Figure 8.

Temperature negatively correlates with Wolbachia abundance, and Wolbachia abundance negatively correlates with West Nile virus (WNV) in Culex pipiens/restuans. (A) Wolbachia, but not WNV, is significantly correlated with mean weekly temperature in samples (n = 538) from Toronto. Daily climate data for 2011–2013 Toronto (WMO identifier 71265) downloaded from: http://climate.weather.gc.ca/historical_data/search_historic_data_e.html. (B) The relative abundance of Wolbachia in Cx. pipiens/restuans pooled samples (n = 591) over 3 years was significantly correlated with West Nile virus prevalence in pooled samples after a 3 week delay (see Table S6). Note that although the correlation coefficient increased from weeks 0 to 3, there was not a significant correlation until 3 weeks had elapsed. Thus, a decrease in Wolbachia abundance may contribute to conditions favoring WNV susceptibility in the succeeding mosquito generation. (C) A conceptual model illustrates the hypothesis that higher temperatures reduce Wolbachia abundance (orange) and lead to higher WNV prevalence (blue) in the following generation. Subsequently, fitness costs of WNV (ω), seasonal reductions in temperature, or lower density of mosquito larvae may drive the cyclical pattern and selection for increased abundance of protective Wolbachia (red).