Evidence for an association between genetic and acoustic variation in a poison frog

Abstract

Acoustic signaling is a key component in both intra- and inter-specific interactions. These signals can be used in many contexts, including mating behavior and agonistic interactions. Poison frogs are of particular interest because the evolution of aposematism has been linked with increased call diversity in this group. Acoustic signals are hypothesized to be used for communication over long distances, whereas at shorter distances they are hypothesized to be used in multi-modal signaling with visual signals. Currently, the genetic variants underlying diversification in acoustic signals in the poison frog family are unknown. The mimic poison frog (Ranitomeya imitator) exhibits acoustic signal phenotypes that vary geographically and coincide with geographic clines of color variation. The advertisement call of this species varies between color morphs with respect to dominant frequency (pitch), pulse rate (number of pulses present in a single note) and note length (grouping of pulses with several notes being a call). My goal throughout this research is to understand the role of two genes (Synaptojanin 1 and Apoptosis Inducing Factor Mitochondria Associated 1) in the acoustic divergence of R. imitator. Research has linked both genes to the maintenance and development of hearing in mammal species, but their roles have not been identified in amphibians. In R. imitator, these genes show significant divergence between two color morphs that have diverged across a geographic cline and are predicted to influence the perception of acoustic signals. Using both field-collected recordings of acoustic signals and genetic samples from ten populations of R. imitator, I identified differences in call traits and genotyped each individual for each gene to test for associations. I found that all 3 call traits had some variation between populations and color morphs. Note length was the most variable trait between individuals and pulse rate and dominant frequency were much more static traits, suggesting different selective pressures are acting on each of these traits. I also found a significant association between aifm1 genotypes and dominant frequency, but I did not find an association with synj1 or the other call traits analyzed. Mechanisms by which acoustic signals diverge in frogs include sexual and ecological selection, geography, morphology, and signal perception. Signal perception can, specifically, be impacted by genetic factors influencing how acoustic signals are perceived and processed. My results suggest acoustic divergence in R. imitator may be influenced by factors other than geographic distance between populations and that these signals are actively diverging between populations based on differences seen in this study and previous studies. In particular, divergence in dominant frequency seems to be driven by the gene aifm1, possibly driven by changes in how signals are perceived by individuals between populations. This is the first study to investigate the role of aifm1 in amphibian signaling. Future studies should focus on how allelic variation in this gene contributes to phenotypic variation and how calls are used to mediate interactions in R. imitator.