Abstract
Recent outbreaks of the West Nile virus have been reported in southern Spain, a region with important wetland habitats for migratory birds. Here, we analyzed the role of species association and abiotic parameters on the abundance of seven mosquito species in the Doñana National Park, Spain. We applied the Joint Species Distribution Models (JSDM), using the Hierarchical Modeling of Species Communities approach to simultaneously model the effect of habitat type, normalized difference vegetation index, hydroperiod, distance to rivers, land surface temperature, and the abundance of seven mosquito species. We created three models with varied parameters and evaluated the effects of abiotic parameters and species-to-species associations, which served as proxies for species interactions. Our models estimated the species-specific responses of the mosquito community to abiotic parameters. A positive association was identified within the Culex community with potential biotic interactions. However, Ochlerotatus caspius, Ochlerotatus detritus, and Anopheles atroparvus had no statistically significant association with each other and with the Culex species after accounting for the effect of abiotic parameters. We also found that the potential species associations estimated at particular sites and abiotic parameters influenced the model projection for the average abundance of mosquito species. The JSDM will allow spatial projection of the abundance of each mosquito species, which is an important parameter for epidemiological models. The JSDM inference of species association is important because predation, competition, and facilitation affect the distribution and abundance of different species.