Vibrio parahaemolyticus is one of the principal bacterial causes for seafood-borne gastroenteritis in the world. In the present study, three sites located on the French Atlantic coast were monitored monthly for environmental parameters over 1 year. The presence of total and pathogenic V. parahaemolyticus in sediment, water and mussel samples was detected following enrichment by culture and real-time PCR (toxR gene, tdh, trh1 and trh2 virulence genes). Using generalized linear models, we showed that the presence of V. parahaemolyticus in water could be explained by a combination of mean temperature over the 7 days before the day of sampling (P < 0.001) and turbidity (P = 0.058). In mussels, an effect of chlorophyll a (P = 0.005) was detected when an effect of the mean salinity over the 7 days before sampling was significant for the sediment (P < 0.001). We did not detect any significant effect of phytoplanktonic blooms or of the number of culturable bacteria on V. parahaemolyticus presence. No sample was revealed positive for tdh. The presence of trh1 and trh2 was positively influenced by the mean temperature during the 2 days before the day of sampling (P < 0.001 and P = 0.032). The importance of these ecological parameters is discussed in relation to the biology of V. parahaemolyticus.

The present study focused on the isolation of culturable bacteria from mussels and sea water to identify Vibrionaceae potentially pathogenic for humans. Three sites located on the French Atlantic coast were monitored monthly (twice each month during summer) for 1 year. Environmental parameters were surveyed (water temperature, salinity, turbidity, chlorophyll a) and bacteria were detected by culture and identified by API 20E(®) systems (BioMérieux) and PCR. A total of seven species were detected (Grimontia hollisae, Photobacterium damselae, Vibrio alginolyticus, V. cholerae, V. fluvialis, V. vulnificus and V. parahaemolyticus) and species diversity was higher at the end of summer. Surprisingly, V. cholerae non-O1/non-O139 was detected in spring. No site effect was detected. Using Sørensen similarity indices and statistical analyses, we showed that chlorophyll a had a significant influence on the bacterial community detected in mussels and assemblages were more similar to one another when chlorophyll a values were above 20 µg l(-1) . No significant effect of any parameter was found on the community detected in water samples. Such surveys are essential for the understanding of sanitary crises and detection of emerging pathogens.