Discussion:
In this study, we evaluated sequence divergence and incidence of recombination in three major endosymbionts (Wolbachia ,Cardinium and Arsenophonus ) to answer whether the ecological community that they are a part of is the primary seat of their horizontal transfer and diversification. We used soil arthropod community because it is relatively insular and has a relatively high habitat endemicity of the residents. Our main goal was to assess whether community members facilitate the spread of endosymbionts as they themselves come in contact with each other for various ecological interactions. To properly assess whether ecological communities are indeed the seat of endosymbiont transfer and diversification, one needed to compare multi-gene phylogenies of such endosymbiont surveys from different communities. However, in spite our extensive literature surveys we could not find any such previous reports. Most surveys of arthropod communities concentrated on the hosts rather than on their endosymbionts (Gonçalves, Pereira, & Liu, 2012). Some studies like Kittayapong et al. (2003) and Sintupachee et al. (2006) did uncover the resident endosymbionts but mostly with single genes. This precluded a cogent comparison of endosymbiont diversity and incidence of recombination with the present study. Another set of studies did indeed sample endosymbionts with multi-gene sequences but concentrated on a few, and not all, host taxa within a community (Bing et al., 2014). Again, such studies are not ideal comparisons with the present one as these were biased towards a few host taxa. To partially overcome this problem, we used statistical models with extensive resampling. We observed that the supergroup A Wolbachia infections andCardinium do indeed show less pairwise divergence, than expected, in accordance with our predictions. However, supergroup BWolbachia and Arsenophonus infections did not show this pattern. In fact, the former shows more variation than expected whereasArsenophonus shows no significant difference. This indicates that these endosymbionts have different propensity and/or rates of horizontal transfer within the community. We speculate what can be the reasons behind this.