Cah. Biol. Mar. (2012) 53 : 77-87

Evolutionary history of Idas sp. Med (: ), a cold seep mussel bearing multiple symbionts

Julien LORION 1, Sébastien HALARY 2, Joana do NASCIMENTO 2, Sarah SAMADI 3, Arnaud COULOUX 4 and Sébastien DUPERRON 2 (1) Japan Agency for Marine Science and Technology (JAMSTEC), Marine Ecosystem Department, 2-15 Natsushima, 237-0061 Yokosuka, Japan Tel: +81-46-867-9570, Fax: +81-46-867-9525, E-mail: [email protected] (2) UMR 7138 UPMC-IRD-MNHN-CNRS, Equipe « Adaptation au Milieux Extrêmes », Université Pierre et Marie Curie, 7 quai St Bernard, 75005 Paris, France (3) UMR 7138 UPMC-IRD-MNHN-CNRS, Equipe « Espèces et Spéciation », Muséum National d'Histoire Naturelle, 57 Rue Cuvier CP 26, F-75231 Paris Cedex 05, France (4) GENOSCOPE, 2 rue Gaston Crémieux, CP 5706, 91057 Evry Cedex France

Abstract: Small mytilids of the genus Idas are related to the large mussels found worldwide at deep-sea hydrothermal vents and cold seeps. They are therefore keys to a better understanding of the colonization of vents and seeps by symbiont-bearing organisms, but still little is known about their biology. For this study, specimens of a mytilid referred to the genus Idas were collected from various substrates in a cold seep area near the Nile deep sea fan. Based on molecular and morphological data, all specimens are confirmed to belong to a single species of the genus Idas , which was previously shown to host six distinct bacterial symbionts. Its larval shell characteristics indicate a long planktonic phase, which could explain its close relationship to a mussel species that occurs in the Gulf of Mexico. 3-D FISH indicates the dominance of sulfur-oxidizing, methane-oxidizing and methylotrophic symbionts in all specimens analysed.

Résumé : Histoire évolutive de Idas sp. Med (Bivalvia : Mytilidae), une moule de suintements froids associée à plusieurs bactéries symbiotiques. Les mytilidés du genre Idas forment un groupe monophylétique avec ceux associés aux sources hydrothermales et aux suintements froids. Ils sont donc la clé d’une meilleure compréhension du scénario de colonisation de ces environnements par les organismes symbiotiques. Pour autant, leur biologie est encore aujourd’hui fort peu connue. Pour la présente étude, des moules ont été collectées à partir de différents substrats aux alentours du delta profond du Nil. Nous utilisons ici une approche moléculaire et morphologique pour démontrer que ces spécimens appartiennent à une espèce du genre Idas , dont il a été montré précédemment qu’elle était en association symbiotique avec 6 lignées bactéri - ennes, un cas sans précédent dans ce taxon. La coquille larvaire suggère une longue phase planctonique qui pourrait expli - quer la parenté avec une espèce présente dans le Golfe du Mexique. Une approche de 3D-FISH indique la dominance de symbiontes de types sulfoxydants, méthanotrophes et méthylotrophes dans tous les individus analysés.

Keywords: Idas l Planktotrophy l Organic falls l Cold seeps l Symbiosis

Reçu le 4 novembre 2010 ; accepté après révision le 26 avril 2011. Received 4 November 2010; accepted in revised form 26 April 2011. 78 BIOLOGY OF IDAS SP. MED

Introduction methane- and two sulphur oxidizers, a methylotroph, and a bacterium of unknown metabolism) and a representative of Bathymodiolin mussels are among the dominant the Bacteroidetes. The closest relative of Idas sp. Med is macrofauna occurring at hydrothermal vents and cold seeps Idas macdonaldi Gustafson, Turner, Lutz & Vrijenhoek, worldwide (Olu-Le Roy et al., 2007). Distel et al. (2000) 1998, which lives at cold seeps in the Gulf of Mexico and first showed that these deep-sea mussels cluster within a harbours sulfur-oxidizing bacteria (Won et al., 2008). It is single clade that also includes other mussels living on striking that such a trans-Atlantic geographical pattern is sunken organic substrates such as wood and vertebrate also reported for other species complexes in the bones. This extended bathymodiolin clade was later bathymodiolin clade, namely Bathymodiolus boomerang confirmed by other phylogenetic studies (Iwasaki et al., Cosel & Olu, 1998 and Bathymodiolus mauritanicus Cosel, 2006; Samadi et al., 2007; Fujita et al., 2009; Kyuno et al., 2002 (Olu-Le Roy et al., 2007; Génio et al., 2008). In these 2009; Miyazaki et al., 2010). Recently, improved sampling species complexes, the large scale at which occurs genetic of mussel species associated with sunken organic substrates structure is explained by the occurrence of highly suggested that several habitat shifts occurred during the dispersive larvae (Lutz et al., 1984; Tyler et al., 2007). evolution of the bathymodiolin clade (Lorion et al., 2010). Until now only four specimens of Idas sp. Med from a Apart from early morphological descriptions, the mussels single collection site have been analysed and this species associated with organic falls have received attention only remains poorly known (Duperron et al., 2008a). In recently (Deming et al., 1997; Gros & Gaill, 2007; particular, there is little or no data regarding (i) its Duperron et al., 2006, Duperron et al., 2008a; Samadi et al., taxonomic status (ii) its developmental mode (iii) the 2007; Southward, 2008; Lorion et al., 2009; Tyler et al., flexibility of its symbiotic associations and (iv) its 2009). Because there are many species of these mussels, an evolutionary history. Recently, three cruises explored the investigation of their biology is needed to improve our eastern Mediterranean cold seeps and yielded additional understanding of the evolution of the whole group and the specimens. These were collected from three substrates: adaptive processes that allowed the colonization of deep- authigenic carbonates, tubes of the siboglinid annelid sea hydrothermal vents and cold seeps. Lamellibrachia sp., and wood colonization devices The presence of symbiotic chemosynthetic bacteria in deployed for a year away from dark sediment patches, mussel gills is one of their most striking characteristics aggregates, or carbonate crusts (and thus hopefully (reviewed in Duperron et al., 2009). Bacteria are mostly from the direct influence of seepage). Here, morphological sulfur-oxidizing Gammaproteobacteria, which sustain part and molecular analyses are employed to test whether all of the mussel nutrition and therefore represent a major specimens correspond to one species, Idas sp. Med, and adaptation to sulfide-rich deep-sea environments discuss its similarity with species already described (Cavanaugh et al., 2006). Among deep-sea mussels, some morphologically in the Mediterranean sea. Dispersal species harbour several bacterial strains. The most common capabilities were evaluated by analysing larval shell bacterial type in addition to sulfur oxidizers is methane- characteristics. The presence of the six reported symbionts oxidizers, which are found in species sampled at methane- was tested in mussel specimens sampled on the different rich vent and seep sites (Cavanaugh et al., 1987). One may substrates (Duperron et al., 2008a), and symbiont relative hypothesize that the acquisition of additional symbionts abundances were estimated using the newly developed 3- allows greater flexibility of the host in the use of substrates Dimensional Fluorescence In Situ Hybridization (3D- and therefore enhances adaptability (Cavanaugh et al., FISH) technique (Halary et al., 2008). Finally, the 2006). Moreover, some species display more than two evolutionary history of Idas sp. Med is discussed here in the lineages of bacteria in their gills, and it has been shown that light of the new biological and phylogenetic data. these multiple associations are highly flexible (Distel et al., 1995; Halary et al., 2008). To better understand how flexibility of symbiotic Material and Methods associations has influenced evolution of deep-sea mussels, the undescribed species so-called Idas sp. Med, sampled in Sampling the eastern Mediterranean, seems a relevant model. Indeed, Idas sp. Med has been collected from cold seeps and The NAUTINIL (2003, chief scientist J. P. Foucher, harbours six distinct symbionts, based on their 16S rRNA, specimens M1 to M4), BIONIL (2006, chief scientist A. although it belongs to a clade including mostly mussels Boetius, specimens M11 to M27) and MEDECO (2007, from organic falls bearing only sulfur-oxidizing symbionts chief scientist C. Pierre, specimens M31 to M49-3) cruises (Duperron et al., 2008a, Lorion et al., 2010). The six from the EU Mediflux / HERMES program allowed the symbionts include five Gammaproteobacteria (one collection of 24 mussel specimens at three locations of the J. LORION, S. HALARY, J. do NASCIMENTO, S. SAMADI, A. COULOUX, S. DUPERRON 79

Nile deep-sea fan (Table 1). Samples were recovered from and evolutionary significant unit (ESU) of the lineage the Central Zone, from the Amon mud volcano, and from which includes Idas sp. Med and I. macdonaldi (lineage Caldera (Kheops mud volcano) in the vicinity of the Nile L6, Lorion et al., 2010; Table 2). According to that study, deep-sea fan at depths between 1150 and 3000 m. Mussels sequences of Bathymodiolus thermophilus Kenk and were found attached to three types of substrates: authigenic Wilson 1985 and B. aff thermophilus were also included carbonate crusts, tubes of the siboglinid tubeworm and these two sister species were used as an outgroup. Lamellibrachia sp., and wood chips consisting of Douglas Datasets were aligned using the module Clustal W in Mega Fir deployed in experiments during the BIONIL cruise. For 4 (Tamura et al., 2007). Positions of the 28S rRNA six specimens, one gill was fixed for in situ hybridization as alignment for which homology was ambiguous due to described previously (Duperron et al., 2008b). Remaining insertions and deletions were not considered in analyses. tissues and the 17 other specimens were stored in 95% The best fitting models of nucleotide substitution were EtOH for genetic analyses. selected using JModeltest and the AICc criterion (Posada & Crandall, 1998). These models were set in Bayesian Morphological examination analyses performed with Beast 1.5.3 (Drummond & Rambaut, 2007). The Yule model was used as the tree prior The shell morphology of 8 new specimens was compared and for each gene, the heterogeneity of the mutation rate with published descriptions and pictures of type specimens across lineages was set under uncorrelated log-normal of Idas species reported from the Mediterranean Sea and clocks for each gene. In an attempt to estimate the age of Atlantic Ocean: I. argenteus Jeffreys, 1876, I. modiolae - the divergence between I. macdonaldi and the species formis (Sturany, 1896), I. simpsoni (Marshall, 1900), I. documented here, the COI mtDNA mutation rate was set dalmasi Dautzenberg, 1927, I. ghisotti Warén & Carrozza, under a normal prior whose 95% confidence interval 1990, I. macdonaldi Gustafson, Turner, Lutz & Vrijenhoek, ranged from 1% to 2% per million years, according to Won 1998 and I. cylindricus Pelorce & Poutiers, 2009. Type et al. (2003). For each dataset, four independent analyses, specimens of I. simpsoni , I. ghisotti and I. cylindricus were starting from distinct coalescent trees, were run over 40 available for direct comparison. Larval shells of two million generations and sampled each 2,000 steps. After specimens (M3 and M49.3) were investigated in detail. analyzing results with Tracer v1.4.1 and discarding 10% Valves were mounted on aluminium stubs, covered with a samples as a burn-in, independent runs were pooled and 700 Å layer of gold-palladium, and examined under a JEOL resampled each 8,000 steps. The maximum clade credibili - JSM 840A Scanning Electron Microscope (SEM). ty tree, the posterior probabilities of its nodes, divergence estimates and corresponding 95% Highest Posterior Host genetics Densities (95% HPD) were calculated from these pooled DNA was extracted using the QIAamp ® DNA Micro Kit results. Kimura two-parameters (K2P) genetic distances (Qiagen). A fragment of the Cytochrome Oxidase I (COI) were calculated with Mega 4. mitochondrial gene was amplified using the primers H691 (5’-GTRTTAAARTGRCGATCAAAAAT-3’), which was Symbiont relative abundances using 3D-FISH designed for deep-sea mussels, and LCO1490 (Folmer et Gill tissues of specimens M31, 32, 33, 38-3, 45 and 46, al., 1994). Domains D1, D2 and D3 (Qu, 1986) of the 28S representing distinct habitats (exposed crust, crust covered rRNA nuclear gene were amplified using primers C1’ with dark sediment, and wood) were embedded in (5’ACCCGCTGAATTTAAGCAT3’) and C4 (5’TCG - Steedman’s wax as described in Duperron et al. (2008a). GAGGGAACCAGCTACTA3’). PCR reactions were Transverse sections of gill filaments, 10 µm-thick, were cut performed in 25 µL final volume, containing approximate - using a microtome (Jung, Germany) and deposited on ly 3 ng template DNA, 1.5 mM MgCl2, 0.26 mM of each Superfrost Plus slides. Hybridizations were performed nucleotide, 0.3 µM of each primer, 5 % DMSO and 0.75 using protocol and probes from Duperron et al. (2008a). unit of Taq Polymerase (Qbiogene). Amplification products Individual hybridization with each of the six probes were generated by an initial denaturation step of 4 min at described in Duperron et al 2008a was performed to test 94°C followed by 35 cycles at 94°C for 40 sec, 50°C (52°C qualitatively the presence of each symbiont. To estimate the for 28S rDNA) for 50 sec and 1 min at 72°C, and by a final percentage of bacterial volume originating from methano- extension at 72°C for 10 min. PCR products were purified and methylotrophs, probes ImedM-138 (methanotrophs) and sequenced in both directions at the Genoscope (Evry, and BhecM2-822 (methylotrophs) were both Cy3-labelled France). Sequences were deposited in the NCBI database and tested against a Cy5-labelled Eub338 probe (Amann et (www.ncbi.nlm.nih.gov, Table 1). al., 1990; Duperron et al., 2008a). To estimate the Sequences obtained were added to a dataset including percentage of bacterial volume originating from sulfur COI mtDNA and 28S rRNA sequences from each species oxidizers, probes Bthio-193 and ImedT2-193, specific for 80 BIOLOGY OF IDAS SP. MED , r - s e 5 o x ) A e r r é S 4 u 7 a b é E u e F

2 6 0 l q c M , S 6 2 8 6 m m l A r

( 5 - - - - A 7 3 5 2 1 0 9 8 7

4 M t e _ u u - a , a G

4 1 7 5 5 5 4 4 4 4 3 3 3 e 1 h 8 t n n , 9 N 3

-

5 5 4 3 s m 5 5 5 5 5 5 5 5 5 a N 9 6 4

4 e e e 5 5 5 5 R t 9 9 9 9 9 9 9 9 9 l R d 7 l 2 5 é 5 A p

r 9 9 9 9

a n 5 5 5 5 5 5 5 5 5 _ E

9 n 9 P y o e 5 5 5 5 v I o i 1 1 1 1 1 1 1

1 1 M b t 5 S 5 r ( u s J J J J J J J s J J 1 1 1 1 N 2 o

1 N 8

1 a m J J J J e / q l F F F F F F F F F l J J

2 8 a é

H 0 F F F F p A i - F f F x n a s 7 P S S o e - I n h n

i F 3 M o F I n é

a - S 4 t , d

o

, ) O é 6 i _ 3 0 D A , e s t

6 2 8 6 4 5 8 C 3 - - - - é a a 7 3 4 5 2 1 0 9 8 7 D d

- N , 0

a - 4 1 7 5 u n 8 8 7 7 7 7 6 6 6 e M e t 9 e 3 i e 8

2 D 8 8 7 6 q 5 5 5 5 5 5 5 5 5 h l a d s 7 t 7 i _ t l o 5 5 5 5 8 8 8 8 8 8 8 8 8 m a

d O i t o 5 5 d b

8 8 8 8 s a 5 5 5 5 5 5 5 5 5 b m u h 8 C 8 n 2

o 1 t a 5 5 5 5 x 1 1 1 1 1 1 1 1 1 i / q

5 I a 5

é E J J J J J J J J J l 1 1 1 1 e l 0 N t o 1 l

1

M J J J J F F F F F F F F F O c 7 J D ( n m e J s

F F F F

M o F n E F C a a w 3 M s l a l

- E

s r , M d r t S e

) a e

u

n é s i e d 3 r , e s é o n n p 2 i o a d r d l a t d f a t y 1 2 2 5 2 2 2 4 2 5 3 2 5 1 2 e t

n

a t t s P M ) s u i d o - a

p p p p p p p p p p p p p p p d o

1 , f

n l e g c a a a a a a a a a a a a a a a 1 é 4 i s 3 e o e o t p 6 r r H H H H H H H H H H H H H H H d u 1 é a 5

N m M p n n i

a e 1 i e , M H

c

s a

8 s l ) e D é , o l

r - 3 3 t p 1 e h a - s t L e 1

u

M 9 m m m m m m m m m m m m m m m m h h ( q

s

N e d 4 t

M t

t s e 9 9 0 0 0 0 0 5 5 5 5 5 5 3 3 7 n ( p s 6 c e p

l 2 2 0 0 0 0 0 9 9 9 9 9 9 a 5 5 5 - L M l e e e

e 8 1 1 7 7 7 7 7 9 9 9 9 9 9 l 1 1 1 d . r à X - l d

b

D 2 2 1 1 1 1 1 1 1 1 2 2 2 2 2 2 u e o S A d a 5 O l o c n n F

4

i N B p t a

S e -

n R s m M s _ d 9 r e (

a n e 2 2 2 9 9 9 9 9 9

8

3 v t 5 5 1 1 1 1 1 e e x 4 4 4 0 0 0 0 0 0 9 1 S t 3 e a 5 5 2 2 2 2 2 e i - ° ° ° ° ° ° ° ° °

7 8

m ° ° ° ° ° ° ° n s D 1 1 1 8 8 8 8 8 8 i o s i

_ R 2 - 9 9 0 0 0 0 0 t c

3 3 3 2 2 2 2 2 2 a e d

b 2 E t 2 2

3 3 3 3 3

l é r I d

2 e w , E E E E E E E E E E E E E E E E p o /

O

n

t t s l N o 0

l a o C 8 8 2 2 2 2 2 2 2 2 8 8 8 8 8 8 A a c 7 s r e A l

p 3 3 2 2 2 3 3 3 3 3 0 0 0 0 0 0 t E

, e N P s h ° ° ° ° ° ° ° ° ° ° ° ° ° ° ° ° s s M g - s e D L 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 b D

e

r 3 n , t t . 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 E u r e i )

4 o * s s o s m 4

3

M * o I L N N N N N N N N N N N N N N N N c

e

n , t h D M s d O ) o - e

l

i

w 3 2 à t C e e

*

3 s c s

p O e s 1 e c n y A l M d u t C e e l

2 M o o o o o o

v o E o s e ( M m t l e a c

n n n n n n e

i

t

D . a a a a a a , c , 3 i c s 1 d - c c c c c c e s E e n e é l 3 r

l l l l l l t n e p u s e o a o o o o o o M i g u r M

q S k b v v v v v v t t (

a

n r r q

. s e a a a p n é d d d d d d

3 i b

s A ) P - A A A A

u u u u u u

m m m u - o o o 1

e 2 2 2 2 s N s a X k t

r 1 n n n m m m m m m

4 e

c c e e e e f n a a a a R l 6

O s s s s s s t t t t

r

o e c c c o i i i i 5 M a

r

l l l p p p p p p

B s s s s p H 1 , e -

m L u S o o o o o o o o o , , , , ,

S 8 e p 8 e e e e e e o 8 v v v . D I e e e e e e e

3 - v y e 3 h h h h h h p 2 n n n n n n n i F t

d d d

d 3 t L - o o o o o o o K K K K K K , M d u u u k c u

( n z z z z z z z e D , , , , , , N t D n

n e

- s m m m

o l l l l l l l a a a a a a é a a 3 6 i

p i s

r r r r r r

2

a a a a a a a - t t s b u n n n e r r r r r r r e e e e e e g r A e t t a t t t t t t n O n . o o o X d d d d d d r n ) c e n e n n n n n n n l l l l l l

i C N C e t

O m m m e e o e e e e e e a a a a a a s . G m E s d n D u y

B . C L C A A A C C C C C C C C C C C t e

é o - a D d n r n s d 9 m e

u o p é E e

t i 3

a t I v s s a M 3 g a

M é s l

O s

’ n g E , e D i i r ) a d - C t c S h u

t t t t t t

7 p s t c 2 r . s

o n n n n n n n e 2 a r o o O e e e e e e ’ p e w v f r o

d M f n C s s é

m m m m m m w

s i , i i i i i i e é

r s E d r m w 6 s s d d d d d d e i o e a ( D u e e e e e e 2 y n r i

b a c l s s s s s s n e é e E h v

n a

M m s r c k k k k k k e ( n m m x a u a M

r r r r r r a i u a

l u u b r n a a a a a a ,

8

c q

a n ) - a s s b e d d d d d d

t

e i n

3 e S n t p l s s a l o h h h h h h l

3 e l n F e i t t t t t t S d i s e l

i i i i i i

e s S .

m u n M s t d i

m y _ w w w w w w A e q e l e

c n à a 5

E o c e e e e e e e e e e e e e e e e o m 0 L t t t t t t t t t t t t t e t c c 1 i v

p t

A p a a a a a a a a a a a a a i 8 a a f c 3 s s t x S

n n n n n n n n n n n n n _ r e G o

e e c e t k .

o o o o o o o o o o o o o b M r d d p t e d s N e ( n r d

b b b b b b b b b b b b b 2 o o

S p / b a r r r r r r r r r r r r r e b e

o n A s o o 3 l . a a a a a a a a a a a a a 0 b c u u - o

e P d

i M

7 n r s C C C C T S C C W W C C C C C C C t s e

X e e . n , a n M h p a * O

n M G e t s , i

d *

. ) B

. d m n s - p 3 m ) é i i d n

a s 8 a s e 2 c

n a i d d 3 é x d * s

l a I . 6 t e i

3 e M p 4 a t - r

a . ’ x s *

, d l u e D o ,

1 e

I M 3

* t - t

a s r * * h

s . e 1 , 3 t 2 e g u

u 4 * i e 2 1 p ) * 2 7 - n

a o 3 L n O h

M 3 y 3 * 2 w -

, a t e e

e p t

. -

,

l l M

C 3 * , p

s o d s 9 M . 1 n M , b b l m 3 E 2 1 2

a i

e é i 1 4 , - M * , - - - a a

* r p w s

k c D 1 3 8 5 6 , 3 6 8 1 7 8 9 9 d a i 1 g T T r e M l M w

3 3 3 4 4 e E 1 2 1 1 2 3 4 4 4 4 4 h a

i ( p s ’

t n w o u m t d 8 u e ( M M M M S M M M M M M M M M M M M M J. LORION, S. HALARY, J. do NASCIMENTO, S. SAMADI, A. COULOUX, S. DUPERRON 81 s e a d

the two sulfur-oxidizing symbionts, were both Cy3- l l e r e i labelled and used along with Cy5-labelled Eub338 a w

’ l s

a probe (Duperron et al., 2008a). Probes CF319 and t 0 8 2 0 6 3 4

5 0 1 1 6 5 0 A e 2 4 , 4 6 7 8 6 5 4

9 9 8 6 4 4 4 5 3 d r N 6 8 7 8 7 8 9 2 2 2 2 1 1 1 ImedG-193 were tested, but yielded very few e 5 5 u t 6 5 5 5 5 5 3 1 1 1 3 3 3 3 R 9 9 e a 6 6 6 6 6 6 6 8 8 8 8 8 8 8 r 5 5 d c positive signals. 9 0 0 0 0 0 8 7 7 7 6 6 6 6 i 1 1 n S J J d U U U U U U Q Y Y Y U U U U o 8 n f F F Image stacks were acquired for Cy3 and Cy5 A A A E E E E i 2 G G G G G G D o

r e r p fluorochromes using an Olympus BX61 epifluores - a

e n d cence microscope (Olympus Optical Co., Tokyo, o i e t u

m Japan). Images were acquired every 0.3 µm over the b A i m 9 0 6 4 2 0 0 5 7 2 r 2 3 2 6 8 4 a N t 3 9 whole thickness of sections using the Optigrid™ 4 0 2 3 5 7 1 7 9 9 0 3 0 1 s g 6 7 5 8 3 3 3 3 3 i D 0

1 1 2 0 2 2 t 6 0 5 9 2 2 2 2 6 a d 0 7 7 7 7 7 7

system (QiOptic, Rochester, NY, USA). The l 6 4 7 4 0 0 0 0 5 1

3 3 3 3 3 3 m c , 9 3 2 6 7 7 7 7

4 i 2 t 9 9 9 9 9 9 I h J J J J J J F a F U Q percentage of voxels corresponding to specific Y Y U U U U t p F F F F F F i O E A E E E E A A G D a b r C signals versus general eubacterial signal was a s g

h d . o ’ s l n e computed using the SymbiontJ plug-in implemented

e s a g , t l

) d s n d i

n into ImageJ as described elsewhere (Halary et al., a U n a d l n S a

s a n o a E i

i 2008). e i o

’ n m g n p n d o o n s o o l d a e u d o r r e m o e )

l r l S ( ) o n h a

u o

l a t t o a c e d o C Results and Discussion p a

c C e

c n i e S u n è A

a x w d

n a w p

e e r s N a & e s n r

d

e V N R e M l N u ’

s r w

t

t e

i , f e d i o a & &

s d f S o n n u

r i e e 8 s s n Species identification and molecular f m k n e e e p n l

2 t d d o i o s s a ,

M p i u s i i t n n t

i p n t V e d l a a a

e

n p G R R i t u l l

, h n i

r e ( i Shells were similar for all 8 examined specimens: s s h l s s s

b a c c c e i i i A b C L i l s i i i t e e e

P

a h f f f

r s a , . k n n n N n n i i i t i i i h yellowish, thin and fragile, often displaying a e P u u e n c c c

o o s &

t t D p p p , i ,

d a E Z a a a u t a a e ( p p p t - m m n n u d P P P

i i i B u u periostracal hairs and ranging from 3 to 10 mm in t

a m

l o l l o a a

w m n

i l i i l n n é u n n n n p p n I e a a a a h o h h o r r r n a a e f e n

w length. Shape was rhomboidal, umbos were sub - e e e

O a J P J S P N P S t V V

d t t t

t o e ) r s s s f f f f f f f f f f V C l e n

a f f f f f f f f f f i a a a

U terminal, ventral margins were straight and formed c s

K E E G O & O O O O O O O O O E N S e s c E n an angle with dorsal margins (Fig. 1a). Shells were

n a r e d o u

( highly similar to those of I. macdonaldi , and of the

s q s é é e s s i i three Mediterranean species I. ghisotti , I. simpsoni

l c s i e m e t

e g m m m m l p u m m m m m m 7

and I. modiolaeformis . Only I. modiolaeformis was

n

r S 4 0 0 0 0 m m a

0 0 0 0 0 0 k

m u 7 2 . 6 6 9

r . 4 0 9 3 0 0 n 1 9 collected from cold seeps (mud volcanoes) in the o

4 k 7 7 3 0 a 2 5 8 9 6 9 9 3 2 p h 2 n 1 1 1 ------5

b 3 1 t - - - - a . . 0 0 0 0 0 0 0 6 n n area where our specimens were collected. These may 0 0 0 0 p b 2 2 6 0 9 8 9 0 5 e o e 6 4 6 8 n i 4 5 2 4 5 6 4 . 9 4 7 1 s G e

D therefore be attributed to that species. However,

s 2 s G e

e c shape and small size correspond to the morphogroup é c m n a o ’ n r II defined by Lorion et al. (2010), for which species f

d o

e e e e d s y

n n n n t o d identification using only shell characters was e o o o o r a d d d d d d u t t e d t é b

p b b b p t .

i o o o o o o n n n e e s , , , , considered not reliable. Moreover, the original e

e m e o o o o o o b o e e A d d d d s s u a V V i S S o o o o B a W W W W W W N n h o o o o b description of I. modiolaeformis was very brief and

H t

R s

r r a e W W W W

l l o based on two shell valves (Sturany, 1896, p. 20, figs

S f e

e 8 d d u 2

) 34-35, 37-38; see also Warén, 1991). Therefore the e q s

” s d i n u n s A 5

e current identification should be taken cautiously, and

a n s 8

i d m n 9 i a n A e

1 that further illustrates the “shell handicap” of the c , a

, N é , s m ’ r i é p n e D

A mytilid systematics (Lorion et al., 2010). c u t

o S n e

s q 8 r . l U i m p i 7 u

d We obtained 23 COI mtDNA sequences of 579 d S I S 9 e T

n W

E . 1 i O ,

(

d 8 M base pairs, among which 5 haplotypes were t n C d e & 9 8

. 7 n r o

r 9 s 5 p a 8 o k M f o

1 identified. The most frequent haplotype was repre - s s 9 n 9 n

f

.

a s r ,

1 t 1 e s e p

e

s u k s , a s l u r e sented on all three substrate types sampled and was K l e

i u B d e l n b s o h I s e b G a n a a h p

D C u . i the same as obtained from the specimen analysed o i

1 l d D U n H m o 2 n i I c d d e L

u S i U U s l o h j . m i e i t i u n n previously (Duperron et al., 2008a). The COI a S S l r p A E 2 k

r i g

a o e n o t E E M S / a i n n e e V

t t h o a i l l . .

. . o t m

s mtDNA dataset included overall 39 sequences of o u i d u h p p p p r b b e . s K L M O P & c a b s i s s s s c e f

a a

i s

r a f c a r h w z 579 base pairs, of which 213 were variable. For the s s s s e T T t U U U U U t i a a e t c

w m a a a a s . u . . S S S . S S

p i c . . d d d d B S I I d L a 28S rRNA gene, we obtained 24 sequences of 1004 A A E E I B E E E I I I 82 BIOLOGY OF IDAS SP. MED

base pairs. All were identical to the 28S rRNA sequence of the specimen analysed by Duperron et al. (2008a). The specimen M33 was not included in analyses because its COI mtDNA sequence was missing. The 28S rRNA aligned dataset included 39 sequences and 1029 base pairs, but positions 900 to 905, 429 to 425, 408 to 417 and 386 to 390 were excluded because of uncertainties on their positional homology. The final 28S rRNA alignment included 1003 positions, of which 69 were variable. In the Bayesian tree, specimens from this study formed a well-supported clade (Fig. 2). The mean K2P genetic distance calculated from COI sequences was 0.1% within the clade, which is very close to genetic distances usually reported within other mussels species from deep-sea chemosynthetic environments (Lorion et al., 2010; Won et al., 2003). Both the monophyly and the small genetic distances confirmed that all specimens belong to a single species. Mussels from the Eastern Mediterranean were closely related to I. macdonaldi (Gulf of Mexico), from which COI mtDNA and 28S rRNA differed in mean by 15 substitutions (K2P = 3.3%) and two 1 base pair indels (K2P = 0), respectively. However, only one 28S rRNA sequence was available for I. macdonaldi , which prevented the testing of putative introgression patterns that are well-known in mussels despite high levels of mitochondrial divergence (Faure et al., 2009). The current dataset therefore did not allow testing whether I. macdonaldi and I. modiolaformis are two allopatric populations of a single species, or two recently differentiated species. As no molecular data is available for I. ghisotti, I. simpsoni , I. argenteus , I. cylindricus and I. dalmasi , their relationships with I. mac - donaldi and I. modiolaeformis remain unclear.

Larval shells All 8 examined specimens exhibited similar larval shells (protodissoconch), approximately 500 µm-diameter, displaying a red colour and clearly distinct from the post- Figure 1. Idas sp. Med. (a) Lateral view of a specimen larval shell (dissoconch). These were also observed in representative of samples collected near the Nile deep-sea fan. (b) SEM-examination of two specimens, which showed that and (c) SEM views of the larval shell of the specimen M3. the most distal part of larval shells displayed numerous Pictures b and c correspond to magnification of x190 and x500, concentric lines, whereas the part closest to the umbo respectively. The white arrow indicates the boundary between consisted of a granulated structure, roughly 100 µm- dissoconch and prodissoconch, while black arrows indicate the diameter (Figure 1b and 1c). Similar larval shells were boundary between prodissoconch I and prodissoconch II. Figure 1. Idas sp. Med. (a) Vue latérale d’un spécimen reported in I. argenteus and in other species associated with représentatif des échantillons collectés aux alentours du delta du hydrothermal vents and cold seeps (Lutz et al., 1984; Dean, Nil. (b) et (c) Vue en SEM des coquilles larvaires du spécimen 1993). They are interpreted as evidence of a planktotrophic M3. Les photos C et D correspondent aux grossissements x190 et larval development. The small size of prodissoconch I x500, respectivement. La flèche blanche indique la limite entre results from the absence of yolk reserves in the eggs, while dissoconque et prodissoconque, et les flèches noires indiquent la the well-developed prodissoconch II suggests a long free- limite entre prodissoconque I et prodissoconque II. swimming feeding phase. J. LORION, S. HALARY, J. do NASCIMENTO, S. SAMADI, A. COULOUX, S. DUPERRON 83

Figure 2. Idas sp. Med. Chronogram obtained from the bayesian analysis of the combined dataset (COI mtDNA and 28S rRNA). Substrates of specimens for this study and environments of other species are indicated. Dark grey bars at nodes indicate 95% Highest Posterior Densities (HPD) of divergence estimates in million years (Myr). Values above nodes correspond to posterior probabilities. Figure 2. Idas sp. Med. Chronogramme obtenu à partir de l’analyse bayésienne du jeu de données combinées (COI mtDNA et 28S rRNA). Le substrat utilisé par les spécimens et leur environnement sont indiqués. Les barres gris sombre aux noeuds indiquent les intervalles comprenant les intervalles de confiance bayésiens à 95% (HPD) des temps de divergence en millions d’années (Myr). Les valeurs aux nœuds correspondent aux probabilités postérieures. 84 BIOLOGY OF IDAS SP. MED

Symbiotic associations Analysis of 12 to 35 image stacks per specimen suggested that methylo- and methanotrophs (together referred to as MOX) represented between 23.3 and 32.5% of bacterial volume in gills of specimens collected from exposed crusts (M31, 32, 33) and 17.1 and 22.2% in specimens from the crust half buried in sulfidic sediment (M45 and M46, Table 3, Fig. 3). With respect to the number of stacks analysed, these results are quite reliable. Overall, 3D-FISH indicated that MOX are much less abundant in the new specimens investigated than observed in our earlier study (compare Fig. 3 with Fig. 3 from Duperron et al., 2008a). Due to low FISH signal intensity, the image analysis tool SymbiontJ could analyse successfully only a single image stack from the wood specimen M38-3, yielding 55.6% MOX. This is the first evidence of methanotrophic and methylotrophic Figure 3. Idas sp. Med. Fluorescence in situ hybridization on bacteria occurring in gills of mussels collected from wood a section of two gill filaments from specimen M45. Sulfur- substrate. oxidizing symbionts appear in red and are more abundant than methanotrophs and methylotrophs which both appear in green. Sulfur oxidizers (together referred to as SOX) represent - Figure 3. Idas sp. Med. Hybridation in situ à l’aide de sondes ed 68.2 to 79.7% of bacteria in specimens from carbonates fluorescentes sur une section de deux filaments branchiaux du and 27% in the wood specimen M38-3. The estimation of spécimen M45. Les symbiontes sulfo-oxydants sont en rouge et volume occupied by sulfur oxidizers was rendered difficult sont plus abondants que les symbiontes méthanotrophes et by the low-intensity FISH signals obtained, despite méthylotrophes, qui apparaissent en vert. Barre d’échelle = 25 µm. attempts, and very few (0 to 4) useable image stacks were obtained per specimen. However, proportions of SOX and MOX in specimens M32, 45 and 46, from which both were Very low numbers of bacteria, and in some sections none estimated, sum up to around 100%. Thus the large at all, hybridized with the probes targeting Bacteroidetes proportions of SOX in these three specimens are likely and Gammaproteobacteria G, two groups previously realistic, and clearly visible on FISH images. SOX were documented as possible symbionts in Idas modiolaeformis . also present in the wood specimen. This suggests that they might not be associated with all

Table 3. Idas sp. Med. Number of 3D-FISH stacks analyzed per symbiont type and specimen collected on three distinct substrates (carbonate crust 1 exposed to seawater, carbonate crust 2 half buried in dark and sulfidic sediment, and surface of a wood log). MOX includes methane- and methylotrophs, SOX includes sulfur-oxidizers T1 and T2 ( sensu Duperron et al., 2008b). Values are provided as percentages ± standard deviations. (-) indicates absence of data. Some SOX stacks could not be analysed because of low FISH signal intensities. Tableau 3. Idas sp. Med. Nombre de coupes 3D-FISH analysées pour chaque type de symbionte et pour les trois types de substrats (croûte carbonatée exposée à l’eau de mer, croûte carbonatée a moitié enfoncée dans un sédiment noir riche en sulfides, et surface de la pièce de bois). MOX inclue les bactéries méthylotrophes et méthanotrophes, SOX incluse les bactéries sulfoxydantes des lignées T1 et T2 ( sensu Duperron et al., 2008b). Les résultats sont fournis sous forme de pourcentages ± déviation standard. (-) indique l’absence de données. Certains empilements d’images SOX n’ont pu être analysées en raison d’un niveau d’intensité insuffisant du signal FISH. MOX SOX Substrate specimen Stacks analysed/ Volume occupied Stacks analysed Volume occupied MOX+SOX Total ± std dev /total ± std dev (% of Eub338) (% of Eub338) (% of Eub338)

Crust 1 – M31 19/19 32.5 ± 22.9 0/7 - 32.5 exposed M32 12/12 23.3 ± 4.2 3/9 68.2 ± 38.4 91.5 M33 35/35 31.1 ± 9.6 0-31.1 Crust 2 – M45 22/22 22.2 ± 14.5 2/11 79.7 ± 8.8 101.9 dark and M46 17/17 17.1 ± 17.0 1/12 68.8 85.9 smelly Wood M38-3 1/10 55.6 4/17 27 ± 2.1 82.6 J. LORION, S. HALARY, J. do NASCIMENTO, S. SAMADI, A. COULOUX, S. DUPERRON 85 specimens, and questions their significance as symbionts. example of amphi-Atlantic species or sister species of In particular, the hypothesis that Bacteroidetes could help mussels, and the first among smaller bathymodiolins from digest the wood (Duperron et al., 2008a) is not supported the genera Idas , Adipicola and Benthomodiolus . because these were absent from the specimen collected on I. macdonaldi and I. modiolaeformis belong to a clade wood. which includes mostly species with a strong affinity to All specimens investigated here harboured both MOX sunken wood and bone. As suggested by a statistical and SOX in significant amounts, a combination which reconstruction of ancestral habitats, these two species likely supposedly allows a greater flexibility with regards to the colonized cold seeps from wood-living ancestors (Lorion et use of reduced compounds (Distel et al., 1995). Only al., 2010; see also Figure 1B). The four specimens of I. sulfur-oxidizing symbionts are reported from other Idas - modiolaeformis collected here from wood substrates like mussels documented from the group including Idas sp. deployed meters away from fluid seepages suggest that C, Idas sp. D (Duperron et a, 2008b, based on molecular sunken wood could still act as stepping stones for the analyses and FISH), I. washingtonia (Southward, 2008, dispersal of the I. modiolaeformis / I. macdonaldi group. based on electron micrographs) and I. macdonaldi (Won et However confirming this hypothesis will require further al., 2008, based on molecular analyses). It is thus experimentation. Indeed, without measurements of reduced reasonable to assume that in this clade, methanotrophs and compounds around the experimental area, the possible methylotrophs were only acquired in I. modiolaeformis . influence of neighbouring seep fluids cannot be completely This multiple symbiosis could have helped I. modiolae - ruled out. formis to adapt to the local environments associated with the different substrates. Conclusion Evolutionary history Specimens examined in this study were assigned to Idas The ingroup presented in the tree of Fig. 1B includes only modiolaeformis , although the ambiguity of shell-based small Idas mussels belonging to the L6 clade defined by identification and the lack of molecular data for other Lorion et al. (2010). Applying a COI mtDNA mutation rate mussel species described from the eastern Mediterranean ranging from 1% to 2% per million years indicated that the stress the need for a taxonomic revision of Idas -like clade L6 started diversifying around 15 million years ago mussels. This species and I. macdonaldi , which are closely (95% HPD: 9-20 Myr, figure 1B). The oldest fossil records related and diverged recently, display an amphi-Atlantic of small Idas -like mussels are from late Eocene, which distribution that could be explained by highly dispersive range from about 40 to 34 Myr ago (Kiel & Goedert, 2006 larvae, and by the existence of intermediate habitats. & 2007; Amano et al., 2007; Kiel et al., 2010). As our Examination of symbiont relative abundances using 3D- phylogenetic tree includes only a sub-clade of Idas -like FISH confirmed that multiple symbiont strains occurred in mussels, the current results are consistent with paleontolo - all I. modiolaeformis specimens analysed, and emphasized gical data. Using that 1-2%/Myr COI mutation rate also the dominance of sulfur oxidizers, methylotrophs and allowed estimating that I. macdonaldi and I. modiolae - methanotrophs as symbionts. The later two being absent in formis diverged about 1.84 million years ago (95% HPD: other members of the clade, it can be suggested that they 0.60-3.61 Myr), well after the Messinian salinity crisis were acquired recently, after the divergence with I. mac - (5.96-5.33 Myr ago) (Duggen et al., 2003). We thus cannot donaldi . Gammaproteobacteria G and Bacteroidetes on the conclude whether the Mediterranean basin played a role in other hand were rarely seen, questioning their significance the divergence between I. macdonaldi and I. modiolae - as symbionts. formis or if these two diverged across the Atlantic. Such geographical patterns are already reported in at least two amphi-Atlantic species complexes of cold seep mussels, Acknowledgements namely the “childressi” and “boomerang” species complexes (Olu-Le Roy et al., 2007; Génio et al., 2008). It This study was supported by the European programs HER - was suggested that these complexes are the result of a MES (CT511234), DIWOOD and CHEMECO, and the progressive colonization using intermediate habitats as French Ministry of Education and Research through a PhD stepping stones (Olu-Le Roy et al., 2007). Such a scenario grant and ANR Deep Oases. Authors are grateful to pilots is plausible for I. modiolaeformis , given its larval and crew of RV Pourquoi Pas?, RV L’Atalante, RV Meteor, development, its wide depth distribution (1150 to 3000 m, manned submersible Nautile, ROV Victor , and ROV Quest. table 1) and the occurrence of cold seep areas in the Eastern We thank chief scientists J. P. Foucher (NAUTINIL), A. Atlantic, such as for example in the Gulf of Cadiz. I. mac - Boetius (BIONIL, and wood samples) and C. Pierre donaldi and I. modiolaeformis thus represent the third (MEDECO). We also thank O. Gros and L. Maurin for their 86 BIOLOGY OF IDAS SP. MED help with sample processing onboard, A. Waren, E. Strong Sulphur-oxidizing extracellular bacteria in the gills of and A. Eschner who provided us with holotype pictures, Y. Mytilidae associated with wood falls. FEMS Microbiology Kantor for his help during SEM observations and T. Comtet Ecology, 63 : 338-349. for helpful comments. SH’s grant was funded by the EU Duperron S., Lorion J., Samadi S., Gros O. & Gaill F. 2009. project HERMIONE. Symbioses between deep-sea mussels (Mytilidae: Bathymodiolinae) and chemosynthetic bacteria: diversity, function and evolution. Comptes rendus Biologies, 332 : 298- References 310. Faure B., Jollivet D., Tanguy A., Bonhomme F. & Bierne N. Amann R., Binder B.J., Olson R.J., Chisholm S.W., Devereux 2009. Speciation in the deep sea: multi-locus analysis of R. & Stahl D.A. 1990. Combination of 16S rRNA-targeted divergence and gene flow between two hybridizing species of oligonucleotide probes with flow cytometry for analysing hydrothermal vent mussels. Plos One, 4: e6485. mixed microbial populations. Applied and Environmental Folmer O., Black M., Hoeh W.R., Lutz R.A.& Vrijenhoek R.C. Microbiology, 56 : 1919-1925. 1994. DNA primers for amplification of mitochondrial Amano K., Little C.T.S. & Inoue K. 2007. A new Miocene cytochrome C oxidase subunit I from metazoan invertebrates. whale-fall community from Japan. Palaeogeography, Molecular Marine Biology and Biotechnology, 3: 294-299. Palaeoclimatology, Palaeoecology, 247 : 236-242. Fujita Y., Matsumoto H., Fujiwara Y., Hashimoto J., Sergey Cavanaugh C., McKiness Z., Newton I. & Stewart F. 2006. V., Galkin S.V., Ueshima R. & Miyazaki J.-I. 2009. Marine chemosynthetic symbioses. In: The Prokaryotes ed. Phylogenetic relationships of deep-sea Bathymodiolus mussels Springer, New-York, pp. 475-507. to their mytilid relatives from sunken whale carcasses and Cavanaugh C.M., Levering P.R., Maki J.S., Mitchell R. & wood. Venus, 64 : 123-134. Lidstrom M.E. 1987. Symbiosis of methylotrophic bacteria Génio L., Johnson S.B., Vrijenhoek R.C., Cunha M.R., Tyler and deep-sea mussels. Nature, 325 : 346-348. P.A., Kiel S. & Little C.T.S. 2008. New record of Dean H.K. 1993. A population study of the bivalve Idas Bathymodiolus mauritanicus Cosel 2002 from the Gulf of argenteus Jeffreys, 1876, (Bivalvia: Mytilidae) recovered from Cadiz (NE Atlantic) mud volcanoes. Journal of Shellfish a submerged wood block in the deep North Atlantic Ocean. Research, 27 : 53-61. Malacologia, 35 : 21-41. Gros O. & Gaill F. 2007. Extracellular bacterial association in Deming J.W., Reysenbach A.L., Macko S.A. & Smith C.R. gills of «wood mussels». Cahiers de Biologie Marine, 48 : 103- 1997. Evidence for the microbial basis of a chemoautotrophic 109. invertebrate community at a whale fall on the deep seafloor: Halary S., Riou, V., Gaill F., Boudier T. & Duperron S. 2008. Bone-colonizing bacteria and invertebrate endosymbionts. 3D FISH for the quantification of methane- and sulphur- Microscopy Research and Technique, 37 : 162-170. oxidizing endosymbionts in bacteriocytes of the hydrothermal Distel D.L., Baco A.R., Chuang E., Morrill W., Cavanaugh vent mussel Bathymodiolus azoricus . ISME Journal, 2: 284- C.M., & Smith C.R. 2000. Do mussels take wooden steps to 292. deep-sea vents? Nature, 403 : 725–726. Iwasaki H., Kyuno A., Shintaku M., Fujita Y., Fujiwara Y., Distel D.L., Lee H.K. & Cavanaugh C.M. 1995. Intracellular Fujikura K., Hashimoto J., Martins L., Gebruk A. & coexistence of methano- and thioautotrophic bacteria in a Miyazaki J.-I. 2006. Evolutionary relationships of deep-sea hydrothermal vent mussel. Proceedings of the National mussels inferred by mitochondrial DNA sequences. Marine Academy of Sciences of the United States of America, 92 : Biology, 5: 1-12. 9598-9602. Kiel S., Campbell K.A. & Gaillard C. 2010. New and little Drummond A.J. & Rambaut A. 2007. BEAST: Bayesian known mollusks from ancient chemosynthetic environments. evolutionary analysis by sampling trees. BMC Evolutionary Zootaxa, 2390 : 26-48. Biology, 7: 1-8. Kiel S. & Goedert J.L. 2007. New mollusks associated with Duggen S., Hoernle K., van den Bogaard P., Rupke L. & biogenic substrates in Cenozoic deep-water sediments of Phipps Morgan J. 2003. Deep roots of the Messinian salinity Washington State. Acta Palaeontologica Polonica, 52 : 41-52. crisis. Nature, 422 : 602-606. Kiel S. & Goedert J.L. 2006. A wood-fall association from late Duperron S., Bergin C., Zielinski F., Blazejak A., Pernthaler Eocene deep-water sediments of Washington state, USA. A., McKiness Z.P., DeChaine E., Cavanaugh C.M. & Palaios, 21 : 548-556. Dubilier N. 2006. A dual symbiosis shared by two mussel Kyuno A., Shintaku M., Fujita Y., Matsumoto H., Utsumi M., species, Bathymodiolus azoricus and Bathymodiolus puteo - Wanatabe H., Fujiwara Y. & Miyazaki J.-I. 2009. Dispersal serpentis (Bivalvia: Mytilidae), from hydrothermal vents along and differentiation of deep-sea mussels of the genus the northern Mid-Atlantic Ridge. Environmental Microbiology, Bathymodiolus (Mytilidae, Bathymodiolinae). Journal of 8: 1441-1447. Marine Biology, ID 625672, 15 pp. doi: 10.1155/2009/625672. Duperron S., Halary S., Lorion J., Sibuet M. & Gaill F. 2008a. Lorion J., Buge B., Cruaud C. & Samadi S. 2010. New insights Unexpected co-occurrence of six bacterial symbionts in the into diversity and evolution of deep-sea mussels (Bivalvia: gills of the cold seep mussel Idas sp. (Bivalvia: Mytilidae). Mytilidae). Molecular Phylogenetics and Evolution, 57 : 71-83. Environmental Microbiology 10 : 433-445. Lorion J., Duperron S., Gros O., Cruaud C. & Samadi S. 2009. Duperron S., Laurent M.C.Z., Gaill F. & Gros O. 2008b. Several deep-sea mussels and their associated symbionts are J. LORION, S. HALARY, J. do NASCIMENTO, S. SAMADI, A. COULOUX, S. DUPERRON 87

able to live both on wood and on whale falls. Proceedings of the gills of mussels of the genera Adipicola and Idas (Bivalvia: the Royal Society B: Biological Sciences, 276 : 177-185. Mytilidae). Journal of Shellfish Research, 27 : 139-146. Lutz R.A., Jablonski D. & Turner R.D. 1984. Larval Sturany R. 1896. Zoologische Ergebnisse. VII. Mollusken I. development and dispersal at deep-sea hydrothermal vents. (Prosobranchier und Opisthobranchier; Scaphopoden; Science, 226 : 1451-1454. Lamellibranchier.) gesammelt von S.M. Schiff “Pola” 1890- Miyazaki J.-I., Martins L.d.O., Fujita Y., Matsumoto H. & 1894. Denkschriften der Kaiserlichten Akademie der Fujiwara Y. 2010. Evolutionary process of deep-sea Wissenschaften, Wien. Mathematische-Naturwissenschaft- Bathymodiolus mussels. PLoS ONE, 5: e10363. lischen, 63 : 1-36, pl. 1-2. Olu-Le Roy K., von Cosel R., Hourdez S., Carney S.L. & Tamura K., Dudley J., Nei M. & Kumar S. 2007. MEGA4: Jollivet D. 2007. Amphi-Atlantic cold-seep Bathymodiolus Molecular Evolutionary Genetic Analysis (MEGA) software species complexes across the equatorial belt. Deep Sea version 4.0. Molecular Biology and Evolution, 24 : 1596-1599. Research Part I: Oceanographic Research Papers, 54 : 1890- Tyler P., Young C., Dolan E., Arellano S., Brooke S. &Baker 1911. M. 2007. Gametogenic periodicity in the chemosynthetic cold- Pelorce J. & Poutiers J.-M. 2009. Une nouvelle espèce de seep mussel “Bathymodiolus” childressi . Marine Biology, 150 : Bathymodiolinae (, Bivalvia, Mytilidae) associ ée a 829-840. des os de baleine coules en Méditerranée . Zoosystema , 31 : Tyler P.A., Marsh L., Baco-Taylor A. & Smith C.R. 2009. 975-985. Protandric hermaphroditism in the whale-fall bivalve mollusc Posada D. & Crandall K.A. 1998. Modeltest: testing the model Idas washingtonia . Deep Sea Research Part II: Topical Studies of DNA substitution. Bioinformatics, 14 : 817-818. in Oceanography, 56 : 1689-1699. Qu L.H. 1986. Structuration et évolution de l’A.R.N ribosomique Warén A. 1991. New and little known Mollusca from Iceland and 28S chez les Eucaryotes: étude systématique de la région 5’ Scandinavia. Sarsia, 76 : 53-124 terminale. Paul Sabatier, Toulouse, 230 pp. Won Y.-J., Jones W.J. & Vrijenhoek R.C. 2008. Absence of Samadi S., Quemere E., Lorion J., Tillier A., von Cosel R., cospeciation between deep-sea mytilids and their thiotrophic Lopez P., Cruaud C., Couloux A. & Boisselier-Dubayle M.- endosymbionts. Journal of Shellfish Research, 27 : 129-138. C. 2007. Molecular phylogeny in mytilids supports the Won Y., Young C.R., Lutz R.A. & Vrijenhoek R.C. 2003. wooden steps to deep-sea vents hypothesis. Comptes Rendus Dispersal barriers and isolation among deep-sea mussel Biologies, 330 : 446-456. populations (Mytilidae: Bathymodiolus ) from eastern Pacific Southward E.C. 2008. The morphology of bacterial symbioses in hydrothermal vents. Molecular Ecology, 12 : 169-184.