THESIS on CSUMB

REPRODUCTIVE STRATEGIES OF THE DOMINANT GASTROPODS

OF THE LAU BASIN HYDROTHERMAL VENT SYSTEM:

ALVINICONCHA HESSLERI AND IFREMERIA NAUTILEI

______

A Thesis

Presented to the

Faculty of the

Moss Landing Marine Laboratories and

California State University Monterey Bay

______

In Partial Fulfillment

of the Requirements for the Degree

Master of Science

Marine Science

______

by Kyle C. Reynolds

Fall 2009

CALIFORNIA STATE UNIVERSITY MONTEREY BAY

The Undersigned Faculty Committee Approves the

Thesis of Kyle C. Reynolds:

REPRODUCTIVE STRATEGIES OF THE DOMINANT GASTROPODS OF THE LAU BASIN HYDROTHERMAL VENT SYSTEM: ALVINICONCHA HESSLERI AND IFREMERIA NAUTILEI

______Nick Welschmeyer, Chair Moss Landing Marine Laboratories

______Michael Graham Moss Landing Marine Laboratories

______Stacy Kim Moss Landing Marine Laboratories

______Approval Date

iii

Kyle C. Reynolds

DEDICATION

I wish to dedicate this thesis to Dr. Ellen Strong, who shall forever after be known as my Fairy Gonadmother.

EPIGRAPH AND AUTHOR CONTRIBUTIONS

The following thesis is comprised of two separate manuscripts, currently in preparation for submission. The first manuscript, which serves as Chapter One, is formatted for submission to the journal Nature. This manuscript reports and puts into context three novel characteristics discovered during the course of this project: 1) a unique embryo transport mechanism, which I discovered and described; 2) a brood pouch within a modified metapodial pedal gland, discovered by Dr. Anders Warén and described by me (taxonomic expertise for this description was provided by Dr. Ellen Young); and 3) a completely new larval form, discovered by me, but the significance and later developmental stages of which were uncovered by Dr. Hiromi Watanabe and Dr. Craig Young, to different degrees.

The second manuscript, Chapter Two of this thesis, compares and describes the reproductive structures and strategies of Ifremeria nautilei and Alviniconcha hessleri at both an organismal and cellular level for the journal Deep-Sea Research. The purpose of this manuscript was to fill in gaps left in previous studies with regard to the reproductive anatomies of both species, and to provide evidence of the significance and derivations of the structures described, with reference to their respective phylogenies.

This project is a culmination of three long years spent personally transporting these samples to experts in various fields across the country, seeking help in investigating the reproductive aspects of these mysterious animals; the likes of which did not exist in textbooks or scientific literature. As first author on these manuscripts, I had the responsibility of the majority of the specimen collections, lab work, literature review, and writing, although significant contributions were made by all co-authors in their respective areas of expertise. The Japanese team of scientists appearing as co-authors on Chapter 1 began work on Ifremeria nautilei larvae just recently, not knowing that my research had been ongoing for three years prior. Their success with rearing the larvae to the veliger stage was a significant contribution and provided information that I had been unable to obtain. We came to the conclusion as a group that our research was too complementary not to combine and authorship order was agreed upon according to contribution of research.

ABSTRACT

Reproductive strategies of the dominant gastropods of the Lau Basin hydrothermal vent system: Alviniconcha hessleri and Ifremeria nautilei

by Kyle C. Reynolds Master of Science in Marine Science Moss Landing Marine Laboratories, California State University Monterey Bay, 2009

Reproductive biology and larval development remain elusive processes to researchers for many vent endemic species due to the cost prohibitive nature of sampling in these environments, as well as the difficulties inherent to laboratory culturing of chemosynthetic organisms. Thus, many of these biological processes and strategies have only been inferred from related species living in shallow marine environments, resulting in a paradigm that broadly attributes phylogenetic constraint to any life-history variation found at vents. Alviniconcha hessleri and Ifremeria nautilei comprise the majority of the dominant megafauna found in the Lau Basin hydrothermal vent system. While they share a number of unique anatomical modifications, overlapping distributions, and a recent common ancestry, they employ disparate reproductive strategies. A planktotrophic mode of larval development has been inferred for A. hessleri from its shell morphology, while I. nautilei has been found to protect its young throughout early development in a brood pouch within its foot. Previous studies of these species involved sexually immature specimens, leaving the most pertinent questions unanswered regarding their reproductive biology. In this thesis, I have examined the reproductive anatomy of both species at both organismal and cellular levels. Evidence of iteroparity and a simultaneously periodic reproductive effort was revealed for both species across vent sites. In addition, the following apomorphic characters were discovered in I. nautilei: a novel brood pouch; a unique embryo transport mechanism; and a new larval form, which we have named Warén’s larva. These findings provide some of the first substantial evidence of evolution of developmental traits in a hydrothermal vent organism.

TABLE OF CONTENTS

PAGE

DEDICATION ...... iv EPIGRAPH AND AUTHOR CONTRIBUTIONS...... iv ABSTRACT ...... v LIST OF FIGURES ...... viii ACKNOWLEDGEMENTS ...... ix CHAPTER 1 INTERNAL BROODING AND A NOVEL LARVAL FORM IN A DEEP-SEA HYDROTHERMAL VENT GASTROPOD ...... 1 Methods Summary / Methods...... 6 References...... 8 Acknowledgements ...... 10 Author Information...... 11 Figures and Captions ...... 11 2 A COMPARATIVE DESCRIPTION OF REPRODUCTIVE STRUCTURES IN IFREMERIA NAUTILEI AND ALVINICONCHA HESSLERI (CAENOGASTROPODA: PROVANNIDAE): HYDROTHERMAL VENT GASTROPODS FROM THE LAU BASIN...... 14 Introduction...... 15 Material and Methods...... 17 Expedition and Collection Data ...... 17 Laboratory Protocol and Sample Treatment...... 17 Results...... 18 Ifremeria nautilei...... 18 Reproductive stage ...... 18 Foot and brood pouch morphology...... 19 Ontogeny ...... 19 Pallial oviduct...... 21

vii

Alviniconcha hessleri ...... 21 Reproductive stage ...... 21 Foot morphology ...... 22 Ontogeny ...... 22 Pallial oviduct...... 22 Discussion...... 23 Species Comparisons...... 23 Reproductive stage ...... 23 Foot morphology ...... 23 Ontogeny ...... 23 Pallial oviduct...... 24 Brooding in Caenogastropods ...... 25 Conclusion ...... 26 Acknowledgements ...... 26 References...... 27 Figures and Captions ...... 28

viii

LIST OF FIGURES CHAPTER ONE PAGE

Figure 1. Brood pouch of I. nautilei...... 11 Figure 2. Development of Warén's Larva...... 12

CHAPTER TWO PAGE

Figure 1. Bathymetric map of the Eastern Lau Spreading Center...... 28 Figure 2 (a-d). In situ photos of Lau Basin mollusks ...... 29 Figure 3. Representative illustration of external morphology for Ifremeria and Alviniconcha ...... 30 Figure 4 (a,b). Sagittal section of head-foot of sexually immature Ifremeria with empty brood pouch, showing morphology of septae...... 30 Figure 5 (a,b). Ifremeria brood pouch location and morphology...... 31 Figure 6. Histological section showing outer brood pouch wall merging continuously with inner septum...... 32 Figure 7. SEM image of embryo transport packet (ETP)...... 32 Figure 8 (a-d). SEM images of progressive developmental stages of Warén’s Larva from different specimens of female Ifremeria...... 33 Figure 9. Histological cross-section of the proximal region of the pallial oviduct complex of Ifremeria...... 34 Figure 10 (a,b). Sagittal sections showing foot morphology of Ifremeria and Alviniconcha to compare metapodial pedal gland locations...... 35 Figure 11. Histological cross-section of the proximal region of the pallial oviduct complex of Alviniconcha...... 36

ACKNOWLEDGEMENTS

I would like to take this opportunity to expand on the acknowledgements provided in the following manuscripts, and give credit to everyone who helped make this thesis possible:

To my advisor, Stacy Kim, who made this fantastic journey of discovery possible.

To Ellen Strong of the Smithsonian National Museum of Natural History, who became a close friend during the course of this project - I can never repay you for the sheer dedication that you put into this from the very moment of your first response to my random email. You breathed new life into this research and taught me so much in such a short period of time.

A sincere debt of gratitude is also owed to Craig Young of the Oregon Institute of Marine Biology for taking me under his wing, offering lab space, histology and SEM equipment, editing support, and doing me the honor of becoming co-author on my publications.

I would like to thank Anders Warén of the Swedish Museum of Natural History for encouraging us to follow up with his initial discovery of Ifremeria’s brood pouch and for his expert opinions that helped me begin to decipher the mystery of these bizarre creatures.

Thank you as well to Paul Tyler of the National Oceanography Centre in Southampton, UK for the generous use of his lab, his equipment, and for training me in the art and science of histology.

To Kamille Hammerstrom and Gabriela Vega, I give thanks for their selfless assistance with lab work, but even more for their much needed encouragement, support, and friendship. And to Marilyn Schotte and Paul Greenhall, whom I barely had the chance to meet, but who took such great care of my samples and worked tirelessly to provide many crucial last-minute slides and SEM images.

I thank my committee members, Nick Welschmeyer and Mike Graham for ideas, advice, and editing; the crew of the ROV Jason and of the R/V Melville for their support with collections; San Jose State University, Oregon Institute of Marine Biology, and the Smithsonian Institute’s National Museum of Natural History for the use of their labs, equipment, and supplies; and finally to the scholarships and funding sources, without which this work would not have been possible: ChEss and Fondation Total Fellowship; Dr Earl H. Myers and Ethel M. Myers Oceanographic and Marine Biology Trust; David & Lucile Packard Student Research & Travel Awards; and the National Science Foundation.

CHAPTER 1

Internal brooding and a novel larval form in a deep-sea hydrothermal vent gastropod Kyle C. Reynolds1, Hiromi Watanabe2, Ellen Strong3, Takenori Sasaki4, Katsuyuki Uematsu5,

Hiroshi Miyake6, Shigeaki Kojima7, Yohey Suzuki8, Katsunori Fujikura2, Stacy Kim1, Craig M.

Young9

1 Moss Landing Marine Laboratories, 8272 Moss Landing Road, Moss Landing, CA 95039, USA.

2 Japan Agency for Marine-Earth Science and Technology, 2-15 Natsushima, Yokosuka 237-

0061 Japan.

3 Smithsonian Institution, National Museum of Natural History, P.O. Box 37012, MRC 163,

Washington, D.C. 20013-7012, U.S.A.

4 The University Museum, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-0033,

Japan

5 Marine Works Japan LTD., 2-15 Natsushima, Yokosuka 237-0061 Japan

6 School of Marine Biosciences, Kitasato University, 160-4 Utou, Okkirai, Sanriku, Ofunato,

Iwate 022-0101, Japan

7 Ocean Research Institute, The University of Tokyo, 1-15-1 Minamidai, Nakano, Tokyo 164-

8639, Japan.

8 Advanced Industrial Science and Technology, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568,

Japan

9 Oregon Institute of Marine Biology, University of Oregon, P.O. Box 5389, Charleston, OR

97420, U.S.A.

Paradoxically, organisms physiologically adapted to extreme thermal and chemical conditions at deep-sea hydrothermal vents seldom display life history modifications unique to this unforgiving environment. Although internal fertilization is more common at vents than in shallow water, developmental modes and larval forms are typically similar in vent animals and their shallow-water relatives1,2,3,4, suggesting that phylogenetic constraint has prevented vent animals from developing life-history traits unique to their extreme habitat5,6. Here we report the first documented exception to this general rule. Ifremeria nautilei, a hydrothermal vent snail from 1700-2900 m depth in the western Pacific, displays two novel life-history traits: 1) internal brood protection in a modified metapodial pedal gland, and, 2) an unusual uniformly ciliated larva that we here name Warén’s larva. This larva, which has a unique external cuticle, is the only known example of a free-swimming pre-veliger larval stage in the higher gastropods and is the first new gastropod larval form to be described in more than 100 years. Warén’s larva emerges from the internal brood pouch as a fully ciliated lecithotrophic larva and swims with its posterior end forward.

After 15 days at room temperature, Warén’s larva metamorphoses into a typical veliger larva. We hypothesize that both the larval cuticle and the internal brood pouch protect sensitive early stages.

The gastropod family Provannidae has traditionally included five genera (Provanna,

Alviniconcha, Cordesia, Debruyeresia, Ifremeria) and is known from hydrothermal vents, hydrocarbon seeps, and large detrital falls (wood, animal carcasses) in the deep sea. Sperm ultrastructure7,8 and molecular data9 reveal affinities with intertidal periwinkles and their relatives

(Littorinoidea) in the Caenogastropoda, the largest of five main gastropod clades, that contains

~60% of living species. A large multi-gene molecular data set has revealed that the deep sea

3 gastropod genus Abyssochrysos (Abyssochrysidae) is nested within the Provannidae10.

Consequently, until a comprehensive revision is completed and family-level relationships refined, we refer to the entire assemblage as the Abyssochrysoidea.

While most abyssochrysoids are small (<20 mm in adult length) grazers on bacteria and detritus, members of Alviniconcha and Ifremeria host chemoautotrophic sulphur oxidizing symbiotic bacteria in their gills that contribute most of the metabolic requirements, allowing them to reach large adult sizes (>85 mm in size)11. Both planktotrophic (feeding) and non- planktotrophic larval development occur among abyssochrysoids, but the spawn and early development are known for only one species, Provanna lomana, which exhibits encapsulated development with adelphophagy (sibling cannibalism). It is not known if the young emerge as veliger larvae or crawling juveniles, but the simple, paucispiral protoconch (larval shell) is indicative of non-planktotrophic development12,13. In Ifremeria, Alviniconcha, Cordesia,

Debruyersia and some fossil members of Provanna, the apical whorls of the protoconch

(protoconch I) are lacking (i.e. decollate) and the opening is sealed with a calcareous plug. In these forms, protoconch II comprises 1.5-2 whorls and is ornamented with numerous orthocline riblets and fine spiral threads13,14. The size difference between the eggs and the decollate larval shells is too large to be explained by lecithotrophy13, and the presence of elaborate, reinforcing ornamentation is also taken to indicate planktotrophic development15,16.

Ifremeria nautilei is unique among abyssochrysoids in possessing a brood pouch in the foot13 (Fig. 1a–d). Small, ovoid, yolky eggs, ~40-45 µm in length are transported from the ovary to the pallial oviduct (Fig. 1e), which lacks the albumen, capsule and jelly glands responsible for producing and provisioning the egg capsule in oviparous and some ovoviviparous species.

Secretions from an unspecialized basophilic epithelium entangle strings of fertilized eggs (Fig.

1g) that are conveyed through a prominent furrow from the mantle cavity to a lateral groove along the sole of the foot. The egg strings enter a central ciliated pore (Fig. 1c–d) on the sole, probably via a deep furrow formed by the foot, as in other gastropods that use the foot to manipulate their spawn. The pore opens to a large brood pouch, roughly circular in cross section, containing thousands of developing embryos. The pouch is surrounded by several layers of circular muscles and is lined with a simple, thin, primarily unciliated, squamous to cuboidal epithelium. A variable number of branching septae (Fig. 1b) radiate from the pore towards the rear and side walls, subdividing the interior into a series of chambers. Muscular tissue from the surrounding layers penetrates the septae. A thin, darkly staining non-cellular substance of unknown composition, possibly cuticular, lines the interior surfaces of the brood pouch (Fig. 1f).

The metapodial pedal gland present in all abyssochrysoids is absent in Ifremeria nautilei and we hypothesize that it has been co-opted for brooding13. Brood pouches within the head-foot have evolved independently several times within the unrelated caenogastropod superfamily

Cerithioidea. However, these penetrate the cephalic hemocoel from a pore on the side of the neck or the foot through modification of an external egg groove. A single non-homologous instance of brooding in a modified anterior pedal gland is known in members of the West African genus

Cymbium (Volutidae)17. The pouch of Ifremeria is the only known instance of brooding in a metapodial pedal gland and no other vent gastropod is known to brood.

Although the duration of incubation is not known, all embryos in the brood pouch of a single individual are at comparable stages of development. Embryos arrive in the pouch soon after fertilization, followed by formation of a stereoblastula. The mode of gastrulation is not known. The first larval stage differs from the classic gastropod trochophore in lacking a girdle of multiciliated trochoblast cells; instead, it is fully ciliated, with simple cilia extending through

5 pores of an external cuticular layer (Fig. 2a–d, f). This resembles the pelagosphaera larvae of sipunculans18 and some polychaetes19, which retain the egg envelope as larvae. Larval cuticles are previously unknown in gastropod larvae, although the pericalymma larvae of protobranch bivalves and neomenioid aplacophorans have cellular tests. Compound cilia on the posterior end are distinctly longer and broader than the other cilia (Fig. 2b,e). At the time of larval release or soon thereafter, two small anterior lobes of slightly different sizes become evident (Fig. 2g).

Larvae were released from the brood pouch at this stage under laboratory conditions on multiple occasions.

Larvae are neutrally buoyant and swim with the compound cilia of the posterior end facing forward. They exhibit no phototactic behaviour when exposed to unidirectional white light.

Although the larva has a mid-ventral stomodeum (Fig. 2 h,j,k), the gut is blind and the larva is packed with yolk granules (Fig. 2i), suggesting that it does not feed.

Approximately 15 days after their release from the brood pouch, some of the larvae reared at room temperature metamorphosed into shelled veligers (Fig. 2l,m); larvae reared at 4˚C developed more slowly and died before metamorphosis. The velum, which develops on the bi- lobed anterior end, possesses simple and compound cilia, but details of shell ornamentation were not yet evident. Settlement and metamorphosis into benthic juveniles were not observed.

The newly released stage is a previously unknown larval type that we name “Warén’s larva” in honor of Anders Warén (Swedish Museum of Natural History), a prolific gastropod systematist and specialist of deep sea gastropods who first noted the brood pouch. The trochophore larva20, veliger larva21 and echinospira larva22 of benthic gastropods and the polytrochous larva23 of pelagic pteropods have all been known since the nineteenth century.

Warén’s larva is the first new gastropod larval form to be described in more than 100 years. No

6 other caenogastropod known possesses a free-swimming pre-veliger larval stage, and no other gastropod larva swims with its posterior end forward. Indeed non-planktonic early development is touted as one of the significant innovations contributing to the evolutionary success of caenogastropods24. However, the significance of releasing motile pre-veliger larvae is uncertain as the distribution of this life history trait is unknown; presence of the decollate larval shell among planktotrophic members of the family may indicate this larval form is widespread within the group and may not be strictly associated with vent habitats. Yet development of a novel mode of internal brooding in conjunction with planktonic early development is remarkable and challenges the notion that the evolution of life-history traits in the deep sea is phylogenetically constrained.

Methods Summary

Specimens were collected from the Lau Back-Arc Basin, Manus Basin, and North Fiji

Basin between 1700 and 2900 m depth using push corers and suction samplers manipulated by

ROVs and manned submersible. Warén’s larvae released onboard were reared in unpressurized filtered seawater at 4ºC and room temperature (about 30 ºC), while a portion were fixed in 4 % paraformaldehyde, 2.5% glutaraldehyde or 80% ETOH. Reproductive tissue and brood pouches of adults were embedded in paraplast, serially sectioned at 7 µm thickness, and stained in hematoxylin and eosin-phloxine for histology. Larvae were observed via light microscopy, confocal microscopy, and scanning electron microscopy using standard methods.

Methods

Expedition and collection data are as follows: YK06-13 05 Sep.-05 Oct. 2006, DSRV

Shinkai6500, PI Y. Suzuki, (n=2). MGLN07MV 06 Sep.-29 Sep. 2006, ROV Jason II dives

230-240, PI C. Fisher, (n=150); TUIM07MV 11 Jun.-26 Jun. 2005, ROV Jason II dives 154-164,

PI C. Fisher (n=1); TUIM05MV 06 Apr.-07 May 2005, ROV Jason II dives 124-139, PI M.

Tivey (n=1); STARMER II Jul. 1989, (n=1); BIOLAU Jul. 1989 (n=3). The JAMSTEC expedition took place in the Manus Basin and the STARMER II expedition was in the North Fiji

Basin. All other expeditions took place in the Eastern Lau Spreading Center (ELSC) of the Lau

Back-arc Basin and included the vent sites Kilo Moana, Abe, Tow Cam, and Tu’i Malila between approximately 1900-2700 m depth .

Pallial oviducts and brood pouches were removed, dehydrated in 100% ethanol, embedded in paraplast, sectioned at 7 µm, and stained with hematoxylin and eosin-phloxine. Thick sections

(8 µm) were examined using a Leica MZ microscope and photographed with a Leica DFC 320 digital camera; measurements were made using Pax-It Software with measurement module.

Warén’s larvae released on the YK06-13 cruise were fixed for sectioning in 4 % paraformaldehyde, washed in filtered (0.22 µm pore size) seawater and post fixed in 2% OsO4 for 2 hours at 4ºC. After rinsing 6 times with distilled water for 10 min at 4ºC, the larvae were stained with 1% aqueous uranyl acetate for 2 hours at 4ºC. They were dehydrated through a graded ethanol series and embedded in EPON 812 resin (TAAB). Semi-thin (0.5 µm) sections were cut with a Leica ULTRACUT S microtome and stained with 1% toluidine blue at 70ºC on glass slides. The sections were observed with a BX51 optical microscope equipped with an

Olympus DP20 digital camera system at JAMSTEC. Warén’s larvae preserved in 80% ETOH were transferred to phosphate buffered saline, stained for confocal microscopy with propidium iodide in phosphate buffered saline with 0.1% Tris, mounted in Vectashield, and visualized on an

Olympus Fluoview 1000 confocal microscope. These same specimens were observed with an

Olympus BX-50 microscope using bright-field and DIC optics and photographed with an

Optronics Microfire digital camera.

Warén’s larvae released on the MGLN07MV cruise were preserved in 80% ETOH, dried using hexamethyldisilazane25 and examined using a Philips XL-30 ESEM at USNM. Additional specimens were critical point dried in ETOH and visualized with a Tescan SEM at OIMB.

Warén’s larvae and veliger larvae obtained on the YK06-13 cruise were fixed with 2.5% glutaraldehyde in filtered natural seawater and washed in filtered seawater, then postfixed with

2% osmium tetroxide in filtered seawater for 2 hours at 4° C. After samples were rinsed with

DW, conductive staining was performed by incubating 0.2% aqueous tannic acid (pH 6.8) for 30 min, washing with distilled water and then treating with 1% aqueous osmium tetra oxide for 30 min. Samples were dehydrated via a graded ethanol series, critical point dried (JCPD-5; JEOL), coated with an osmium plasma coater (POC-3; Meiwa Shoji Co., Osaka, Japan) and observed with a FE-SEM (JSM-6700F; JEOL) at an acceleration voltage of 5 kV.

References

1. Bouchet, P. & Warén, A.. Ontogenetic migration and dispersal of deep-sea gastropod larvae.

In: Young, C.M. & Eckelbarger, K.J. (eds) Reproduction, Larval Biology, and Recruitment of the

Deep Sea Benthos. Columbia University Press, New York, pp 98-118 (1994).

2. Eckelbarger, K.J. & Watling, L. Role of phylogenetic constraints in determining reproductive patterns in deep-sea invertebrates. Invertebr. Biol. 114, 256-269 (1995).

3. Tyler, P. & Young, C.M. Reproduction and dispersal at vents and cold seeps. J. Mar. Biol.

Ass. U.K. 79, 193-208 (1999).

4. Young, C.M. Reproduction, development and life history traits. In: Tyler, P.A. (ed.), Ecosystems of the World, Volume 28: Ecosystems of the Deep Oceans. pp. 381-426 (2003).

5. Lutz, R.A., Jablonski, D., Turner, R.D. Larval development and dispersal at deep-sea hydrothermal vents. Science 226 (4681),1451 (1984).

6. Gustafson, R.G., Littlewood, DTJ., Lutz, RA Gastropod Egg Capsules and Their Contents

From Deep-Sea Hydrothermal Vent Environments. Biol. Bull. 180(1),34-55 (1991).

7. Healy, J.M. Taxonomic affinities of the deep-sea genus Provanna: new evidence from sperm structure. J Mollusc Stud. 56, 119-122 (1990).

8. Healy, J.M. Dimorphic spermatozoa of the hydrothermal vent prosobranch Alviniconcha hessleri: systematic importance and comparison with other caenogastropods. Bull Mus Nat His

Nat (Paris) 4 sér A 14, 273–291 (1992).

9. Colgan, D.J., Ponder, W.F., Beacham, E., Macaranas, J. Molecular phylogenetics of

Caenogastropoda (Gastropoda: Mollusca). Mol Phylogenet Evol. 42, 717–737 (2007).

10. Johnson SB, Warén A, Lee R, Kaim A, Davis A, Kano Y, Strong EE, and Vrijenhoek RC. Living Fossils: Rubyspira, a New Genus of Bone-Eating Snails from the Deep-Sea. In review. 11. Henry, M.S., Childress, J.J., Figueroa, D. Metabolic rates and thermal tolerances of chemoautotrophic symbioses from Lau Basin hydrothermal vents and their implications for species distributions. Deep Sea Rsch Part I: Oceanographic Research Papers 55, 679-695 (2008). 12. Warén A, Ponder, WF. 1991. New species, anatomy, and systematic position of the hydrothermal vent and hydrocarbon seep gastropod family Provannidae fam. n. (Caenogastropoda). Zoologica Scripta, 20: 27-56. 13. Kaim, A., Jenkins, R.G., Warén, A., Provannid and provannid-like gastropods from the late Cretaceous cold seeps of Hokkaido (Japan) and the fossil record of the Provannidae (Gastropoda, Abyssochrysoidea). Zool J Linnean Soc 154, 421–436 (2008). 14. Warén, A., Bouchet, P. New gastropods from deep-sea hydrocarbon seeps off West Africa,

Deep–Sea Rsch II, doi:10.1016/j.dsr2.2009.04.013 (2009).

15. Bouchet, P. & Warén, A. Planktotrophic larval development in deep-water gastropods. Sarsia

64, 37-40 (1979).

16. Jablonski, D., Lutz, R.A. Molluscan larval shell morphology. Ecological and paleontological applications. In: Rhoads, D., Lutz, R., eds. Skeletal Growth of Aquatic Organisms, New York: Plenum. pp. 323-377 (1980). 17. Marche-Marchad, I. Un nouveau mode de developpement intracapsulaire chez les

Mollusques proso-branches neogastropodes: l'incubation intrapedieuse des Cymba (Volutidae). C

R Hebd Seances Acad Sci 266D, 706-709 (1968).

18. Rice, M. A comparative study of the development of Phascolosoma agassizii, Golfingia pugettensis, and Themiste pyroides with a discussion of developmental patterns in the sipuncula.

Ophelia 4, 143-171 (1967).

19. Eckelbarger, K.J., Chia, F.S. Morphogenesis of larval cuticle in the polychaete

Phragmatopoma lapidosa. Cell Tiss. Res. 186, 187-201 (1978).

20. Patten, W. The embryology of Patella. Arb. Zool. Inst. Wien, 6, 149-174 (1886).

21. Lankester, E.R. Contributions to the developmental history of the Mollusca. Phil. Trans. R.

Soc. Lond. 165 (1875)

22. Krohn, A. Ub. Einen neuen mit Wimpersegeln versehenen Gasteropoden. Archiv f.

Naturgesch., Jahrg. Xix, Bd. Ii., pp. 223-226 (1853)

23. Gegenbauer, C. Untersuchungen über Pteropoden und Heteropoden. Ein beitrag zur Anatomie und Entwicklungsgeschichte dieser Thiere. Engelmann, Leipzig. (1855). 24. Ponder, W. F., et al. Caenogastropoda. In: Ponder, W.F. & Lindberg, D.L. (eds.) Molluscan

Phylogeny. U. California Press, 331-383 (2008).

Acknowledgements We are grateful to Anders Warén for bringing the presence of the brood pouch to our attention, for providing SEM images of some of the larvae, and for loaning samples from his own collection. We thank Paul

Tyler for histological expertise and help with sample processing, OIMB and the Smithsonian Institution for use of

SEM, Paul Greenhall, Marilyn Schotte, Gabriela Vega, Kamille Hammerstrom and Maya Wolf for assistance with handling specimens. Special thanks to funding sources: National Science Foundation (Grants # OCE-0241250 and

OCE-0527139), JSPS Research Fellowships for Young Scientists, KAKENHI (18405006) and grants from the

ChEss/Fondation Total, Myers Trust, and Packard Foundation.

Author Information Correspondence and requests for materials should be addressed to K.C.R.

([email protected]) or H. W. ([email protected]).

Figure 1| Brood pouch of I. nautilei. a, Sagittal section of head-foot showing position of an empty brood pouch (bp). Ventral side of foot is on the left. Nerve ring (nr), odontophore (od), and snout (sn) are visible. b, Detail of empty brood pouch, with surrounding circular muscle

12 layers and branching septae (s) inside lumen. c, Brood pouch (bp) full of embryos. Anterior pedal gland (ap), brood pore (p), cephalic tentacle (t), nerve ring (nr), and snout (sn) are visible. d, Histological section of foot (f), brood pouch (bp) with embryos (e), and brood pore (p). e, Cross section of pallial oviduct with unspecialized basophilic epithelium (be) and furrow (f) through which egg strings pass. Mantle cavity (mc) with hypobranchial gland (hg) is on the right. Scale bar, 100 µm. f, Detail of brood pouch with embryos (e). Note septum (s) with muscle fibers (m) and possibly cuticular lining (l). g, SEM of an egg string.

Figure 2| Development of Warén’s Larva. a, Light micrograph of early Warén’s larva showing simple anterior lobe (al), posterior lobe (pl) and larval cuticle (lc). Scale bar, 50 µm. b, Posterior lobe, showing tuft of compound cilia (cc) adjacent to the field of simple cilia (sc) that cover the rest of the larva. Scale bar, 25 µm. c, Detail of posterior lobe, showing larval cuticle (lc) and compound cilia (cc) penetrating cuticle. Scale bar, 10 µm. d, Scanning electron micrograph (SEM) of larval cuticle, with simple cilia and obvious holes from which cilia have previously emerged. Scale bar, 2 µm. e, SEM of posterior lobe, showing large compound cilia (cc) and

14 smaller simple cilia. Scale bar, 10 µm. f, 0.5µm plastic section of an early Warén’s larva showing cells full of large, darkly staining lipid globules (lg) and the larval cuticle (lc) penetrated by cilia. Scale bar, 20µm. g, SEM of late Warén’s larva, showing complete ciliation, two anterior lobes (al) and posterior lobe (pl). Scale bar, 50 µm. h, Semi-thin (0.5µm) section of later Warén’s larva stained with toludene blue. Lipid globules are present in many of the large cells and the larval cuticle (lc) is clearly visible. The mouth (mo) is visible, as well as a posterior tissue of smaller cells that may become the shell gland (psg). Scale bar, 20µm. i, Laser confocal projection of a late bi-lobed Warén’s larva stained with propidium iodide. Cell boundaries are not visible, but lipid globules (lg) are common in the equatorial region and a cup-shaped putative shell gland (psg) is present. The paired anterior lobes (al) and single posterior lobe (pl) are visible. Scale bar, 50 µm. j, Anterolateral view of Warén’s larva showing position of mouth. Scale bar, 50 µm. k, Detail of mouth (mo). Scale bar, 50 µm. l, Early veliger larva (20-days-old under 4 ºC conditions) with anterior lobes (al) still intact and the larval protoconch (pr) developing. Scale bar, 100µm. m, Veliger larva (15-days-old under room temperature conditions) with a well-developed velum (ve), protoconch (pr) and operculum (op). Scale bar, 20µm.

CHAPTER 2

A COMPARATIVE DESCRIPTION OF REPRODUCTIVE STRUCTURES IN IFREMERIA NAUTILEI AND ALVINICONCHA HESSLERI (CAENOGASTROPODA: PROVANNIDAE): HYDROTHERMAL VENT GASTROPODS FROM THE LAU BASIN

KYLE REYNOLDS1, ELLEN E. STRONG2 , CRAIG YOUNG3 & STACY KIM1

1 Benthic Ecology Lab Moss Landing Marine Laboratories 8272 Moss Landing Road Moss Landing, CA 95039, U.S.A. Phone: (831) 771-4112 Email: [email protected]

2 Smithsonian Institution National Museum of Natural History P.O. Box 37012 MRC 163 Washington, D.C. 20013-7012, U.S.A.

3 Oregon Institute of Marine Biology PO Box 5389 Charleston, OR 97420, U.S.A.

INTRODUCTION

The Lau Back-arc Basin in the southwest Pacific encompasses a region of hydrothermal vents in an area of diverse chemical and geological gradients. In sheer biomass, its community composition is dominated by molluscan megafauna similar to communities found at other

Western Pacific vent systems (refs from Podowski et al: Tunnicliffe & Fowler 1996; Van Dover et al. 2002; refs from Kojima 2000: Both et al., 1986; Okutani and Ohta, 1988). Within the Lau

Basin, the Eastern Lau Spreading Center (Fig 1) consists of several discrete venting sites characterized by basaltic substrate to the north and andesitic substrate to the south. The three dominant chemoautotrophic molluscs found in this community include the provannid snails

Alviniconcha hessleri (Okutani & Ohta 1988) and Ifremeria nautilei (Bouchet & Warén 1991), as well as the mytilid Bathymodiolus brevior (Fig 2). These species co-occur in a somewhat concentric, overlapping distribution pattern around areas of active venting; Alviniconcha consistently inhabit the hottest, most sulfide rich zones, closest to the venting source, surrounded by Ifremeria and then Bathymodiolus. A trend of decreasing temperature and sulfide levels can be found from Ifremeria populations to outermost Bathymodiolus populations (Podowski et al.

2009). This paper is primarily concerned with Alviniconcha and Ifremeria, gastropods that share a common ancestry.

Of all the living gastropod groups, freshwater, marine or terrestrial, caenogastropods have the highest species richness and possess the most diverse morphologies, physiologies, and life history strategies (Colgan et al 2007; Strong et al 2008). Ifremeria and Alviniconcha are highly modified provannids of the Abyssochrysoidea (Johnson et al., in prep.); this group has recently been placed in the Zygopleuroid Group within the Caenogastropoda – a superorder comprising

17 over 60% of all known gastropods (Bouchet & Rocroi 2005). Molecular analyses involving sequencing of the cytochrome oxidase I (COI) regions of mitochondrial genes has shown

Ifremeria and Alviniconcha to be sister species (Johnson et al., in prep). The two species share a unique anatomical organization involving an exchanged positioning of the digestive gland with the gonad (a proposed synapomorphy), reduced digestive glands, hypertrophied gills, primitive, open pallial gonoducts (Warén & Bouchet 1993), and unusually large gonads extending well into the neck when sexually mature. However, despite their recent common ancestry, shared anatomical modifications, and overlapping distribution patterns, these two species employ markedly different reproductive strategies. From larval shell morphology, Alviniconcha has been reported to have planktotrophic larvae (Warén & Bouchet 1993) while Ifremeria protects its embryos in a highly modified brood pouch (Chapter 1; the first report of brooding in a hydrothermal vent gastropod).

Original descriptions of these species involved sexually immature specimens, and much of the reproductive anatomy was underdeveloped or absent (Warén & Ponder 1991; Warén & Bouchet

1993). The goals of this investigation were: 1) to provide detailed descriptions of the reproductive anatomies and mechanisms of sexually mature Ifremeria and Alviniconcha specimens; 2) to describe these species’ reproductive strategies through the use of multiple temporally variant collections; and 3) to combine these data with phylogenetic data to infer the derivation of novel adaptations for either species, as well as any shared characteristics between the two.

MATERIAL AND METHODS

Expedition and collection data

MGLN07MV 06 Sep.-29 Sep. 2006, ROV Jason II dives 230-240, PI C. Fisher, (Ifremeria n=150; Alviniconcha n=90); TUIM07MV 11 Jun.-26 Jun. 2005, ROV Jason II dives 154-164,

PI C. Fisher (Ifremeria n=11; Alviniconcha n=23); TUIM05MV 06 Apr.-07 May 2005, ROV

Jason II dives 124-139, PI M. Tivey (Ifremeria n=1; Alviniconcha n=3); STARMER II Jul.

1989, (Ifremeria n=1); BIOLAU Jul. 1989 (Ifremeria n=3). The STARMER II specimen is from the Mussels Valley vent in the North Fiji Basin. All other expeditions took place in the

Eastern Lau Spreading Center (ELSC) of the Lau Back-arc Basin and included the vent sites

Kilo Moana, Abe, Tow Cam, and Tu’i Malila between approximately 1900-2700 m depth .

Samples were obtained with push corers and suction samplers manipulated by the ROVs.

Laboratory protocol and sample treatment

Shell heights were determined to be arbitrary measures due to their decollate nature and the inconsistent dimensions of their calcareous plugs. Instead, maximum shell width was used to gauge specimen size; measured from the outermost edge of the shell opening to the opposing side of the first whorl. Specimens were removed from their shells onboard the ship and transferred to 70% ethanol after an initial 24 hr fixation in 10% formalin. The exact handling of specimens from the 1989 collections is unknown, however, samples were ultimately preserved in

70% ethanol.

For histology, tissues from the gonad of all specimens from the 2005 and 2006 collections, and brood pouches from 60 of the female Ifremeria from 2006 were dissected out, dehydrated in

100% isopropanol for 2 hours, cleared in Histoclear for 8-12 hours, impregnated and embedded

19 in paraffin wax in a 65 ºC oven for ~6 hours, and sectioned via microtome at 7 µm thicknesses.

Slide-mounted sections were stained with Haematoxylin ‘Z’ and Eosin using standard histological protocol (via methods used by Tyler, et al. 2007). Slides were examined using an

Olympus BX50 light microscope fitted with a Canon Powershot S3-IS with Ultrasonic Image

Stabilizer.

For scanning electron microscopy, Ifremeria embryo transport packets were removed from various parts of the reproductive tract and neck furrow, and embryos were removed from the brood pouch; both of which were dehydrated to 100% EtOH, critical point dried using an

Emscope CPD 750 Thermoelectric Control, and photographed with a Tescan Vega scanning electron microscope at varying magnifications. In addition, one individual on MGLN07MV, released embryos into a bucket, which were harvested, critical point dried and examined. Brood pouches from female Ifremeria were dissected out, critical point dried, and imaged as well to view septae, pouch pore, and outer wall structures.

RESULTS

Family Provannidae (Fig. 3)

Ifremeria nautilei

Reproductive stage: All specimens from 2005 and 2006 were sexually mature (i.e. in a progressed state of gametogenesis), with gonads extending anteriorly from 0.20 whorl from the caudal end, to 0.1 whorl from the edge of the mantle. Oocytes were observed from previtellogenic to vitellogenic stages within the ovary. Other collections ranged from sexually immature (or in a non-reproductive period) to early sexual maturity. Histology was not performed on the gonads of specimens from other collections.

Foot and brood pouch morphology: The foot is oval, with rows of warts along the outer edge.

An anterior pedal gland resides along the anterior edge of the propodium. Subepithelial glands line the surface of the sole. In female Ifremeria, a modified metapodial pedal gland exists, but it has migrated anteriorly to the center of the foot, has apparently lost its glandular function, and instead functions as a brood pouch. Ventrally, a non-ciliated pore opens to the exterior within a small slit in the center of the longitudinal groove of the foot sole. The pore extends dorsally into the smooth muscle of the foot and opens to a pouch filled with septae. When devoid of embryos, the septae appear as folded layers of tissue radiating from the opening of the pouch pore (Fig 4).

As embryos enter the pouch via this pore, they fill the chambers created by the septae and expand the overall diameter of the pouch into a roughly spherical shape (Fig 5a,b). The function of the septae appears to be mainly structural and they may also serve as baffles to prevent loss of embryos to the outside. The pore is the only line of communication; there is no direct communication with internal reproductive anatomy.

A capsule surrounding the brood is lacking; the pouch and septae are instead lined with a thin continuous, non-cellular and non-ciliated membrane of undetermined origin, presumably secreted by surrounding cells (Fig 6). The pouch is surrounded by several layers of smooth muscle.

Ontogeny: The eggs of Ifremeria are small and packed with yolk. As they exit the ovary into a broadened, glandular region of the renal oviduct, a small grouping of eggs are bundled into a

21 transparent mesh-covered packet. This packet, equipped with a pointed head and long tail (10 mm in total length), passes through two glandular chambers which supply the packet with a dense mass of sperm from the adjoining seminal receptacles and then secrete a coating of what appears to be a fibrous mucus, sealing off the eggs and sperm (Fig. 7). This completed embryo transport packet (ETP) is guided by cilia along the pallial oviduct and out of the mantle cavity where it travels within a deep furrow along the side of the neck to the edge of the foot. From here, the packet is somehow directed to the center of the foot sole where it enters the brood pouch. It appears that this process continues repeatedly until the entire contents (tens of thousands of eggs) have been transported to the pouch via multiple ETPs. Adult size has no apparent correlation with empty or full brood pouches and onset of sexual maturity appears to occur very early. Sections through the brood pouch show sterroblastulae with small micromeres forming a cap on top of large yolky macromeres. The protoconch of the larval shell is unknown.

Females were found in one of two stages: either with brood pouches devoid of embryos or with brood pouches filled with thousands of embryos at roughly similar stages of development. The novel larval form of Ifremeria, known as Warén’s larva (Reynolds et al, in prep) was found in all brood pouches in a narrow range of morphologies. The earliest stages were globular in shape, progressing through a modified trochophore stage to a pre-veliger stage (Fig 8a-d), with a veliger stage being discovered and described later in a separate study (Chapter 1, Fig 2).

We observed one specimen releasing pre-veliger stage larvae in a collection bucket. The larvae were observed under a dissecting microscope and were swimming independently.

Pallial oviduct: The renal oviduct extends proximally from the posterior ovary wall and widens, becoming more glandular just past a connection to the gonopericardial duct. Here, the embryo transport packet (ETP) is apparently formed and filled with unfertilized oocytes before entering a dorsal spongy, glandular chamber that is open to several finger-like receptaculae seminalis on its proximal end. For simplicity, we will call this the mucus gland, as its function appears to involve a secretion of spongy, fibrous mucus over the ETP after sperm is combined to the oocytes within (Fig 9). Continuing distally, a narrow duct opens to the posterior region of the pallial oviduct.

The open pallial oviduct continues distally as a 2.5 mm high skin flap along the right side of the body wall, tapering to a point approximately 0.5 whorl anteriorly. From posterior to anterior, the epithelium of the ventral canal becomes less glandular and is not consistent with a typical capsule gland. Within the pallial oviduct sits a free, convoluted finger of ciliated epithelium that extends past the distal end of the skin flap, through the mantle skirt, becoming the right ridge of the neck furrow. This structure obviously functions in guiding the ETP out of the mantle cavity anteriorly to the epipodial furrow, and the edge of the foot sole.

Alviniconcha hessleri

Reproductive Stage: Specimens from the 2006 collection appear to be sexually mature, with gonads extending anteriorly from 0.20 whorl from the caudal end, to 0.1 whorl from the edge of the mantle. Oocytes were observed from previtellogenic to vitellogenic stages within the ovary.

Histology was not performed on the gonads of specimens from other collections.

Ontogeny: The eggs of Alviniconcha are small and packed with yolk. They appear similar in size and content to Ifremeria eggs. No embryos or larvae were found for this species.

Foot morphology: A metapodial pedal gland is present in female Alviniconcha, opening via a long duct to the posterior fifth of the foot sole (Fig 10). No brood pouch is present in the foot.

Pallial oviduct: Overall, the pallial oviduct is very weakly glandular. The renal oviduct extends proximally from the posterior ovary wall, and just past a connection to the gonopericardial duct it broadens and becomes glandular before leading into what is presumably analogous to an albumen gland, but will here be called the proximal gland (Fig 11). The inner epithelium of the gland is coated in short cilia and lined with long columnar cells containing basal nuclei. The epithelium is very convoluted, filling much of the interior of the gland; however, it appears to be much less glandular than that of a typical albumen gland. Fingers of receptaculae seminalis open up on the ventral side of this gland and canals lead dorsally to a connection with the open pallial oviduct.

The pallial oviduct continues distally as a thin skin flap for approximately 0.5 whorl, all the while becoming less glandular within the epithelial lining. The histological appearance of the distal pallial oviduct is not consistent with a typical capsule gland.

DISCUSSION

SPECIES COMPARISONS

Reproductive stage:

It is significant that all 162 Ifremeria and 116 Alviniconcha specimens collected in the April

2005, May 2005 and September 2006 samples across four distinctly separate hydrothermal vents

(a total distance of approximately 175 km) were in a state of sexual maturity regardless of their size. This is in contrast to the sexually immature specimens of both species collected and described by Warén and Ponder (1991) and Warén and Bouchet (1993) from May of 1989. This may suggest non-continuous, or even simultaneously coordinated reproduction. Since all specimens within a given collection were found either sexually immature or sexually ripe, without regard to body size, it suggests that reproduction in both species is iteroparous (Young,

2003).

Foot morphology: Histological evidence exists to confirm that the brood pouch of Ifremeria is in fact a modified and relocated metapodial pedal gland. Histology also confirmed the presence of an unmodified metapodial pedal gland in Alviniconcha in the proper anatomical location for such a structure; a brood pouch is lacking in the foot or elsewhere. The presence of a metapodial pedal gland in these and other provannids, whether modified or not, is a pleisiomorphy of this family, as the littorinids and other abyssochrisids have lost this feature (David Reid, pers. comm.).

Ontogeny: The extensive ciliation of Ifremeria’s unique larval form provides a means for independent locomotion, presumably aiding in its dispersal, and thus may provide the larvae an

25 avoidance mechanism for intolerable environmental conditions (Chapter 1). Duration of development within the brood pouch and time of release is unknown, however, larvae have been shown to be capable of successful development outside of the brood pouch after being released at a very early pre-veliger stage (Chapter 1).

No embryos or larvae have been recorded for Alviniconcha.

Pallial oviduct: It can be assumed that the proximal gland on the pallial oviduct in both

Ifremeria and Alviniconcha is analogous to an albumen gland. Also, although the histology does not support the presence of a capsule gland in the distal pallial oviduct, it is conceivable that the mucus gland of Ifremeria is analogous to a capsule gland (albeit a modified and relocated capsule gland). More thorough analysis would be needed of the cellular structure and gland products to determine this conclusively. Regardless, it is evident from the cellular morphology of the pallial oviduct that neither species provides their embryos with egg capsules. In fact,

Alviniconcha appears to provide even less in terms of protective provisioning for their embryos than Ifremeria; a puzzling finding for a planktotrophic species. Reduced glandularity in the pallial oviduct is generally indicative of a brooding strategy, as a protected brood requires less complex encapsulation. This might suggest some kind of external brood protection strategy in

Alviniconcha, but without any evidence of this, and in light of the planktotrophic morphology of the larval shell, it is impossible to speculate.

The open condition of the pallial gonoduct is a feature shared with the littorinids, and apparently of functional significance. Reid (1989) speculated that this condition may benefit a quicker

26 transport of sperm than could occur in a narrow, closed tube – a trait that could be important for littorinids to help them avoid dessication and dislodging while copulating in the intertidal. For

Alviniconcha and Ifremeria, the functional benefit is not as apparent and will need further investigation.

There is no context in which to put the ETP mechanism of Ifremeria, as no similar mechanism has been reported with a dual functionality of combining unfertilized eggs with sperm as well as aiding in internal and subsequent external transport to a brood pouch. It is tempting to draw an analogy between an ETP and an egg capsule, however, the similarities are rather weak and the functionalities appear vastly different. Further analysis of this structure is obviously required.

BROODING IN CAENOGASTROPODS

Brood pouches have evolved many times and in many anatomical positions in Caenogastropoda, including: pouches within modified sections of the reproductive tract, as in Littorina saxatalis

(Reid, 1989) and Polygireulima (Warén 1983); or pouches formed by deep ectodermal invaginations within the head-foot, as in several cerithioideans (Seshaiya 1940; Houbrick 1987;

Strong 2002). Brooding can also be found within the mantle cavity, as in Littorina scabra

(Struhsaker 1966; Rosewater 1980), or even within the folds of the foot, as in Bullia mellanoides

(Ansell & Trevallion 1970)

There is only one other report of a brood pouch in a modified pedal gland - it is reported in a volutid species (Marche-Marchad 1968). However, this is where the similarity to Ifremeria’s brood pouch ends, as the volutid pouch occurs in the non-homologous anterior pedal gland and

27 contains larvae in egg capsules that undergo normal development and emerge as crawl-away shelled veligers. In fact, no caenogastropods have ever been reported to be released prior to the veliger stage, as is found in Ifremeria.

CONCLUSIONS

The novel brood pouch, ETP mechanism, and larval form of Ifremeria are apomorphies, not only relative to this clade, but relative to all known gastropods. The recently described larval form of

Ifremeria is unique not only among marine gastropods, but among all known marine invertebrate larval forms. Further research is necessary to determine the significance of Warén’s larva, and to conclusively determine the mode of development of Alviniconcha larvae.

ACKNOWLEDGEMENTS

We are grateful to Anders Waren for bringing the presence of the brood pouch to our attention.

We thank SJSU & OIMB for use of SEM; Paul Tyler for histology facilities; Paul Greenhall and

Marilyn Schotte for assistance with handling specimens; Marilyn Schotte for some of the histology and SEM; and Kamille Hammerstrom and Gabriela Vega for assistance with dissections and histology.

REFERENCES

Ansell, A.D. & A. Trevallion. 1970. Brood Protection in the Stenoglossan Gastropod Bullia mellanoides (Deshayes). Journal of Natural History, 4: 369-74.

Houbrick, R.S. 1987. Anatomy, reproductive biology and phylogeny of the Planaxidae (Cerithiacea: Prosobranchia). Smithsonian Contr. Zool. 445: 1-57.

Marche-Marchad, I. 1968. Un nouveau mode de developpement intracapsulaire chez les Mollusques proso-branches neogastropodes: l'incubation intrapedieuse des Cymba (Volutidae). Comptes rendus hebdomadaires des seances de l'Academie des Sciences. 266D:706-709.

Ponder, W. F., D. Colgan, J. Healy, A. Nützel, L. R. L. Simone & E. E. Strong. 2008. Caenogastropod Phylogeny. In: W.F. Ponder & D.L. Lindberg (eds.) Molluscan Phylogeny. U. California Press, 331-383. Reid, D.G. 1989. The Comparative Morphology, Phylogeny and Evolution of the Gastropod Family Littorinidae. Philosophical Transactions of the Royal Society of London. Biol. Sci. 324: 1-110.

Rosewater, J. 1980. Subspecies of the gastropod Littorina scabra. Nautilus. 94: 158-162.

Seshaiya, R. 1940. A free larval stage in the life history of a fluviatile gastropod. Curr. Sci., 9: 331-332.

Strong, E. E. & M. Glaubrecht. 2002. Evidence for convergent evolution of brooding in a unique gastropod from Lake Tanganyika: anatomy and affinity of Tanganyicia rufofilosa (Smith, 1880) (Caenogastropoda, Cerithioidea, Paludomidae). Zoologica Scripta, 31: 167-184.

Strong, E. E. 2003. Refining molluscan characters: morphology, character coding and the phylogeny of the Caenogastropoda (Gastropoda). Zoological Journal of the Linnean Society, 137: 447- 554.

Strong, E. E. & M. Glaubrecht. 2007. The morphology and independent origin of ovoviviparity in Tiphobia and Lavigeria (Caenogastropoda, Cerithioidea, Paludomidae) from Lake Tanganyika. Organisms, Diversity and Evolution, 7: 81-105.

Struhsaker, J.W. 1966. Breeding, spawning, spawning periodicity and early development in the Hawaiian Littorina: L. pintada (Wood), L. picta Phillippi and L. scabra (Linné). Proc. Malac. Soc., London, 37: 137-166.

Warén, A. 1983. A generic revision of the family Eulimidae (Gastropoda, Prosobranchia). J. Moll. Stud., 13:1-95.

FIGURES

Figure 1. Bathymetric map of the Eastern Lau Spreading Center. Black dots represent sites of active venting. (Image from http://venturedeepocean.org)

A B

C D

Figure 2 (a-d). In situ photos of Lau Basin mollusks: A Alviniconcha hessleri, B Ifremeria nautilei, C Bathymodiolus brevior, D aggregation of all three species in their typical concentric distribution pattern around diffuse hydrothermal venting (Photos courtesy of Chuck Fisher)

Figure 3. Representative illustration of external morphology for Ifremeria and Alviniconcha. Abbreviations: dg, digestive gland (grey); f, foot; ms, mantle skirt; op, operculum; ov, ovary (pink)

Figure 4 (a,b). Sagittal section of head-foot of sexually immature Ifremeria with empty brood pouch, showing morphology of septae. Abbreviations: ap, anterior pedal gland; bp, brood pouch; nr, circum-esophageal nerve ring; od, odontophore; s, septae; sn, snout.

B Figure 5 (a,b). Ifremeria brood pouch location and morphology. A Sagittal section of head- foot of sexually ripe Ifremeria with brood pouch full of embryos. (Photo courtesy of Anders Warén); B SEM image of cross-sectioned brood pouch showing embryos lining walls of septae. Note: majority of embryos were removed for better viewing. (SEM image courtesy of SJSU Physics Laboratory)

Em O W

Figure 6. Histological section showing outer brood pouch wall merging continuously with inner septum. Note the continuous acidophilic, non-cellular lining that continues along the outer wall and around the septum (black arrows). Abbreviations: Em, embryo; OW, outer brood pouch wall; S, septum.

Figure 7. SEM image of embryo transport packet (ETP). White arrows point to embryos that were stuck to the outside. Strings of embryos reside within the head. Abbreviations: H, head; T, tail.

A B

C D

Figure 8 (a-d). SEM images of progressive developmental stages of Warén ’s Larva from different specimens of female Ifremeria. A Early undifferentiated globular stage, B Trochophore-like stage with ciliation beginning (Note broad non-ciliated swath near the broad end), C Trochophore-like stage with full ciliation, D Pre-veliger stage with velar lobes forming at broad end.

ETP

MG PG

POC

Figure 9. Histological cross-section of the proximal region of the pallial oviduct complex of Ifremeria. Abbreviations: ETP, embryo transport packet; MG, mucus gland; PG, proximal gland; POC, pallial oviduct canal; SR, seminal receptacles.

Figure 10 (a,b). Sagittal sections showing foot morphology of Ifremeria and Alviniconcha to compare metapodial pedal gland locations. A Ifremeria metapodial pedal gland migrated anteriorly, modified internally into a brood pouch, and communicating to exterior by a pore. B Alviniconcha retains the position and morphology of a normal metapodial pedal gland opening at the posterior of the foot. (Arrow shows duct leading dorsally). Abbreviations: ap, anterior pedal gland; bp, brood pouch; mp, metapodial pedal gland; nr, circum-esophageal nerve ring; od, odontophore; p, pore; sn, snout; t, cephalic tentacle.

SR PG

Figure 11. Histological cross-section of the proximal region of the pallial oviduct complex of Alviniconcha. Abbreviations: BS, blood sinus, PG, proximal gland; PO, pallial oviduct; SR, seminal receptacle.