first study of vibrios in larval cultures of pullet carpet shell clam (venerupis corrugata) in hatchery

Protocol for hatchery culture of the pullet carpet shell clam Venerupis corrugata spat is currently under development, as the only reliable means of providing spat to replenish natural beds or to support aquaculture activities. Among other variables, the microbiota has been demonstrated to be critical for successful bivalve culture. Shellfish hatcheries are hindered by fatal outbreaks of disease, regardless the bivalve species. These mass mortalities are mainly caused by opportunistic bacteria belonging to genus Vibrio and constitute one bottleneck for this economic activity. Different species, as V. tubiashii, V. pectenicida, V. splendidus, V. neptunius, V. ostreicida and V. bivalvicida, have been identified as responsible of mortalities in hatchery-reared larvae, affecting a wide range of bivalves. This is the first report of the microbiota associated with larval cultures of the pullet carpet shell clam. We present the results of the microbiological analyses of two larval cultures of pullet carpet shell reared in the Centro de Investigacións Mariñas (CIMA, Xunta de Galicia) de Ribadeo (Galicia, NW Spain) following the procedures developed in the institution. Each batch, A and B, was obtained from broodstocks collected in natural environment but in different geographical locations, the stock A (SW Galicia) and the stock B (NW Galicia). Previous records of mortalities led us to divide each batch in two. One sub-batch (A1 and B1) was cultured following the routine procedures. Antibiotic was experimentally added to the other sub-batch (A2 and B2) with the aim of evaluating the effects on the culturable bacterial population (total marine bacteria and presumptive vibrios) and on larval survival. Chloramphenicol, formerly the most commonly used antibiotic in bivalve hatcheries, was supplied with each change of seawater during larval development. Microbiological samples of broodstock, larvae and seawater in culture tanks were taken and processed immediately following Prado et al. (2005, 2014). The bacteriological media used were Marine Agar, for heterotrophic marine bacteria, and Thiosulphate-Citrate-Bile-Sucrose, selective for vibrios. Bacterial counts were determined in all samples and expressed in colony forming unit (cfu) per millilitre or gram. The different types of colony in the samples of larval cultures were isolated, purified and subjected to a basic phenotypic characterization. Further studies were carried out with the isolates that shared the main features of the genus Vibrio, including phenotypic tests and the genetic characterization by sequencing of 16S rRNA gene, as described by Prado et al. (2014). The counts of marine heterotrophic bacteria associated to broodstock gonad were similar in both groups (≈105 cfu/g). However, substantial differences were observed in the counts of presumptive vibrios, with higher values in the stock B (≈104 cfu/g) than in the stock A (≈102 cfu/g). Interestingly, the larval batch A showed high bacterial counts in the initial samples, including presumptive vibrios, while the batch B began with lower numbers of total marine bacteria and vibrios. The survival rates at the end of the larval development were higher in batch B (40% B1, 32% B2) than in batch A (24% A1, 27% A2). The use of antibiotic did not imply a significant enhancement of the survival. The total counts in the seawater from the culture tanks were similar. The presumptive vibrios were found mainly associated to larvae, regardless the batch or the use of antibiotic. The differences on the composition of bacterial populations will be analysed, focused on the identification of the presumptive vibrios. Preliminary results showed the dominance of species within the Splendidus clade of the genus Vibrio, as expected taking into account that its members have been isolated from marine environments and many of them from bivalves. Due to the high phenotypic diversity within this clade and the difficult to differentiate species, the correlation between phenotypic and genotypic identification will be evaluated.