Evolution of body size and wing shape trade-offs in arsenurine silkmoths

One of the key objectives in biological research is understanding how evolutionary processes have produced Earth9s biodiversity. These processes have led to a vast diversity of wing shapes in insects; an unanswered question especially pronounced in moths. As one of the major predators of nocturnal moths, bats are thought to have been involved in a long evolutionary arms race with their prey. In response, moths are thought to have evolved many counter strategies, such as diverse wing shapes and large body sizes. However, the tradeoffs between body size and wing shape are not well understood. Here we examined the evolution of wing shape in the wild silkmoth subfamily Arsenurinae (Saturniidae). By using phylogenomics and geometric morphometrics, we established the framework to evaluate potential evolutionary relationships between body size and wing shape. The phylogeny was inferred based on 781 loci from target capture data of 42 arsenurine species representing all 10 recognized genera. We found there are evolutionary trade-offs between body size, wing shape, and the interaction of fore- and hindwing shape. Namely, body size decreases with increasing hindwing length, but increases as forewing shape becomes more complex. Additionally, hindwing shape has a significant effect on forewing shape complexity. The complex wing shapes that make Arsenurinae, and silkmoths as a whole, so charismatic are likely driven by the strong forces of natural selection and genomic constraints. One other important outcome was discovering within our data one of the most vexing problems in phylogenetic inference - a region of a tree that possesses short branches and no "support" for relationships (i.e., a polytomy). These parts of the Tree of Life are often some of the most interesting from an evolutionary standpoint. To investigate this problem, we used reciprocal illumination to determine the most probable generic relationships within the Arsenurinae by inspecting differing phylogenetic inferences, alternative support values, quartets, and phylogenetic networks to reveal hidden phylogenetic signal.