1 Reproductive ecology of the deep-sea protobranch bivalves Yoldiella ecaudata, Yoldiella 2 sabrina, and Yoldiella valettei (Yoldiidae) from the Southern Ocean 3 4 Adam J. Reed1, James P. Morris1, Katrin Linse2, Sven Thatje1 5 6 7 8 9 1 University of Southampton, Ocean and Earth Science, National Oceanography Centre Southampton, European Way, Southampton, SO14 3ZH, UK 2 British Antarctic Survey, Natural Environment Research Council, High Cross, Madingley Road, Cambridge, CB3 0ET, UK 10 11 12 Corresponding Author: Adam J. Reed 13 email: ajr104@soton.ac.uk 14 15 16 17 18 19 20 21 22 23 24 1 25 Abstract 26 The protobranch bivalves of the Southern Ocean are poorly understood ecologically, despite 27 their high abundances in soft sediments from the shelf to the deep sea. The subclass has a 28 long evolutionary history pre-dating the formation of the polar front, and knowledge of 29 their reproductive biology is key to understanding better the successful radiation into the 30 Southern Ocean, and within deep-sea basins. In this study we investigate the reproductive 31 biology of three deep-water protobranchs for the first time; Yoldiella ecaudata from 500 m 32 in the Amundsen Sea; Y. sabrina from between 200 and 4730 m in the Amundsen Sea, 33 Scotia Sea, and South Atlantic; and Y. valettei from 1000 m in the Scotia Sea. All three 34 species demonstrate evidence of lecithotrophic larval development with maximum egg size 35 of 130.4 µm, 187.9 µm, and 120.6 µm in Y. ecaudata, Y. sabrina, and Y. valettei, respectively, 36 further supported by prodissoconch I measurements. There is evidence for simultaneous 37 hermaphroditism in Y. valettei. Asynchronous oocyte development within specimens of Y. 38 ecaudata and Y. valettei is described, and also between populations of Y. sabrina separated 39 by depth. The reproductive characteristics, comparable to those of North Atlantic deep-sea 40 protobranch species, are discussed in the context of the cold-stenothermal conditions 41 prevailing on the deep-Antarctic continental shelf and deep sea. The requirement for 42 reclassification of this complex subclass is also discussed in relation to observed soft 43 anatomy and shell characteristics. 44 45 46 2 47 Introduction 48 The Protobranchia are a subclass of small deposit feeding bivalves containing over 600 49 species globally, and represent one of the oldest clades of bivalve with fossil records dating 50 back to the early Ordovician ~500 Ma (Allen 1978). They share in common highly specialised 51 characteristics for their preferred fine sediment environment, including a strong gill 52 structure solely used for respiration, long hindguts for feeding in food impoverished 53 environments, and adapted feeding palps to consume large quantities of sediment (see 54 review by Zardus 2002). Most commonly found in the deep-sea, protobranchs are also well 55 represented in shallow water environments, and are of interest because of their long 56 evolutionary history (Allen 1978; Zardus 2002), potentially long life spans (Turekian et al. 57 1975; Nolan and Clarke 1993; Gage 1994; Peck and Bullough 1993; Peck et al. 2000), and 58 diverse ecology. 59 Despite representing between 50 and 90% of bivalve species on the continental slope and 60 abyssal plain, respectively, deep-sea protobranch bivalves are ecologically poorly 61 understood. They are considered an important macrofaunal taxon in the deep sea for their 62 sediment turnover and biomass (Sanders et al. 1965; Allen 2008), and yet their taxonomy 63 remains largely unclear, with many synonyms as a result of previous misidentifications of 64 new species (Verrill and Bush 1898; Knudson 1970). Allen and Hannah (1986), who state 65 that names used within the group are ‘a confusing array of regional uses that are not 66 collated’, reorganised the group omitting erroneous names, and subsequent studies have 67 outlined the defining characteristics of different genera (Allen and Sanders 1973; 1982, 68 1996; Allen and Hannah 1989; Allen et al. 1995). More recently, genetic tools have been 69 used to analyse widespread distributions of species revealing contrasting levels of genetic 3 70 differentiation and homogeneity among different species, often relating to bathymetry and 71 geography (Chase et al. 1998; Etter et al. 2005, 2011; Zardus et al. 2006). 72 Reproduction of deep-sea protobranchs has been studied in few species, and in most detail 73 from those collected in the Rockall Trough in the North Atlantic, where a research station at 74 ~2900m was sampled extensively between 1973 and 1983 (Lightfoot et al. 1979; Tyler et al. 75 1992), but also the Gay Head-Bermuda transect (Scheltema and Williams 2009). These 76 studies identified seasonality in Ledella pustulosa, Yoldiella jeffreysi, Nucula delphinodonta, 77 and N. annulata while asynchronous reproduction prevailed in Malletia cuneata, Ennucula 78 similis, E. granulosa, Deminucula atacellana, and Brevinucula verrilli, representing both 79 lecithotrophy and direct larval development. To date, this represents the most detailed 80 reproductive 81 chemosynthetic environments. 82 Protobranch bivalves have a unique, short dispersing pericalymma larva, in which the larval 83 shell develops beneath an external layer of cells and cilia (the test) to help movement 84 (Zardus and Morse 1998; Zardus and Martel 2001). This distinct shell development results in 85 a smoother but distinctly pitted prodissoconch I (first larval shell) and absent prodissoconch 86 II (second larval shell) (Ockelmann 1965). To date, known larval development in 87 protobranchs can be pelagic lecithotropthic or direct larval development with parental care, 88 commonly inferred by the larval shell and oocyte size (Ockelmann 1965; Scheltema and 89 Williams 2009), with prodissoconch and oocyte size above 135 and 90µm, respectively, 90 suggesting lecithotrophic development over planktotrophic development (Ockelmann 91 1965). biology known within deep-sea protobranch bivalves from non- 4 92 In the Antarctic, the protobranch bivalves are well represented between 1 and 5000m but 93 research has predominantly focussed on the ecology of the comparatively large and shallow 94 water protobranch Yoldia eightsii (Davenport 1988; Nolan and Clarke 1993; Peck and 95 Bullough 1993; Peck et al. 2004). Despite frequent sampling, even the reproduction of 96 Yoldia eightsii is still poorly understood and is demonstrative of the difficulties associated 97 with time constraints for seasonal sampling in the Antarctic. The Southern Ocean 98 protobranchs Yoldiella sabrina (Hedley 1916), Yoldiella ecaudata (Pelseneer 1903), and 99 Yoldiella valettei (Lamy 1906) are commonly found from the Antarctic shelf to the deep-sea 100 basins (200-4000m+) and have wide distributions (Gofas 2012) that are likely to be circum- 101 Antarctic, although locally distinct morphotypes have been identified (Reed et al. 2013a). 102 Additionally, Yoldiella sabrina has a known distribution outside of the Polar Front into the 103 South Atlantic (Linse et al. 2007). 104 This study, for the first time, describes the reproduction of deep-water Southern Ocean 105 bivalves Yoldiella ecaudata, Y. sabrina, and Y. valettei collected from the South Atlantic, 106 Amundsen Sea, and Scotia Sea. This knowledge will enhance our understanding of the 107 benthic ecology of protobranchs in the cold-stenothermal environments of the deep sea 108 and Southern Ocean, and provide an improved ecological framework for future studies on 109 biogeography and population connectivity in the Southern Ocean. The taxonomic status of 110 these species is also discussed in the context of anatomical observations made through 111 histology and comparisons with other genera. 112 Materials and Methods 113 Sample collection 114 Yoldiella valettei, Y. ecaudata, and Y. sabrina were collected during a number of research 115 expeditions to the Weddell Sea, Scotia Arc, Antarctic Peninsula, and Amundsen Sea between 5 116 2005 and 2008; ANDEEP III 2005 (Fahrbach 2005); BIOPEARL 2006 (Linse 2006); BIOPEARL II 117 2008 (Enderlein and Larter 2008) (Table 1). Benthic samples were collected primarily using 118 an epibenthic sledge (EBS) (Brandt and Barthel 1995; Brenke 2005) or Agassiz trawl (AGT), 119 and the deployment methods and conditions are described in the relating reports. Samples 120 containing the target protobranch species were taken in depths ranging from 192m to 121 4900m. Specimens collected by EBS were fixed in either 96% ethanol or 4% buffered 122 formaldehyde immediately after coming on deck and later sorted, while specimens 123 collected by AGT were sorted alive and then fixed in 96% ethanol or 4% buffered formalin. 124 Differences in fixing/sorting procedure were the consequence of different science priorities. 125 Histology 126 Histology was done according to the protocol of previous studies (Tyler et al. 1992; Reed et 127 al. 2013b). Specimens were dissected from their shell or decalcified in rapid decalcifying 128 solution (with HCL). Where possible, specimens were photographed before and after shell 129 removal for identification of anatomy post processing. Soft anatomy and internal shell 130 characteristics was used to confirm identification of species. 131 Whole animals were used for histology to avoid damaging gonads in dissection (see Higgs et 132 al. 2009). Samples were dehydrated in graded isopropanol and cleared in histoclear 133 (CellPath). Tissue was set in wax and serial sectioning of tissue was done at 6µm where 134 possible, to ensure the gonad was cut. 135 A total of 37 Yoldiella sabrina from 500m in the Amundsen Sea, 4730m in the South Atlantic 136 and from 200-3500m in the Scotia Sea; 34 Y. valettei from the 1000m in the Scotia Sea; and 137 15 Y. ecaudata from 500m in the Amundsen Sea, were used for histology, representing the 138 size range of each species where possible. Microphotographs were taken using a Nikon 6 139 D5000 mounted on a microscope and Feret diameter was used to measure egg size using 140 SigmaScan Pro 4 software. 141 Prodissoconch size 142 Prodissoconch length of ten Y. ecaudata and eleven Y. valettei were measured by scanning 143 electron microscopy (SEM) of the umbo region of the shell. Prodissoconch sizes of Y. sabrina 144 were not measured as only adult specimens with eroded umbones were available for this 145 study. Specimens for SEM were mounted onto stubs, coated with 5µm gold layer and 146 analysed using a Leo 1450 VP SEM. The maximum length and height of the prodissoconch I 147 (PI) were measured using sub-adult shells in good condition. Shells were positioned so that 148 the prodissoconch was lying flat to the SEM beam. Length measurements were taken tilting 149 the shell along the mid-axis until the maximum length was obtained. 150 To retain consistency when measuring the maximum length, the line of measurement was 151 always kept parallel with the hinge line. Maximum height was always perpendicular to the 152 length. The PI has been previously found to be directly linked to size of offspring, 153 representing the size of shell secreted by the larval stage, and so linked to egg size and 154 nutritional content (Ockelmann 1965; Goodsell and Eversole 1992). 155 Results 156 Reproductive ecology 157 Yoldiella sabrina 158 In total 25 animals were identified with active gonads and 12 with no identifiable gonads. 159 The gonad occupied an area dorsal laterally on both the left and right external to the 160 digestive gland and internally around the digestive tissue and gut loop (Fig. 1, 2). Oocytes 7 161 and testis were observed beside the intestinal loop wall and there was no evidence of 162 hermaphroditism (Fig. 2). 163 The smallest female was found at 3.3 mm shell length from the Scotia Sea with 164 previtellogenic oocytes observed internally around the digestive gland (Fig. 2a, b). Small 165 vitellogenic oocytes were observed at 5.5 mm shell length with up to 22 oocytes in 166 specimens from Scotia Sea (Fig. 3a), the largest measuring 55.7 µm in diameter. No more 167 than ten large vitellogenic oocytes were measured in specimens of 9.1 mm shell length from 168 the Scotia Sea and the largest female at 10.3 mm from the South Atlantic (Fig. 3a). The 169 largest eggs measured 187.9 µm and 174.0 µm respectively. Only one female with small 170 vitellogenic oocytes was observed in the Amundsen Sea at 5.8mm shell length. 171 Males were observed from 3 – 13 mm shell length, identified by extensive testis surrounding 172 the digestive glands, and hind gut (Fig. 2c). The right dorsal-ventral gonad surrounding the 173 hind gut appeared less extensive however, than the left dorsal ventral gonad (Fig. 2c). 174 Spermatogonia and spermatocytes were observed, but no accumulations of spermatozoans 175 were visible in the specimens from any location. Males were well represented in specimens 176 studied from the Scotia Sea, Amundsen Sea, and South Atlantic, with 16 males to 8 females 177 in all areas. A relationship of two males to every female, or higher, was also observed within 178 each area studied. 179 Yoldiella ecaudata 180 Nine specimens of Y. ecaudata were identified has having active gonads and in six no 181 reproductive tissue was found. The gonad is situated dorsal-laterally, external to the 182 digestive gland and around the hind gut (left and right), and internally around the digestive 183 gland (Fig. 4a digestive gland; 4b for gonad). Testis or oocytes were observed in areas 8 184 otherwise occupied by the digestive gland, which appeared reduced in specimens with well- 185 developed gonad; there is no evidence of hermaphroditism 186 Female Y. ecaudata were found between 2.2 and 2.8 mm shell length with the largest 187 oocyte measured 130.4 µm in a female of 2.8 mm shell length (Fig. 3b). Up to 13 eggs were 188 counted in a single female of 2.6 mm shell length. Only large vitellogenic oocytes were 189 observed in females although the size range measured varied within and between 190 individuals (Fig. 3b). Only three male Y. ecaudata were observed between 2.4 and 2.7 mm 191 shell length, but no evidence of spermatazoan accumulation. 192 Yoldiella valettei 193 In total, 21 specimens from the Scotia Sea were identified with active gonads. The gonad 194 occupied an area dorso-laterally on both the left and right sides of the body, external to the 195 digestive gland, and internally around digestive tissue and hind gut loop (Fig. 5). The gonad 196 extended dorsally to join the left and right portions of gonad (Fig. 5a). In males, the testis 197 extended anteriorly under the mantle epithelium (Fig. 5b). 198 Females were found between 2.2 and 3.2 mm shell length with vitellogenic oocytes 199 observed in the gonad. The largest oocyte measured 120.6 µm diameter and up to 56 200 oocytes were counted in a single individual (3.1 mm) (Fig. 3c). In one specimen of 3.1 mm 201 shell length, both testis and vitellogenic oocytes were observed with spermatozoa 202 accumulation in the lumen (Fig. 5a). Testis in the hermaphrodite was only observed dorso- 203 laterally on the right side, external to the digestive gland and hind gut loop. Male Y. valettei 204 were found between 2.3 and 3.2 mm shell length with well-developed testis often 205 enveloping the bivalve, externally to the digestive glands and along the mantle cavity 206 epithelium (Fig. 5b). Spermatogonia and spermatocytes were observed and accumulations 9 207 of spermatozoans were visible in some specimens. There was no evidence of protandry or 208 protogyny. 209 Prodissoconch measurements 210 Prodissoconch I was identified by a characteristic pitted surface. Both Y. valettei and Y. 211 ecaudata showed similar prodissoconch lengths of 184.4 µm ±7.2 SD (n=11) and 191.6 µm ± 212 4.0 SD (n=10), respectively (Fig. 6), and no prodissoconch II was observed in any specimens. 213 Discussion 214 Reproductive biology of deep-sea protobranch bivalves is not well known; species with 215 known reproduction include the nuculanid Ledella pustulosa and yoldiid Yoldiella jeffreysi 216 from 2900 m in the Rockall Trough, North Atlantic (Lightfoot et al. 1979; Tyler et al. 1992). 217 Gonad morphology and anatomy in all three studied Southern Ocean species are similar to 218 that described in Tyler et al (1992); the gonad is found external to the digestive gland and 219 hind gut loop, but histology also revealed oocytes and testis formation internal to the 220 digestive gland, and dorsally above the stomach. 221 Although not intended as a taxonomic revision, the deep-water protobranchs Y. valettei, Y. 222 ecaudata, and Y. sabrina appear most morphologically and anatomically similar to deep-sea 223 species of the genera Yoldiella (Allen et al. 1995), Ledella (Allen and Hannah 1989) and 224 Portlandia (Killeen and Turner 2009), respectively. These observations highlight the 225 requirement for a comprehensive morphological and molecular taxonomic revision of the 226 current Yoldiidae and related small sized protobranch families. The soft anatomy of Y. 227 ecaudata appears most similar to the nuculanid Ledella pustulosa with a characteristically 228 large stomach and extensive hind gut coil on both the left and right sides of the body, while 10 229 the shell is short and compact, rostrate with strong concentric sculpture and strong chevron 230 teeth (Allen and Hannah 1989). Sharma et al. (2013) regarded Y. ecaudata as Ledella 231 ecaudata in a recent molecular study, although without a taxonomic diagnosis. Gonad 232 morphology and oocyte size is also similar with L. pustulosa having a maximum oocyte size 233 of 120 µm (Tyler et al. 1992) while Y. ecaudata have maximum oocyte size of 130.4µm 234 (Table 1). Likewise, the similar Southern Ocean Y. valettei and North Atlantic Y. jeffreysi have 235 ovate and fragile shell structure with very fine concentric sculpture, and maximum oocyte 236 sizes of 120.6 µm and 120 µm respectively. While only representing a ‘snapshot’ of the 237 reproduction in these species, these comparisons of maximum oocyte size with regularly 238 sampled material from the North Atlantic may infer that the largest oocytes observed are 239 fully developed. Prodissoconch measurements of Y. valettei and Y. ecaudata together with 240 the oocyte sizes between 85 and 140 µm suggest a pelagic lecithotrophic development 241 (Ockelmann 1965), as found in other deep-sea protobranch bivalves (Tyler et al. 1992; 242 Scheltema and Williams 2009). 243 Yoldiella sabrina share more morphological characters with the genus Portlandia (Yoldiidae 244 although high levels of plasticity may cross over with characteristics of Yoldiella spp.; the 245 shell is oblong with an extended hinge margin, a distinct lunule and escutcheon, and a single 246 hind gut loop to the right of the body, (Verrill and Bush 1897, Allen and Hannah 1986, 247 Killeen and Turner 1999). Reproductively, over a range of depths and across biogeographic 248 ranges, we find a higher proportion of males, which contrasts to Y. ecaudata and Y. valettei 249 observed having one male to two females (Table 1). A maximum oocyte size for Y. sabrina of 250 187.9 µm is larger than observed in Y. ecaudata and Y. valettei and is close to the predicted 251 range for direct larval development (Ockelmann 1965). The mean prodissoconch I 11 252 measurements of Y. sabrina from published data suggest a length of approximately 200 µm 253 with a barely discernible prodissoconch II (Hain and Arnaud 1992), which would, however, 254 support lecithotrophic larval development. 255 Using an underlying knowledge of deep-sea protobranch life history published to date, 256 evidence of seasonal or continuous reproduction can be identified from our samples, 257 despite only having a snapshot of their reproductive cycle. Lecithotrophic larval 258 development could be uncoupled from seasonality of food supply in the Antarctic benthos 259 as found in the asteroid Porania sp. (Bosch and Pearse 1990). Protobranch bivalves are 260 specially adapted to living in food-impoverished conditions (Allen 2008) reducing the effect 261 of food limitation into the deep-sea (Gage and Tyler 1991). Cold Antarctic deep-sea 262 conditions can store high levels of organic matter in sediments, acting as an annual food 263 supply to deposit feeding fauna (Mincks et al. 2005; Glover et al. 2008), further reducing any 264 effect of seasonality. This study found variation in oocyte size within individuals of Y. 265 ecaudata and Y. valettei, which may be representative of asynchronous oocyte 266 development within individual specimens, as found for Y. jeffreysi in the North Atlantic 267 (Tyler et al. 1992). The smaller egg size range within Y. ecaudata could also represent an 268 earlier stage of oocyte development at smaller sizes. Oocyte development is often slow in 269 Southern Ocean and deep-sea molluscs taking between 12 and 24 months (Picken 1979, 270 1980; Richardson 1979; Brey and Hain 1992; Higgs et al. 2009; Reed et al. 2013b) and 271 multiple cohorts may exist within individuals. The smaller eggs seen in some specimens may 272 represent a second cohort of eggs being produced simultaneously to the current cohort. The 273 differences observed in egg size and depth in Y. sabrina could also indicate a difference in 274 seasonality of reproduction with depth, or an earlier stage of oocyte development with 12 275 smaller shell size. However, the bathymetric and subtle ecological differences between 276 samples may also have a role in the timing and tempo of reproduction at the benthos. 277 Additionally, biogeographic related plasticity in morphology and growth in the target species 278 have already been identified, and should not be underestimated in the Southern Ocean 279 (Reed et al. 2013a). 280 Pelagic lecithotrophic larval development not only has the advantage of being uncoupled 281 with primary productivity, but also offers the potential for long periods of dispersal, 282 essential for maintaining a homogenous circum-Antarctic distribution. Bosch and Pearse 283 (1990) determined the development period of the demersal and pelagic lecithotrophic 284 asteroids Porania sp. and Acodontaster hodgsoni to be 78 and 106 days, respectively, while 285 deep-sea Atlantic protobranchs have been modelled to disperse between 237 and 749 km 286 (McClain et al. 2012). Furthermore, a study into the metabolic rates and biomass of larvae 287 and eggs in Antarctic echinoderms suggested larval survival over extended periods in the 288 Antarctic, in contrast to much shorter periods in similar temperate species (Shilling and 289 Manahan 1994). Although untested in the field, the large egg size of the Antarctic 290 protobranch bivalves may provide the energetic requirements for long periods of 291 development, facilitating a wide dispersal of these species in the Southern Ocean. Deep- 292 Antarctic waters have complex currents driven by cold bottom water formation in the 293 Weddell Sea and the flooding of warmer waters from the Pacific. Together with the 294 Antarctic circum-polar current, continental counter currents, and large scale eddy formation 295 characteristic of the Weddell and Scotia Seas, larval dispersal has great potential if 296 suspension and survival can be maintained in the pelagic environment (for review see Thatje 297 2012). 13 298 The occurrence of hermaphroditism is rare in deep-sea protobranchs (Zardus 2002) 299 although not unusual in Southern Ocean molluscs, and bivalves, globally. Simultaneous and 300 sequential hermaphroditism has been described in Laternula elliptica (Bosch and Pearse 301 1988), Thracia meridonalis (Sartori and Domaneschi 2005), Mysella charcoti and M. narchii 302 (Passos and Domaneschi 2009), Lissarca notorcadensis (Prezant et al. 1992) and Lissarca 303 miliaris (Reed et al. 2013b). In the deep-sea context, simultaneous hermaphroditism could 304 contribute to the successful radiation of Y. valettei, increasing the likelihood of successful 305 fertilisation in areas of low species abundance. Simultaneous hermaphroditism observed in 306 Y. valettei may be key to a successful species radiation throughout the Southern Ocean. 307 From an evolutionary perspective, simultaneous hermaphroditism may have enabled fast 308 recolonisation of the Antarctic continental shelf during ice retreat in glacial-interglacial 309 cycles of the Late-Cenozoic. Past glacial maximums are modelled to have grounded at the 310 edges of the continental shelf seen today (Huybrechts 2002), which may have forced many 311 benthic species into the adjacent deep-sea. Evidence however, suggests that ice advance 312 was unlikely to be synchronous around the Southern Ocean, and temporary benthic shelters 313 may have facilitated species survival on the shelf (Thatje et al. 2005, 2008, Convey et al. 314 2009). Simultaneous hermaphroditism, as observed in Y. valettei, would allow for species to 315 rapidly colonise new areas by increasing the chance of both sexes being available for 316 fertilisation, or even the possibility of self-fertilisation. Single specimens could be all that is 317 needed to found new populations in areas of high ice impact, and may have facilitated their 318 likely radiation from deep-sea basins beyond the Polar Front, a hypothesis, which should be 319 tested in future molecular studies. Hermaphroditism in deep-sea protobranchs may be 320 underestimated globally through a lack of life history studies, but may have implications to 321 understanding their ecological success over wide geographic and bathymetric ranges. 14 322 Conclusion 323 Yoldiella ecaudata, Y. sabrina, and Y. valettei all show evidence of lecithotrophic larval 324 development and gonad morphology consistent with what is known for other species of 325 deep-sea protobranch bivalves in the Atlantic Ocean. Large egg size, often described in 326 deep-sea benthic invertebrates, enables the uncoupling from primary production, and could 327 drive 328 hermaphroditism will increase the chances of successful fertilisation in areas of low species 329 density or in colonising areas following ice retreat. Given the contrasting morphology and 330 anatomy of these three Yoldiella species and similarities with different protobranch genera, 331 a thorough review of the protobranch taxa of the Southern Ocean should be completed to 332 understand evolutionary links to other deep-sea basins. 333 Acknowledgements 334 Adam J. Reed was supported through a Natural Environment Research Council PhD 335 studentship. We are grateful to the cruise leaders, captains, officers, and crews of PFS 336 Polarstern (ANDEEP III) and of RRS James Clark Ross (JR144 BIOPEARL I, JR179 BIOPEARL II) 337 who enabled us to collect the samples for this study. We thank Richard Pearce for help with 338 SEM analysis, and Paul Tyler and John Grahame for their feedback on an earlier version of 339 this manuscript. This study is part of the British Antarctic Survey Polar Science for Planet 340 Earth Programme funded by The Natural Environment Research Council (NERC), and is 341 ANDEEP publication #192. continuous or quasi-continuous reproduction. Additionally, simultaneous 342 343 References 15 344 Allen JA (1978) Evolution of the deep-sea protobranch bivalves. Phil Trans R Soc, B 284:387- 345 401 346 Allen JA (2008) Bivalvia of the deep Atlantic. Malacologia 50:57-173 347 Allen JA, Sanders HL (1973) Studies on deep-sea Protobranchia (Bivalvia); the families 348 Siliculidae and Lametilidae. Bull Mus Comp Zool 145:263-310 349 Allen JA, Sanders HL (1982) Studies on deep-sea Protobranchia (Bivalvia); the subfamily 350 Spinulinae (family Nuculanidae). Bull Mus Comp Zool 145:263-310 351 Allen JA, Hannah FJ (1986) A reclassification of the recent genera of the subclass 352 Protobranchia (Mollusca: Bivalvia). J Conch 32:225-249 353 Allen JA, Hannah FJ (1989) Studies on the deep sea Protobranchia (Nuculanidae). Bull Br 354 Mus Nat Hist 55:123-171 355 Allen JA, Sanders HL, Hannah FJ (1995) Studies on the deep-sea Protobranchia (Bivalvia); the 356 subfamily Yoldiellinae. Bull Br Mus Nat Hist 61:11-90 357 Allen JA, Sanders HL (1996) The zoogeography, diversity and origin of the deep-sea 358 protobranch bivalves of the Atlantic: The epilogue. Prog Oceanogr 28:95-153 359 Bosch I, Pearse JS (1988) Seasonal pelagic development and juvenile recruitment of the 360 bivalve Laternula elliptica in McMurdo Sound, Antarctica. Am Zool 28:89-89 361 Bosch I, Pearse JS (1990) Developmental types of shallow-water asteroids of McMurdo 362 Sound, Antarctica. Mar Biol 104:41-46 363 Brandt A, Barthel D (1995) An improved supra- and epibenthic sledge for catching 364 Peracarida (Crustacea, Malacostraca). Ophelia 43:15-23 365 Brente N (2005) An epibenthic sledge for operations on Marine Soft Bottom and Bedrock. 366 Mar Tech Soc J 39:10-21 367 Brey T, Hain S (1992) Growth, reproduction and production of Lissarca notorcadensis 368 (Bivalvia: Philobryidae) in the Weddell Sea, Antarctica. Mar Ecol Prog Ser 82 219-226 16 369 Chase MR, Etter RJ, Rex MA, Quattro JM (1998) Bathymetric patterns of genetic variation in 370 a deep-sea protobranch bivalve Deminucula atacellana. Mar Biol 131:301-308 371 Convey P, Stevens MI, Hodgson DA, Smellie JL, Hillenbrand C-D, Barnes DKA, Clarke A, Pugh 372 PJA, Linse K, Cary SC (2009) Exploring biological constraints on the glacial history of 373 Antarctica. Quat Sci Rev 28:3035-3048 374 Davenport J (1988) The feeding mechanism of Yoldia (=Aequiyoldia) eightsi (Courthouy). 375 Proc R Soc B: 232:431-442 376 Enderlein P, Larter R (2008) JR179/BIOPEARL II Cruise Report. 377 www.bodc.ac.uk/data/information_and_inventories/cruise_inventory/report/8277 378 Etter RJ, Rex MA, Chase MR, Quattro JM (2005) Population differentiation decreases with 379 depth in deep-sea bivalves. Evolution 59:1479-1491 380 Etter RJ, Boyle EE, Glazier A, Jennings RM, Dutra E, Chase MR (2011) Phylogeography of a 381 pan-Atlantic abyssal protobranch bivalve: implications for evolution in the Deep Atlantic. 382 Mol Ecol 20:829-843 383 Fahrbach E (2005) The Expedition ANTARKTIS-XXII/3 of the Research Vessel “Polarstern” in 384 2005. Berichte Polar Meeresforsch 533:1-246 385 Gage JD (1994) Recruitment ecology and age structure of deep-sea invertebrate 386 populations. In: Young CM et al (eds) Reproduction, larval biology and recruitment of the 387 deep-sea benthos. pp. 223-242 388 Gage JD, Tyler PA (1991) Deep-sea biology: A natural history of organisms at the deep-sea 389 floor. Cambridge University Press 390 Glover AG, Smith CR, Mincks SL, Sumida PYG, Thurber AR (2008) Macrofaunal abundance 391 and composition on the West Antarctic Peninsula continental shelf: Evidence for a sediment 392 ‘food bank’ and similarities to deep-sea habitats. Deep-Sea Res Part II 55:2491-2501 17 393 Gofas, S (2012) Yoldiella A. E. Verrill & Bush, 1897. Accessed through: World Register of 394 Marine Species at http://www.marinespecies.org/aphia.php?p=taxdetails&id=138673 on 395 2012-03-07 396 Goodsell JG, Eversole AG (1992) Prodissoconch I and II length in Mercenaria taxa. Nautilus 397 106:119-122 398 Hain S, Arnaud PM (1992) Notes on the reproduction of high-Antarctic molluscs from the 399 Weddell Sea. Polar Biol 12:303-312 400 Higgs ND, Reed AJ, Hooke R, Honey DJ, Heilmayer O, Thatje S (2009) Growth and 401 reproduction in the Antarctic brooding bivalve Adacnarca nitens (Philobryidae) from the 402 Ross Sea. Mar Biol 156:1073-1081 403 Huybrechts P (2002) Sea-level changes at the LGM from ice-dynamic reconstructions of the 404 Greenland and Antarctic ice sheets during the glacial cycles. Quart Sci Rev 21:203-231 405 Killeen IJ, Turner J (2009) Yoldiella and Portlandia (Bivalvia) from the Faroe-Shetland 406 Channel and Rockall Trough, Northeast Atlantic. J Conch 39:733-778 407 Knudsen J (1970) The systematics and biology of abyssal and hadal Bivalvia. Galathea Rep 408 11:1-241 409 Lightfoot RH, Tyler PA, Gage JD (1979) Seasonal reproduction in deep-sea bivalves and 410 brittlestars. Deep-Sea Res 26A:967-973 411 Linse 412 www.bodc.ac.uk/data/information_and_inventories/cruise_inventory/report/jr144-149.pdf 413 Linse K, Cope T, Lörz AN, Sands C (2007) Is the Scotia Sea a centre of Antarctic marine 414 diversification? Some evidence of cryptic speciation in the circum-Antarctic bivalve Lissarca 415 notorcadensis (Arcoidea: Philobryidae). Polar Biol 30:1059-1068 416 McClain CR, Stegan JC, Hurlbert AH (2012) Dispersal, environmental niches and oceanic- 417 scale turnover in deep-sea bivalves. Proc R Soc B 279:1993-2002 K (2006) Cruise Report; JR144, JR145, JR146, JR147 and JR149. 18 418 Mincks SL, Smith CR, De Master DJ (2005) Persistence of labile organic matter and microbial 419 biomass in Antarctic shelf sediments: evidence of a sediment ‘food bank’. Mar Ecol Prog Ser 420 300:3-19 421 Nolan CP, Clarke A (1993) Growth in the bivalve Yoldia eightsi at Signy Island, Antarctica, 422 determined from internal shell increments and calcium-45 incorporation. Mar Biol 117:243- 423 250 424 Ockelmann KW (1965) Developmental types in marine bivalves and their distribution along 425 the Atlantic coast of Europe. In: Peake J (eds) Proceedings of the first European 426 Malacological Congress, London, 1962, pp 25-35 427 Passos FD, Domaneschi O (2009) The anatomical characters related to the brooding 428 behavior of two Antarctic species of Mysella Angas, 1877 (Bivalvia, Galeommatoidea, 429 Lasaeidae), with direct and indirect evidences of ovoviviparity. Polar Biol 32:271-280 430 Peck LS, Bullough LW (1993) Growth and population structure in the infaunal bivalve Yoldia 431 eightsii in relation to iceberg activity at Signy Island, Antarctica. Mar Biol 117:235-241 432 Peck LS, Colman JG, Murray AWA (2000) Growth and tissue mass cycles in the infaunal 433 bivalve Yoldia eightsii at Signy Island, Antarctica. Polar Biol 23:420-428 434 Peck LS, Webb KE, Bailey DM (2004) Extreme sensitivity of biological function to 435 temperature in Antarctic marine species. Funct Ecol 18:625-630 436 Picken GB (1979) Non-pelagic reproduction of some Antarctic prosobranch gastropods from 437 Signy Island, South Orkney Islands. Malacologia 19:109-128 438 Picken GB (1980) Reproductive adaptations of Antarctic benthic invertebrates. Biol J Linn 439 Soc 14:67-75 440 Prezant RS, Showers M, Winstead RL, Cleveland C (1992) Reproductive ecology of the 441 Antarctic bivalve Lissarca notorcadensis (Philobryidae). Am Malacol Bull 9:173-186 19 442 Reed AJ, Morris JP, Linse K, Thatje S (2013a) Plasticity in shell morphology and growth 443 among deep-sea protobranch bivalves of the genus Yoldiella (Yoldiidae) from contrasting 444 Southern Ocean regions. Deep-Sea Res Part I 81:14-24 445 Reed AJ, Thatje S, Linse K (2013b) An unusual hermaphrodite reproductive trait in the 446 Antarctic brooding bivalve Lissarca miliaris from the Scotia Sea, Southern Ocean. Polar Biol 447 36:1-11 448 Richardson MG (1979) The ecology and reproduction of the brooding Antarctic bivalve 449 Lissarca miliaris. Br Antarct Surv Bull 49:91-115 450 Sanders HL, Hessler RR, Hampson GR (1965) An introduction to the study of deep-sea 451 benthic faunal assemblages along the Gay Head-Bermuda transect. Deep-Sea Res 12:845- 452 867 453 Sartori AF, Domaneschi O (2005) The functional morphology of the Antarctic bivalve Thracia 454 meridionalis Smith, 1885 (Anomalodesmata: Thraciidae). J Molluscan Stud 71:199-210 455 Scheltema RS, Williams IP (2009) Reproduction among protobranch bivalves of the family 456 Nuculidae from sublittoral, bathyal, and abyssal depths off the New England coast of North 457 America. Deep-Sea Res Part II: 56:1835-1846 458 Sharma PP, Zardus JD, Boyle EE, González VL, Jennings RM, McIntyre E, Wheeler WC, Etter 459 RJ, Giribet G (2013) Into the deep: A phylogenetic approach to the bivalve subclass 460 Protobranchia. Mol Phylogenet Evol 69: 188-204 461 Shilling FM, Manahan DT (1994) Energy metabolism and amino acid transport during early 462 development of Antarctic and temperate echinoderms. Biol Bull 187:398-407 463 Thatje S (2012) Effects of capability for dispersal on the evolution of diversity in Antarctic 464 benthos. Integr Comp Biol 52:470-482 465 Thatje S, Hillenbrand CD, Larter R (2005) On the origin of Antarctic marine benthic 466 community structure. Trends Ecol Evol 20:534-540 20 467 Thatje S, Hillenbrand CD, Mackensen A, Larter R (2008) Life hung by a thread: endurance of 468 Antarctic fauna in glacial periods. Ecology 89:682-692 469 Turekian KK, Cochran JK, Kharkar DP, Cerrato RM, Vaišnys JR, Sanders HL, Grassle JF, Allen JA 470 (1975) Slow growth rate of a deep-sea clam determined by 228Ra chronology. Proc Natl Acad 471 Sci 72:2829-2832 472 Tyler PA, Harvey R, Giles LA, Gage JD (1992) Reproductive strategies and diet in deep-sea 473 nuculanid protobranchs (Bivalvia: Nucloidea) from the Rockall Trough. Mar Biol 114:571-580 474 Verrill AE, Bush KJ (1898) Revision of the deep water Mollusca of the Atlantic Coast of North 475 America with descriptions of new genera and species. Part 1 - Bivalvia. Proc US Nat Mus 476 20:775-901 477 Zardus JD (2002) Protobranch Bivalves. Adv Mar Biol 42:1-65 478 Zardus JD, Morse MP (1998) Embryogenesis, morphology and ultrastructure of the 479 pericalymma larva of Acila castrensis (Bivalvia: Protobranchia: Nuculoida). Invert Biol 480 117:221-244 481 Zardus JD, Martel AL (2001) Phylum Mollusca: Bivalvia. In: Young CM, Sewell MA, Rice ME 482 (eds) Atlas of Marine Invertebrate Larvae. Academic Press, London. pp. 289-325 483 Zardus JD, Etter RJ, Chase MR, Rex MA, Boyle EE (2006) Bathymetric and geographic 484 populations in the pan-Atlantic deep-sea bivalve Deminucula atacellana. Mol Ecol 15:639- 485 651 486 487 488 489 490 491 21 492 Figure Legends 493 Fig. 1 Photomicrographs of decalcified Yoldiella sabrina before histology sectioning (shell 494 length 7.75mm) from 200m in the Scotia Sea. a) view of the left sowing large digestive 495 gland; b) view of the right showing digestive gland around hind gut loop. m – muscle; ct- 496 ctenidia; dg – digestive gland; hg - hind gut. Scale bars = 1mm. 497 Fig. 2 Transverse histology sections of Yoldiella sabrina. a) 9.1mm shell length from 3500m 498 Scotia Sea showing general anatomy; b) left side digestive gland and oocytes from 5.5mm 499 shell length from 200m Scotia Sea; c) left side digestive gland and testis from 9.7mm shell 500 length from 3500m in the Scotia Sea. oc – oocyte; dg – digestive gland; hg – hind gut; st – 501 stomach; t – testis. Scale bars 100µm unless stated otherwise. 502 Fig. 3 Oocyte diameter measured from histology sections against shell length. a) Yoldiella 503 sabrina collected in February from the South Atlantic (SA) and Scotia Sea (SS); b) Yoldiella 504 ecaudata collected in March 2008 from the 500m in the Amundsen Sea. N.B. two specimens 505 at 2.6mm shell length offset for clarity; c) Yoldiella valettei collected in February from 506 1000m in the Scotia Sea. 507 Fig. 4 Photomicrograph and transverse histology section of Yoldiella ecaudata from 500m in 508 the Amundsen Sea. a) photomicrograph of 2.8mm shell length specimen showing digestive 509 gland and hind gut loop; b) histology of 2.6mm shell length female showing vitellogenic 510 oocytes in the dorsal portion of the gonad above the stomach and surrounding digestive 511 gland. m – muscle; dg – digestive gland; hg – hind gut; st – stomach; oc – oocyte. 512 Fig. 5 Transverse histology sections of Yoldiella valettei from 1000m in the Scotia Sea. a) 513 section of entire 3.1mm hermaphrodite showing oocytes and testis development in relation 22 514 to digestive gland; b) section of a 3.1mm shell length male showing testis enveloping 515 digestive gland and hind gut loop, extending dorsally above the stomach. dg – digestive 516 gland; hg – hind gut; oc – oocyte; t – testis; st – stomach. Scale bars = 100µm. 517 Fig. 6 Prodissoconch I (PI) images taken using scanning electron microscopy. Arrow indicates 518 artefact on shell. a) Yoldiella valettei; b) Yoldiella ecaudata. Scale bars =100µm. 519 23 520 521 Table 1 Species collection and reproductive data for Yoldiella ecaudata, Yoldiella sabrina, 522 and Yoldiella valettei from the Southern Ocean. ‘n’ indicates total number of animals used 523 to obtain egg sizes and sex ratios, including those of unknown sex. Species Location & Date Cruise& Gear Yoldiella ecaudata Yoldiella sabrina Yoldiella sabrina Yoldiella sabrina Yoldiella sabrina Yoldiella valettei Amundsen Sea March 2008 Amundsen Sea March 2008 Scotia Sea February 2005 Scotia Sea February 2005 South Atlantic February 2005 Scotia Sea February 2006 BIOPEARL II EBS BIOPEARL II EBS ANDEEP III AGT ANDEEP III AGT ANDEEP III EBS BIOPEARL I EBS Depth (m) male/female Max. oocyte size (µm) n 500 3/6 130.4 15 500 3/1 / 10 200500 6/5 55.68 17 3500 2/1 187.9 5 4730 3/1 132.0 5 1000 9/12 120.6 34 524 525 24 526 Fig. 1 Photomicrographs of decalcified Yoldiella sabrina before histology sectioning (shell 527 length 7.75mm) from 200m in the Scotia Sea. a) view of the left sowing large digestive 528 gland; b) view of the right showing digestive gland around hind gut loop. m – muscle; ct- 529 ctenidia; dg – digestive gland; hg - hind gut. Scale bars = 1mm. 530 531 25 532 Fig. 2 Transverse histology sections of Yoldiella sabrina. a) 9.1mm shell length from 3500m 533 Scotia Sea showing general anatomy; b) left side digestive gland and oocytes from 5.5mm 534 shell length from 200m Scotia Sea; c) left side digestive gland and testis from 9.7mm shell 535 length from 3500m in the Scotia Sea. oc – oocyte; dg – digestive gland; hg – hind gut; st – 536 stomach; t – testis. Scale bars 100µm unless stated otherwise. 537 26 538 Fig. 3 Oocyte diameter measured from histology sections against shell length. a) Yoldiella 539 sabrina collected in February from the South Atlantic (SA) and Scotia Sea (SS); b) Yoldiella 540 ecaudata collected in March 2008 from the 500m in the Amundsen Sea. N.B. two specimens 541 at 2.6mm shell length offset for clarity; c) Yoldiella valettei collected in February from 542 1000m in the Scotia Sea. 543 27 544 Fig. 4 Photomicrograph and transverse histology section of Yoldiella ecaudata from 500m in 545 the Amundsen Sea. a) photomicrograph of 2.8mm shell length specimen showing digestive 546 gland and hind gut loop; b) histology of 2.6mm shell length female showing vitellogenic 547 oocytes in the dorsal portion of the gonad above the stomach and surrounding digestive 548 gland. m – muscle; dg – digestive gland; hg – hind gut; st – stomach; oc – oocyte. 549 28 550 Fig. 5 Transverse histology sections of Yoldiella valettei from 1000m in the Scotia Sea. a) 551 section of entire 3.1mm hermaphrodite showing oocytes and testis development in relation 552 to digestive gland; b) section of a 3.1mm shell length male showing testis enveloping 553 digestive gland and hind gut loop, extending dorsally above the stomach. dg – digestive 554 gland; hg – hind gut; oc – oocyte; t – testis; st – stomach. Scale bars = 100µm. 555 29 556 Fig. 6 Prodissoconch I (PI) images taken using scanning electron microscopy. Arrow indicates 557 artefact on shell. a) Yoldiella valettei; b) Yoldiella ecaudata. Scale bars =100µm. 558 30