Ecological determinants of the occurrence and dynamics of Vibrio parahaemolyticus in offshore areas

Abstract

The life cycle of Vibrio parahaemolyticus has been conventionally associated with estuarine areas characterized by moderate salinity and warm seawater temperatures. Recent evidence suggests that the distribution and population dynamics of V. parahaemolyticus may be shaped by the existence of an oceanic transport of communities of this organism mediated by zooplankton. To evaluate this possibility, the presence of V. parahaemolyticus in the water column of offshore areas of Galicia was investigated by PCR monthly over an 18-month period. Analysis of zooplankton and seawater showed that the occurrence of V. parahaemolyticus in offshore areas was almost exclusively associated with zooplankton and was present in 80% of the samples. The influence of environmental factors assessed by generalized additive models revealed that the abundance and seasonality of V. parahaemolyticus in zooplankton was favoured by the concurrence of downwelling periods that promoted the zooplankton patchiness. These results confirm that offshore waters may be common habitats for V. parahaemolyticus, including strains with virulent traits. Additionally, genetically related populations were found in offshore zooplankton and in estuaries dispersed along 1500 km. This finding suggests that zooplankton may operate as a vehicle for oceanic dispersal of V. parahaemolyticus populations, connecting distant regions and habitats, and thereby producing impacts on the local community demography and the spread of Vibrio-related diseases.

Figure 1

Temporal variation in the predicted log-transformed levels of V. parahaemolyticus in zooplankton throughout the period of study (a) and estimated effects of different variables on the predicted abundance of V. parahaemolyticus (log MPN counts) when the effect of time is incorporated into the GAM regression models. Except for total abundance and Cnidaria, the effect of all environmental variables and zooplankton groups is linear when the time effect is included in the GAM regression models (not shown), and their non-linear effects are then associated with their specific seasonality as shared by these variables (b, c, d, f, h and i). Total abundance and Cnidaria abundance are the only variables retaining a non-linear effect on the response independent of the seasonal trend (e and g). The dashed lines indicate the 95% point-wise confidence bands for the predictions.

Figure 2

Effects of the different variables on the abundance of V. parahaemolyticus (log MPN counts) in zooplankton obtained from GAMs: seawater temperature (a), salinity (b), upwelling (c), total abundance of zooplankton (d), Copepoda abundance (e), Cnidaria abundance (f), Cladocera abundance (g), Gastropoda larvae abundance (h) and Lamellibranchia larvae abundance (i). The dashed lines indicate the 95% point-wise confidence bands for the predictions.

Figure 3

Effects of the relative abundance (%) of Copepoda (a), Cladocera (b) and Cnidaria (c) on the V. parahaemolyticus levels in zooplankton. The dashed lines indicate the 95% point-wise confidence bands for the predictions.

Figure 4

Pulsed-field gel electrophoresis dendrogram generated by the BioNumerics software and drawn using the online iTOL software package, showing the genetic relationships among the 42 trh+/tdh− V. parahaemolyticus strains isolated in offshore waters in the present study and the 24 strains obtained from estuarine areas along 1500 km of the coast of Galicia in previous studies. Read from the inner to outer ring, the rings indicate the area of origin (offshore and estuary) and the season of isolation, respectively, for each strain, and the numbers at the dendrogram branches represent the percentage of similarity among the different groups.