Portal de Periódicos da CAPES

Diversity of vibrios in the haemolymph of the spider crab Maja brachydactyla

2010; Oxford University Press; Volume: 109; Issue: 3 Linguagem: Inglês

10.1111/j.1365-2672.2010.04718.x

1365-2672

Bruno Gómez‐Gil, Ana Roque, Beatriz Lacuesta, Guiomar Rotllant,

Invertebrate Immune Response Mechanisms

Journal of Applied MicrobiologyVolume 109, Issue 3 p. 918-926 ORIGINAL ARTICLEFree Access Diversity of vibrios in the haemolymph of the spider crab Maja brachydactyla B. Gomez-Gil, B. Gomez-Gil CIAD, A.C., Mazatlán Unit for Aquaculture and Environmental Management, Mazatlán, MexicoSearch for more papers by this authorA. Roque, A. Roque Institut de Recerca i Tecnologia Agroalimentàries-Centre d'Aqüicultura, Sant Carles de la Rápita, Tarragona, SpainSearch for more papers by this authorB. Lacuesta, B. Lacuesta Institut de Recerca i Tecnologia Agroalimentàries-Centre d'Aqüicultura, Sant Carles de la Rápita, Tarragona, SpainSearch for more papers by this authorG. Rotllant, G. Rotllant Institut de Recerca i Tecnologia Agroalimentàries-Centre d'Aqüicultura, Sant Carles de la Rápita, Tarragona, SpainSearch for more papers by this author B. Gomez-Gil, B. Gomez-Gil CIAD, A.C., Mazatlán Unit for Aquaculture and Environmental Management, Mazatlán, MexicoSearch for more papers by this authorA. Roque, A. Roque Institut de Recerca i Tecnologia Agroalimentàries-Centre d'Aqüicultura, Sant Carles de la Rápita, Tarragona, SpainSearch for more papers by this authorB. Lacuesta, B. Lacuesta Institut de Recerca i Tecnologia Agroalimentàries-Centre d'Aqüicultura, Sant Carles de la Rápita, Tarragona, SpainSearch for more papers by this authorG. Rotllant, G. Rotllant Institut de Recerca i Tecnologia Agroalimentàries-Centre d'Aqüicultura, Sant Carles de la Rápita, Tarragona, SpainSearch for more papers by this author First published: 16 August 2010 https://doi.org/10.1111/j.1365-2672.2010.04718.xCitations: 18 Bruno Gomez-Gil, CIAD, A.C. Mazatlán Unit for Aquaculture and Environmental Management, AP 711 Mazatlán, Sinaloa, México 82000. E-mail: bruno@ciad.mx AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract Aims: The aim of this study was to characterise and identify vibrios isolated from the haemolymph of apparently healthy adult spider crabs (Maja brachydactyla) wild-caught in the Spanish localities of Galician coast and in the Canary Islands and also from captive animals held at IRTA's facilities in the Ebro Delta of Catalonia, north-west Spanish Mediterranean coast. Methods and Results: A total of 277 bacterial isolates were obtained, and of these, 171 were characterised with rep-PCR, resulting electrophoretic bands were analysed and clusters formed. Identification of representative strains of each cluster was made by sequencing the 16S rRNA. Samples from animals caught in Galicia and captive at IRTA (around 15–18°C) rendered mostly species belonging to the Splendidus clade (72·2 and 76·6% respectively), commonly found in cold waters (below 20°C). Higher species diversity was found in the haemolymph of the captive animals. In the warmer Canary Islands waters (around 21°C), the diversity of vibrios is dominated by three clades, Harveyi (Vibrio core group, 39·3%), Orientalis (23·2%) and Splendidus (21·4%) with a species diversity that equals that of the colder captive animals. Conclusions: Differences in the vibrios populations were found in the haemolymph extracted from animals collected from the three localities. Potential new species were found, and their description is under way. Significance and Impact of Study: As with other invertebrates, spider crabs also contain a diverse population of vibrios. These findings should help researchers to diagnose when a crab is infected. Introduction The European spider crab Maja brachydactyla (Crustacea: Decapoda: Majidae) constitutes an important fishery resource on the coast of Galicia (NW Spain) extracting more than 90% of the annual recruitment; consequently, this species has been classified as overexploited (Freire et al. 2002). Captures of spider crab are not enough to cover the market demand, and aquaculture under controlled intensive conditions might be a solution to the problem. From the aquaculture perspective, several characteristics of the species, such as its high fecundity rate, short embryonic and larval development and rapid growth (Gonzalez-Gurriaran et al. 1995; Rotllant et al. 2007), make this spider crab species a promising candidate for aquaculture. Currently, research on this species is directed towards culture development to diversify aquaculture (Andres et al. 2007, 2008; Rotllant et al. 2008; Guerao and Rotllant 2009). Bacteria of the genus Vibrio are ubiquitous in marine and estuarine aquatic ecosystems where M. brachydactyla occurs naturally (Thompson et al. 2004; Urakawa and Rivera 2006). Many of these species have also been associated with bacterial infections in cultured crustaceans (Hameed et al. 1996; Roque et al. 2000; Jayasree et al. 2006; Soto-Rodriguez et al. 2006) and are generally agreed to be opportunistic pathogens causing disease when the immune system of crustaceans is compromised (Lightner and McVey 1993). However, the presence of bacteria in the haemolymph of Crustacea, previously interpreted as a disease indicator (Lightner 1975; Lightner and Lewis 1975), is currently viewed more as a natural process (Gomez-Gil et al. 1998). Bacteria have been isolated from the haemolymph of apparently healthy Pacific white shrimp (Penaeus vannamei; Gomez-Gil et al. 1998), American lobster (Homarus americanus; Cornick and Stewart 1966; Bartlett et al. 2008), blue crab (Callinectes sapidus; Haskell et al. 1975; Tubiash et al. 1975; Davis and Sizemore 1982; Welsh and Sizemore 1985), red swamp crawfish (Procambarus clarkii; Scott and Thune 1986), diverse crab species (Plagusia dentipes, Schizophrys aspera, Atergatis spp., and Aniculus aniculus; Ueda et al. 1993), and freshwater prawns (Macrobrachium rosenbergii; Brady and Lasso-de-la-Vega 1992). Bacterial density often does not exceed 103−4 CFU ml−1 of haemolymph (Tubiash et al. 1975; Welsh and Sizemore 1985; Ueda et al. 1993; Gomez-Gil et al. 1998) in 88·3% of the analysed organisms (Bartlett et al. 2008). These studies clearly reinforce the need to know the bacterial populations in the haemolymph of healthy crustaceans to correctly interpret diagnostic results. Therefore, the aim of this study is to analyse and compare the diversity of Vibrionaceae present in the haemolymph of spider crab recently fished in Galicia and in the Canary Islands with the Vibrionaceae present in the haemolymph of spider crab kept in an experimental culture system for over a year. Material and methods Sample collection and processing Three batches of ten adult spider crab females were selected with an average wet weight of 1122 ± 195 g and a carapace length of 159 ± 10 mm and sampled individually. All crabs appeared to be healthy and were devoid of obvious signs of disease. The first batch of crabs was collected in the Ría A Coruña, Galician coast (43°21′N, 8°22′W, 20 m depth) on the 6th of December 2005, and water conditions were as follows: temperature 15°C, salinity 37·25 ppt; pH 8·24. These crabs were caught by trawl or gillnet and transported to the nearest port in tanks with cold water. Immediately, they were shipped to Centre for Aquaculture of the Research Institute and Agro-alimentary Technologies (IRTA, Sant Carles de la Ràpita, Tarragona, Spain. 40°37′40″N, 0°39′36″E) in high humidity containers at a temperature below 8°C for 24–36 h. The haemolymph was sampled immediately after arrival. The second batch of crabs was collected also at the Ría de A Coruña but 1 year earlier and was kept at the rearing facilities at IRTA. The water conditions during the maintenance of the animals and at sampling were as follows: temperature 18 ± 1°C, salinity 35 ± 1 ppt and pH 7·9. The haemolymph of this second batch was sampled on the 21st of November 2005. The third batch was collected off from the Melenara beach, Taliarte, Gran Canaria, Canary Islands (27°59′N, 15°22′W), on 30th of November 2006, and water conditions were as follows: temperature 21°C, salinity 37·5 ppt and pH 8·20. All animals were anaesthetised on ice until they did not respond to external stimuli, and the base of the periopod (coxa) disinfected with 70% ethanol before sampling to assure the sample was not contaminated with surface bacteria. From each organism, approx. 200 μl of haemolymph was drawn with a 1-mL syringe attached to a needle of 25Gx5/8″. One hundred microlitres of undiluted haemolymph were spread-plated onto thiosulfate citrate bile salt agar (Scharlau, Barcelona, Spain) and incubated at 28°C for 24 h. Ten colonies from each plate were randomly taken for further analysis. These isolates were purified in tryptic soy agar (TSA) + 2% NaCl (Scharlau) and cryopreserved at −80°C in TSA + 2% NaCl + 15% glycerol (Gherna et al. 1994) until use. Molecular analysis Total DNA was extracted from each isolate using the Wizard Genomic DNA purification kit (Promega, Madison, WI) following the instructions of the manufacturer, and final DNA concentration was spectrophotometrically adjusted to 50 ng μl−1 with sterile distilled water. DNA fingerprinting of all strains was performed using the (GTG)5 primer (Doll et al. 1993), and the PCR amplification was performed as previously described (Cabanillas-Beltran et al. 2006) with the Invitrogen Taq DNA polymerase recombinant enzyme (Invitrogen Corp., Carlsbad, CA). PCR products were electrophoresed in a 15 × 15 cm, 1·2% agarose gel for 70 min at 90 V at room temperature using as DNA markers the 1 kb Plus DNA Ladder (Invitrogen Corp.). Gels were stained a posteriori with ethidium bromide at room temperature for 15 min and then rinsed with tap water. Gels were visualised, and the image captured as TIFF files in a gel documentation system (UVP, Upland, CA). Images were analysed with the GelCompar II software ver. 4.5 (Applied-maths, Sint-Martens-Latem, Belgium). Band patterns obtained from the isolates were used to calculate a similarity matrix (Dice coefficient), which was used to construct a dendrogram (Ward algorithm). Representative strains from each cluster and nonclustering strains were identified by sequencing the 16S rRNA gene. The 16S rRNA was amplified with universal primers (V16S-9F, AGA GTT TGA TCA TGG CTC AG; V16S-1491R, AGC GCT ACC TTG TTA CGA CTT) to obtain an almost complete sequence (>1450 bp). The amplification programme used was one cycle at 94°C for 2 min, 35 cycles at 94°C for 1 min, 56°C for 1 min, 72°C for 1 min and one final cycle at 72°C for 5 min. PCR purification and sequencing of the leading DNA strand and sometimes the lagging strand were carried out at Macrogen (Seoul, Korea). For a few strains that were not clearly identified with the 16S rRNA gene, the toxR gene was amplified as described earlier (Pascual et al. 2009). All sequences were deposited at Genbank with accession numbers GQ386820 through GQ386822 and GQ454927 through GC455018 for 16S rRNA sequences, and GQ455019 through GQ455024 for toxR sequences. The dendrogram with the 16S sequences was performed with the first third (approx. 500 bp) of the gene; the evolutionary history was inferred using the neighbour-joining method, and the evolutionary distances calculated with the Juker-Cantor method employing the mega ver. 4 software (Kumar et al. 2001). The radial dendrogram was drawn with the Dendroscope ver. 2.2 program (Huson et al. 2007). To differentiate between closely related species, identified by 16S, API 20E phenotyping was performed as recommended by the manufacturer but with the addition of 2% NaCl to the inoculation bacterial suspension. For the case of species belonging to the Vibrio splendidus complex, the differentiation table proposed by Beaz-Hidalgo (2009) was followed. Results A total of 277 bacterial isolates were obtained from the haemolymph of wild-caught spider crab samples taken from the Galician coast (162 isolates), Canary Islands (59 isolates) and from crabs kept in captivity at IRTA (56 isolates). Of these isolates, 175 were randomly selected for fingerprinting with rep-PCR (Table S1), and from 171 isolates, a band pattern was obtained. The average number of bands obtained was 14·12 (20–6, n = 152), ranging from 10 700 bp to 65·1 bp (average 1436·52 bp, standard deviation 1014·5). The 16S rRNA gene of 103 isolates was sequenced partially (first 445–955 bp, 82 isolates), and of those strains was the first segment that does not clearly differentiate between closely related species (24 isolates); the almost complete (>1371 bp) sequence was obtained. The rep-PCR analysis (Fig. 1, optimisation 0·97%, band position tolerance 0·7%) showed the formation of 31 clusters composed of one species each. Most of these clusters as well as unclustered strains formed together groups of the same phylogenetic clade (Fig. 1, grey boxes). Six clonal groups were found among all the isolates analysed, and these were defined as those having above 99% band similarity with the parameters employed. One of these clonal groups comprehended four strains isolated in the Canary Islands, whereas the other five groups were composed of 2–3 isolates each (Fig. S1). Therefore, out of the 171 isolates analysed by rep-PCR, 162 different strains were found. Figure 1Open in figure viewerPowerPoint Dendrogram produced from rep-PCR band patterns obtained from vibrios isolated from the haemolymph of Maja brachydactyla from wild-caught animals in the Galician coast and Canary Islands and held in captivity at IRTA in the Ebro delta. Strains with asterisks have 16S rRNA sequences. Nonvibrio strains and members of the Aliivibrio, Anguillarum and Nigripulchritudo were omitted. H = Harveyi clade, O = Orientalis clade, P = Photobacterium clade, S = Splendidus clade. From the fingerprinting analysis supported with the 16S sequence results (Fig. S2), 37 operating taxonomic units (OTUs) were obtained. Twenty-one OTUs could be identified to species level, 15 to genus level and one was not identified. Of the 37 identified OTUs, five did not belong to the Vibrionaceae (one isolate per each OTU). Of the Vibrionaceae (32 OTUs), one OTU was identified as Aliivibrio, six as Photobacterium and 25 as Vibrio. The OTUs found were assigned to clades as defined by Sawabe et al. (2007) based on a Multilocus Sequence Analysis (MLSA) study. The Vibrionaceae clades found comprehended 165 isolates distributed in eight clades (Fig. 2). One hundred isolates were assigned to the Splendidus clade (60·6%), 22 to the Harveyi clade (13·3%), 20 to the Photobacterium clade (12·1%) and 17 to the Orientalis clade (10·3%). The rest of the isolates (6) belonged to three minor clades, Aliivibrio (3), Anguillarum (1), Nigripulchritudo (1). From one isolate (Mj73), neither a reliable 16S sequence nor a rep-PCR clustering could be obtained, but even so the blastn of the sequence finds Vibrio as the closest genus, and therefore it has been considered as a vibrio and coded as Vib.OTU2. Figure 2Open in figure viewerPowerPoint Clades found in the haemolymph of Maja brachydactyla from wild animals caught in two different geographic locations (Galician coast and Canary Islands) and animals kept in captivity at IRTA (Ebro delta). (, Ebro; , Galicia; , Canary islands). Splendidus clade Isolates identified as belonging to the Splendidus clade were spread in 20 clusters, 14 of these clusters were grouped with other Splendidus isolates forming five Splendidus clade groups (Fig. 1). In total, 14 OTUs were obtained; eight of these were identified to species (Table S1) and six only to genus level as the 16S sequences were below 98·9% with known species of this clade. Thirty-six isolates were obtained from haemolymph samples extracted from animals at IRTA (Ebro river), 52 from the Galician coast batch and 12 from the Canary Islands batch. Only relevant information will be described of the clusters and groups, but all the information is presented in Table S1, Figs 1 and S1–S4. The Splendidus group 1 (Fig. 1) is composed of a cluster of five strains (Fig. 1, S1) closely related to Vibrio gallaecicus but with a 16S sequence similarity of 98·5% (GQ45011, 1453 bp) for strain Mj71 and 98·9% (GQ454967, 1471 bp) for strain Mj206. This OTU was coded Vib.B, and the strains were isolated from IRTA samples at the Ebro delta. Two other strains were grouped with this cluster, but they were identified as other species of this clade. The Splendidus group 2 is composed of three clusters (Fig. 1; S3, S4 and S5) and four unclustered strains. Cluster S4 is composed of three similar strains (77·8–88·9% rep similarity) identified as Vibrio crassostreae, two isolated from the Galician coast and the other from IRTA. Cluster S5 (six strains) were identified as Vibrio cyclitrophicus, four from Galicia and two from the Ebro. Unclustered strain Mj134 (Vib.4) had a 16S sequence similarity of 97·7% (GQ454937) with V. splendidus but only 519 bp were sequenced. Splendidus group 3 was the largest found with 30 strains and four clusters (Fig. 1; S7, S8, S9 and S10). Cluster S7 contained 16 strains, all but one (Mj249) identified as V. splendidus and all of these isolated from the Galician coast samples. The other strain was identified as Vibrio fortis (Mj249, GQ454988, 99·4% 16 similarity, 532 bp), a member also of the Splendidus clade, and isolated from the haemolymph of Canary Islands specimens. Two strains that did not belong to the Splendidus clade were grouped here, Mj265 (Vib.I, GQ454998) belonging to the Orientalis clade and Mj220 to the Photobacterium clade (Ph.B, GQ454978). Splendidus group 4 is composed of 16 strains and four clusters (Fig. 1; S13, S14, S15, and S16). Cluster S14 is composed of three strains identified as Vibrio sp. (Vib.A), strain Mj201 has a sequence similarity of 98·7% (GQ454964) with V. gallaecicus and the three have a rep-PCR similarity between 86·7 and 64·5%. In this Splendidus group, three strains not belonging to this clade were grouped here; Mj233 (identified as Vib.I, Orientalis clade), Mj261 (Vibrio tubiashii, Orientalis clade, GQ454996) and Mj279 (Ph.1, Photobacterium clade, GQ455005). Splendidus group 5 has five isolates, four strains, and one cluster (Fig. 1, S19). Cluster S19 is composed of two isolates with a rep-PCR similarity of 100%, thus considered clones and identified as Vibrio tasmaniensis, both isolated from IRTA samples. Clusters S2, S6, S11, S12, S17, S18 and S20 did not get grouped together (Fig. 1). Cluster S11 is composed of two clonal isolates (rep-PCR 100%) identified as Vibrio crassostreae. Cluster S20 with four strains, three from the Galician coast and one from IRTA, was identified as Vibrio sp. (Vib.3). Mj190 has a 16S sequence similarity of 97·1% (GQ454961) with V. splendidus with only 514 bp (first third of the 16S gene). In total, three OTUs (Vib.A, Vib.B and Vib.E) were found to be similar to V. gallaecicus by their 16S sequences and also by the toxR sequences. By the 16S sequence analysis, all species of the Splendidus clade and these three OTUs appear together with a bootstrap value support of 60%, but Vib.B is closer to V. gallaecicus than the other two, which appear closer together. The toxR analysis also places these OTUs independently but closer to V. gallaecicus than to the other species. The same pattern is also observed when both sequences (16S and toxR) are concatenated (Fig. S4). These analyses lead us to believe that these three OTUs could represent new species within the Splendidus clade and might even become, with V. gallaecicus, a new clade. Harveyi clade Of the 22 isolates belonging to the Harveyi clade, 15 clustered in four clusters and seven remained unclustered. All were isolated from the Canary Islands crab samples. Clusters H1 and H2 (Fig. 1, H1 and H2) were identified as the newly described Vibrio atypicus (Wang et al. 2009) because of the high 16S sequence similarity of strain Mj273 (99·3%, GQ455002, 1451 bp, cluster H1), of strain Mj275 (99·1%, GQ455003, 1450 bp, cluster H2) and of the unclustered strain Mj268 (99·1%, GQ455000, 1462 bp). Cluster H3 is composed of three strains identified as Vibrio rotiferianus; the other four strains identified as V. rotiferianus did not cluster with any other and were identified by their high (>99·4%) 16S sequence similarities. Cluster H4 is composed of eight isolates identified as Vibrio campbellii, and isolates Mj236, Mj237 and Mj245 are clones (rep-PCR similarity of 100%). Photobacterium clade Twenty isolates, 19 strains and three clusters were found belonging to the Photobacterium clade (Fig. 1; P1, P2, and P3). Clusters P1 (four strains, Ph.A) and P2 (six strains, Ph.A) from the Galician coast samples were identified only to the genus Photobacterium because the 16S sequence similarities of these strains ranged between 97·4% (Mj110, GQ454928, 1475 bp) and 96·6% (Mj123, GQ454932, 454 bp) to Photobacterium iliopiscarium and Photobacterium aplysiae, respectively. Differences found could be explained because size (bp) differences were obtained in the 16S sequences. Cluster P3 was composed of three isolates belonging to Canary Islands samples, two of which were clones (Mj241 and Mj255); they were identified as Photobacterium sp. (Ph.1). Another strain belonging to the three batches (Mj279, Mj212, and Mj220) did not cluster with the other Ph.1 strains, all had similar 16S sequences to Ph.1 strains and differ between 95·8 and 96·8% from the closest phylogenetic neighbour Photobacterium frigidiphilum. Strain Mj214 (Ph.C) from IRTA samples did not cluster with any other, and its 16S sequence (GQ454974, 1454 bp) was 97·6% similar to strain Mj110 from the Galician coast sample and 96·9% to P. iliopiscarium. Strain Mj232 (Ph.2) from animals caught in the Canary Islands also did not cluster; its 16S sequence (GQ454982, 944 bp) was 99·9% similar to the recently proposed Photobacterium jeanii (Chimetto et al. IJSEM). Strain Mj246 (Ph.3) also from Canary Islands batch, another unclustered strain, was 97·6% and 95·4% similar to P. jeanii and Photobacterium leiognathi (16S GQ889359, 943 bp). Orientalis clade Seventeen isolates/strains were identified as belonging to the Orientalis clade, 12 strains clustered in four clusters (Fig. 1; O1, O2, O3 and O4) and the other five remained unclustered. Three strains from Canary Islands samples were identified as Vibrio hepatarius (Fig. 1, O2 and the unclustered Mj270) because of their high 16S sequence similarities; 98·8% of Mj250 (GQ889361, 936 bp) and 99·3% of Mj270 (GQ455001, 1451 bp). Five strains from Canary Islands batches were identified as Vibrio tubiashii, three clustered together (Fig. 1, O4) and two did not. Strain Mj248 from cluster O4 belonging to the Canary Islands samples had a 16S sequence similarity of 98·9% with V. tubiashii, and the API 20E profile differed only in the utilisation of amygdalin (negative for V. tubiashii). Unclustered strains Mj261 and Mj274 also from Canary Islands samples had a 16S similarity of 99·5% (GQ454996, 947 bp) and 99·6% (GQ889362, 908 bp) with V. tubiashii, respectively. The remaining nine strains from the Canary Islands and IRTA samples formed two clusters (Fig. 1. O1 and O3) and two strains remained unclustered; the 16S sequences of three strains of this group yielded similarity values between 98·1 and 98·8% to V. tubiashii and V. hepatarius. The 16S differences between them were <99·8% (GQ889354, GQ454998 and GQ454999). Other clades Three other clades were found, Aliivibrio with three strains, and Anguillarum and Nigripulchritudo with one each. The three strains from the Galician coast samples assigned to the Aliivibrio clade were identified as Aliivibrio fischeri as they had 16S similarity values between 99·4 and 100% (Mj109 GQ454927, 535 bp; Mj165 GQ454950, 516 bp; Mj230 GQ454981, 948 bp) with the type strain of the species. Strain Mj76 (Vib.J) from samples taken from animals kept at IRTA had a 16S sequence similarity with Vibrio aestuarianus of 97·3% (GQ455012, 1447 bp); this species belongs to the Anguillarum clade. The strain belonging to the Nigripulchritudo clade (Mj264) was isolated from the haemolymph of animals caught in Canary Islands and was identified as Vibrio nigripulchritudo because the 16S similarity with this species was 99·3% (GQ454997, 952 bp). Simpson's reciprocal (1/D) indexes showed a higher diversity of vibrios from the crab's haemolymph extracted from Canary Islands animals (14·42), followed by those belonging to crabs kept in captivity at IRTA (13·35), whereas the Galician coast samples showed a much lower value (6·34). The diversity index considering all localities was 15·83. A rarefaction analysis (Fig. S2) permitted to observe that for the Galician coast samples, a plateau was almost reached, whereas from the samples of the other two batches, still more OTUs could be found, especially in the haemolymph of the animals maintained in captivity. When all the localities were considered, it was evident that still more OTUs could be found because a plateau was not reached. Discussion This is the first report on the bacterial flora associated with M. brachydactyla. Until recently, there had been no interest in the culture of this species or similar ones, but with a demand that cannot be met by the fishery because of overfishing (Freire et al. 2002) and the world need to diversify crustacean cultures (mainly shrimps) because of overproduction and diseases (Lem 2006), the need for culture became eminent. It is assumed that, similarly to other species, disease problems will arise, and knowing the bacterial diversity associated with apparently healthy spider crabs will contribute towards a more informed diagnosis. The animals caught in the Galician coast (wild and captive crabs) share more bacterial species than with those belonging to the southern Canary Islands, where the species of vibrios found are generally more tropical–temperate (e.g. Vibrio alginolyticus, Vibrio harveyi) (Maeda et al. 2003). It has been found that the Vibrio species' composition changes as the seawater temperature changes (Miguez and Combarro 2003); members of the Harveyi clade (Vibrio core group) predominate when the temperatures are above 20°C, and those from the Splendidus clade when it is below that (Maeda et al. 2003). Both northern localities studied had a water temperature ranging from 15 to 18°C and the southern above 20°C. Although no time analysis was performed here, a different species composition was also found between the localities suggesting a temperature as well as a geographic selection of species (Urakawa and Rivera 2006). In animals caught in the Ría de A Coruña, members of the Splendidus clade are the dominant species, 76·6% in crabs from the captive animals (IRTA samples) and 69·4% from the wild animals (Galician coast samples), whereas those from the Canary Islands samples only represent 21·4% of the microbiota. As the crabs kept at IRTA were originally brought from Galicia, it might be speculated that the strains found there could be originated in Galicia, but the band patterns generated by the rep-PCR analysis of the strains from both localities are different. Only in one case, a couple of Vibrio cyclitrophicus strains (Mj13 and Mj157) from both batches have a very high similarity, but still slight band differences can be observed (Fig. S2). This is to be expected because the residence time of bacteria in the haemolymph of crustaceans is from minutes to a few hours (Martin et al. 1993; Alday-Sanz et al. 2002; Burgents et al. 2005). Therefore, it can be suggested that the bacterial population in the haemolymph could be originating from the surrounding environment. Thorough studies of the Vibrionaceae found in molluscs of several localities in Galicia also showed a predominance of members of the Splendidus clade (Beaz Hidalgo et al. 2008) with up to 91·5% of the cultivable vibrios belonging to this clade (Guisande et al. 2008). The number of different OTUs was slightly lower in the species caught in the wild (15) in Galicia than the ones maintained in captivity (17) at IRTA's facilities, this is contrary to what has been observed in other studies; the venerid bivalve Callista chione presented a less diverse bacterial flora in individuals kept in captivity than in recently caught individuals (Delgado et al. 2007). This difference could be because different methodologies were employed in both studies and also because of the different life forms. In this study, many OTUs (11) from IRTA samples (captive crabs) remained identified only to genus level, most belonging to the Splendidus clade. These OTUs need further analysis to clearly identify them, as some might be even new species. One strain, Mj205 and identified as Staphylococcus epidermidis, was most probably a postsample contamination as this species is a normal inhabitant of the human skin (Gotz et al. 2006). The rarefaction analysis also showed that not enough isolates or samples were analysed to obtain a representative of all species potentially found in the haemolymph of the captive crabs. Therefore, it can be concluded that the captive stock has a higher diversity (1/D 13·35) of vibrios than the wild-caught animals (1/D 6·34) for the cold-water sites. The cause of such a higher diversity in the captive animals could be because of the species richness in the locality where the water is taken from; molluscs collected from a nearby locality have rendered up to 21 Vibrio species (Montilla et al. 1994). The species in both studies do not coincide, but this could be because the identification carried out by Montilla et al. (1994) was phenotypical. For th