Evolutionary history structures plant-hawkmoth interactions frequency, beyond its effect through ecological traits

Multiple non-exclusive neutral or niche-based processes influence species interactions and shape the structure of ecological networks.Although evolutionary processes may influence interactions via shared inherited traits, their relative importance in comparison to ecological processes remain poorly understood.Here we investigate the effects of evolutionary history compared to ecological processes on both interaction frequencies and the emergence of modularity in a plant-hawkmoth interaction network.Finally, we tested the phylogenetic signals on hawkmoth ecological traits relevant to pollination interactions.We analyzed a planthawkmoth network on Pampa grasslands for which we collected morphological, phenological, abundance and phylogenetic data for every species in the network to evaluate the effect of neutral-based processes (i.e., random chance of encounter dictated by species abundances) or niche-based processes (i.e., morphological matching) on species interactions.The best model predicting interactions indicated that, beyond increasing interaction frequencies with increased morphological matching and phenological overlap, intermediate levels of phylogenetic isolation confer higher probabilities of interaction for hawkmoths while lower phylogenetic isolation confer higher probabilities of interaction to plants Further, by evaluating whether hawkmoth composition differed among interaction modules, we found that modules encompassed phylogenetically distant and functionally distinct hawkmoth species.Thus, module composition suggests that, despite evolutionary history and ecological traits influencing interactions' frequency, modules emergence may be also potentially modulated by processes such as environmental filtering or stochastic factors acting on other traits than the ones evaluated here.Such processes may have influenced species composition in the community and ultimately also shape the observed interaction patterns.