composed of relatively big cells with a nucleus of great size. These regions of ciliate epithelium are limited by simple epithelium with flattened non-ciliated cells (Figure 2B).
The germinal epithelium of tubular acini is surrounded by a thin layer of connective tissue (Figure 2B). Spermatogonia, spherical in shape and granular in appearance, line the inside of the acini. The diameter of spermatogonia is 5.6±0.5 µm. Toward the lumen of acini, it is possible to distinguish sexual cells in sequentially mature stages—spermatocytes, spermatides and spermatozoids. The mature spermatozoid has an elongated, bullet-shaped head measuring 3.7±0.2 µm in length from the tip of the acrosome to the base of middle piece constituted by a complex of 4 spherical mitochondria (Figure 6). The length of the flagellum is 26.5±2.5 µm.
Fig. 6. Lateral and posterior views of the head of spermatozoids of Calyptogena gallardoi. H, head of the spermatozoid; m, mitochondria; mp, middle plate.
Morphometric analysis of the valves
Calyptogena gallardoi has a stout shell up to 45 mm in length. The valves are elongate–elliptical, with a slightly convex, ventral margin. Significant differences in shell shape between males and females were found by comparing the relation of valve area to the ratio between the length and height. This relation, which is significantly higher in females (443.8) than in males (133.8) (t-test, P < 0.01), reflects that the shells of the females tend to be more rounded, while those of the males are more elongated (Figure 7). Males also have a more angulated posterior margin, sloping in its dorsal region. Furthermore, on average, the size of the female (31.4±7.3 mm) is significantly larger than the male (21.5±0.7; t-test, P = 0.017).
Fig. 7. Sexual dimorphism in Calyptogena gallardoi. External and internal views of the valves of a female (left) and of a male (right). Shells of females are larger, more rounded while males have a more angular posterior edge.
DISCUSSION
The new reproductive data gathered on C. gallardoi has been summarized in Table 1, along with the respective data for the seven other vesicomyid species that have been studied to date. These species have separate sexes, and the reported sex ratio for ‘Calyptogena’ magnifica is 1:1 (Berg, Reference Berg1985). All studied vesicomyids have large gonads embedded in the posterior–dorsal part of visceral mass adjacent to the digestive gland and surround the reduced gut. Genital apertures are slit-like and located in excurrent chambers near the base of posterior pedal retractor. The gonad of C. gallardoi, having tubular acini with branched divisions, follows this common pattern of gonadal organization.
Table 1. Reproductive characteristics of some representatives of the family Vesicomyidae.
Female reproductive system
In C. pacifica, ‘C.’ kilmeri and a vesicomyid from Blake Ridge, the developing gametes are arranged peripherally around a central lumen within reproductive acini, which simultaneously contain gametes in all stages of development (Lisin et al., Reference Lisin, Hannan, Kochevar, Harrold and Barry1997; Heyl et al., Reference Heyl, Gilhooly, Chambers, Gilchrist, Macko, Ruppel and Van Dover2007). The succession of the stages of oogenesis is typical for bivalves: oogonia, previtellogenic oocytes, pedunculated oocytes, vitellogenic oocytes and mature oocytes. In general, the gonadal organization in C. gallardoi is similar. The simultaneous presence of all developing stages of oogenesis would suggest that C. gallardoi displays continuous partial spawning, as it was proposed for other species of vesicomyids (Berg, Reference Berg1985; Heyl et al., Reference Heyl, Gilhooly, Chambers, Gilchrist, Macko, Ruppel and Van Dover2007).
The average diameter of the female gametes in C. gallardoi (ranging from 11.6±2.5 µm in oogonia to a maximum of 273.8±23.1 µm in mature oocytes) is larger than the average size of the oocytes of most bivalves, and corresponds to, approximately, the dimensions of oocytes of most vesicomyids (Table 1). Among vesicomyids, the largest oocyte diameter is found in ‘C.’ magnifica (Berg, Reference Berg1985) which reaches 482 µm. In this species, 1–2 eggs are enclosed in gelatinous capsules (Berg & Alatalo, Reference Berg and Alatalo1982).
In C. pacifica, a species of comparable size, the variation of the average diameter during the development of the oocytes is similar, and goes from 68.0–74.4 µm in previtellogenic oocytes, to 80–100 µm in vitellogenic oocytes, to 180–220 µm in mature oocytes (Lisin et al., Reference Lisin, Hannan, Kochevar, Harrold and Barry1997). The size of mature oocyte is related to the quantity of yolk and the type of feeding strategy of the larvae (Bruce, Reference Bruce1991; Lisin et al., Reference Lisin, Hannan, Kochevar, Harrold and Barry1997; Tyler & Young, Reference Tyler and Young1999). The large dimensions of the oocytes of vesicomyids indicate a lecithotrophic type of development. Beninger & Le Pennec (Reference Beninger and Le Pennec1997), in discussing the large oocytes of Acharax alinae, pointed out that the traditional interpretation of lecithotrophic development as short-term planktonic stage in conditions found in reducing habitats could be not true. Possibly, large amounts of yolk enable a prolonged larval stage to disperse through deep-sea oligotrophic areas which surround the scattered reducing environments. This could also be true for vesicomyids, and would explain the wide but disconnected areas inhabited by some species.
A particular type of ciliature was observed in the evacuator conducts of C. gallardoi. This ciliature has been described as modified normal cilia, and named by several authors as ‘paddle’ cilia (Ehlers & Ehlers, Reference Ehlers and Ehlers1978; Campos & Mann, Reference Campos and Mann1988; Deiner & Tamm, Reference Deiner and Tamm1991; Deiner et al., Reference Deiner, Tamm and Tamm1993). The paddle-shaped cilia were previously seen in the veliger larval stage of some molluscs. Campos & Mann (Reference Campos and Mann1988) reported the presence of disc-like cilia and palette cilia in larvae of the bivalves Mulinia lateralis and Spisula solidissima, and suggested a possible locomotive and sensorial function for them. There are also indications that the palette cilia (as well as other modifications in the ciliature of molluscs) are probably artefacts generated by the formaldehyde, sodium phosphate and sodium cacodylate solutions used during the SEM treatment (Ehlers & Ehlers, Reference Ehlers and Ehlers1978) or by the exposure of the samples to hypotonic media with variations in salinity (Deiner & Tamm, Reference Deiner and Tamm1991; Deiner et al., Reference Deiner, Tamm and Tamm1993). However, the results of this study suggest that the paddle-shaped cilia found in the evacuator conducts of C. gallardoi are not artefacts. Besides being observed through SEM, they were also found in gonadal tissue preserved in 70% ethanol and subjected to histological treatment. There are no reports of the occurrence of this type of artefact due to ethanol preservation. Paddle cilia in C. gallardoi are thought to have a transportation function inside the ovary, helping the movement of mature oocytes during their transit through the conduct. Further study of the nature of this ciliature is needed to confirm its function.
Male reproductive system
Spermatozoid morphology is a useful tool for taxonomic investigations, especially in such a taxonomically complicated group as vesicomyids. Data on spermatozoid morphology are available for ‘Calyptogena’ soyoae (Fujiwara et al., Reference Fujiwara, Tsukahara, Hashimoto and Fuikura1998); for ‘C.’ magnifica and C. pacifica only some data on dimensions has been published (Le Pennec unpublished, in Beninger & Le Pennec, Reference Beninger and Le Pennec1997). The species, ‘C.’ soyoae and C. gallardoi, have 4 spherical mitochondria in the middle piece. This number of mitochondria is typical for most bivalves and for all studied veneroids (Popham, Reference Popham1979). The shape of the spermatozoid head of C. gallardoi and ‘C.’ soyoae is slightly different: both species have a bullet-shaped head, but in C. gallardoi it is more elongated than in ‘C’ soyoae (Y. Fujiwara, personal communication).
The spermatozoids of C. gallardoi and those of its closely related species C. pacifica differ considerably in their flagellum length, with C. gallardoi having nearly twice the length of that found in C. pacifica (Le Pennec unpublished, in Beninger & Le Pennec, Reference Beninger and Le Pennec1997).
There is an aspect of the male gonadal structure of C. gallardoi that has not been reported in other vesicomyids. In C. gallardoi, the walls of some large acini, located close to the genital opening consist of two types of specialized (probably secretor) epithelia: (i) cubical ciliate epithelium, with relatively large cells, each containing a nucleus of great size; and (ii) simple epithelium with flattened, non-ciliated cells that limit these acini. In this zone, only mature sperm that is ready for spawning has been found. This suggests the zone is a place for sperm storage, which is a characteristic that has not been reported for other members of this family. A similar receptacle for storing sperm is reported for the hermaphroditic bivalves Mysella bidentata, Barrimysia siphonosomae (Jespersen & Lützen, Reference Jespersen and Lützen2001) and Pseudopythina ochetostomae (Jespersen et al., Reference Jespersen, Lützen and Morton2002). According to Jespersen & Lützen (Reference Jespersen and Lützen2001), it is not usual for a seminal receptacle to occur in gonochoric bivalves. The presence of a seminal receptacle in C. gallardoi would contribute to the continuous production of sperm. Once reaching maturity in the acini, sperm would pass into the seminal sac which would allow the acini to continue sperm producing. Further histological investigation would probably indicate such structure for other vesicomyids.
Reproductive cycle of vesicomyids
So far, information concerning temporal variations and seasonal-relative reproductive patterns of vesicomyids is unclear. A temporal investigation of gametogenesis was conducted only for ‘C.’ kilmeri and C. pacifica (Lisin et al., Reference Lisin, Hannan, Kochevar, Harrold and Barry1997). While in ‘C.’ kilmeri, analysis of gonadal development through the year suggest a peak in reproductive output during winter, in C. pacifica analyses of variations in reproductive tissues were insufficient to resolve seasonal changes. In both species however, there were oocytes of spawning size found during all sampling periods (Lisin et al., Reference Lisin, Hannan, Kochevar, Harrold and Barry1997).
In a vesicomyid from Blake Ridge, five reproductive condition stages were distinguished for the female gametes, all of them were simultaneously present and most females were in the developing and ripe reproductive stages, while all males were in ripe stage (Heyl et al., Reference Heyl, Gilhooly, Chambers, Gilchrist, Macko, Ruppel and Van Dover2007). A continuous reproductive pattern was also suggested for ‘C.’ magnifica given the simultaneous presence of all stages of gametes in the gonads (Berg, Reference Berg1985).
For ‘C.’ soyoae, over the course of 1.5 years, eleven in situ spawning events have been recorded at the same site (Fujiwara et al., Reference Fujiwara, Tsukahara, Hashimoto and Fuikura1998). These data, together with the TEM observations of gonads having similar maturity in specimens collected in summer and autumn, suggest a continuous spawning. The absence of seasonal peaks does not mean that there is no synchrony in reproduction efforts, at least among part of the population. Spawning of ‘C.’ soyoae was shown to be induced by a 0.1–0.2°C temperature rise. Males spawn first usually followed 7–11 minutes later by females releasing eggs. However, not every male spawning event was followed by the spawning of females (Fujiwara et al., Reference Fujiwara, Tsukahara, Hashimoto and Fuikura1998). This observation supports the view that the presence of a sperm sac could be a generalized feature in vesicomyids. Such a structure would allow for the storage of large volumes of spermatozoids ready to be spawned once unpredictable appropriate environmental conditions occur.
Sexual dimorphism
Although most bivalves are gonochoric, external sexual dimorphism is unusual. Saleuddin (Reference Saleuddin1964) describes the sexual dimorphism for Astarte elliptica, A. sulcata and A. borealis, reporting that the shell of the female is more elongated than that of the male. Except for these and a few other cases, the sex of bivalves can only be determined by direct examination of the gonads (generally under microscope) or by the observation of the spawning process (Sastry, Reference Sastry, Giese and Pearse1979).
For C. pacifica, variation in the shell shape was described by Dall (Reference Dall1981) and this was later suggested to demonstrate sexual dimorphism impressed in the shell shape: the females have larger and more expanded posterodorsal shells (J.P. Barry, personal communication, in Coan et al., Reference Coan, Scott and Bernard2000). Later the dimorphism was proposed for the whole genus Calyptogena (Krylova & Sahling, Reference Krylova and Sahling2006), but it was not confirmed.
Variations in the shape of the shell of C. gallardoi were reported by Sellanes & Krylova (Reference Sellanes and Krylova2005), who indicated the presence of more or less elongated specimens. They also stated that the outline of posterior margin varied from rounded to tapering. However, these observations were based on the valves of dead specimens. Currently, the parallel investigation of the reproductive system of C. gallardoi has allowed us to confirm that these morphs actually exhibit sexual dimorphism. Among vesicomyids described to date, sexual dimorphism is reported only for the genus Calyptogena s.s.
Concluding remarks
Many reproductive aspects of C. gallardoi such as the presence of external sexual dimorphism and the presence of large eggs, suggesting a lecithotrophic development of the larvae, confirm previous observations (often poorly documented) in other vesicomyids. Other aspects, like the presence of a sperm sac in males, as well as the evacuator conduits of the gonads lined by paddle cilia, though reported for different bivalves, are novel for this family. Future studies of the recently discovered seep habitats in the south-eastern Pacific will certainly shed light on further interesting biological aspects of this species, as well as on other representatives of the rich local fauna.
ACKNOWLEDGEMENTS
We thank the captain and crew of AGOR ‘Vidal Gormáz’ of the Chilean Navy for support at sea. We are also indebted to Karin Lohrmann and to Maria Soledad Romero (UCN, Facultad de Ciencias del Mar), who provided laboratory facilities for the histological treatment and help with the SEMs, respectively. Our thanks also go to Jennifer González and Carlos Neira (Scripps Institution of Oceanography, USA) and to two anonymous referees for providing corrections and constructive comments. We are grateful to Yoshihiro Fujiwara (JAMSTEC) for providing unpublished TEM images of spermatozoids of ‘Calyptogena’ soyoae for comparison and to Anatoly Drozdov (Institute of Marine Biology, Vladivostok, Russia) for fruitful discussion.
This work was funded by FONDECYT project No. 1061217 to J.S. and the research Direction and the Centre of Oceanographic Research in the Eastern-South Pacific (COPAS) of the University of Concepción. FONDECYT project No. 1061214 to Práxedes Muñoz; SCRIPPS Institution of Oceanography through NOAA Ocean Exploration programme, and the Office of Naval Research of the US Navy provided extra funding for ship time.
