Sexual selection in the neotropical treefrog Scinax crospedospilus (Anura, Hylidae): integrating testes, calls and social interactions

Abstract

Sexual selection is classically known to act on premating traits such as ornaments and armaments to increase mating success. In the last decades, however, the consensus is that, when females mate with multiple males, sperm can compete to fertilize the ova, so much so that advantageous characteristics of ejaculates should be also under selection (postmating sexual selection). Pre- and postmating traits should covary in two different directions. The theory of sperm competition predicts a fundamental trade-off (negative correlation) between pre- and postmating traits when resources are limited. In another direction, the phenotype-linked fertility hypothesis predicts a positive correlation in which females may access male fertility status through exaggerated phenotypic traits. Frogs are well known for their calling behaviour, although details on the social context of emission are still lacking, and new studies have investigated postmating traits such as ejaculate characteristics. We used the treefrog Scinax crospedospilus (Anura: Hylidae) as a model to test possible relationships between call traits vs. sperm and testes traits for the first time in amphibians. We analyzed calls recorded in the field (Botanical Garden of the Universidade Federal de Juiz de Fora, Brazil) and sperm number, morphology and testes mass of collected males in the lab. In addition, we made video and audio recordings and direct observations in the field to study the social interactions of individuals, as well as site fidelity to bromeliads in a mark-recapture approach. According to the trade-off hypothesis, we found that sperm head length and sperm tail length traded off against note duration and second band dominant frequency, respectively. According to the phenotypelinked fertility hypothesis, we found that sperm tail length, sperm total length and sperm number correlated positively with call rate and that sperm head length correlated negatively with first band dominant frequency, i.e., males invested equally in two types of traits that are advantageous at the same time. We described for the first time the species’ encounter call, and male-male interactions that include chasing and four types of visual signals. Individuals showed some degree of site fidelity in bromeliads, which they might use as a refuge and calling site. This is the first study to show covariations between vocalization and ejaculate characteristics in frogs, and we highlight the importance to incorporate sperm size in the context of postmating sexual selection to understand the total opportunity for sexual selection in animals. Regarding the reproductive behaviour of Scinax crospedospilus, our results expand the communication repertoire for the species long after its description and should contribute for future experimental studies on sexual selection in frogs, especially within the Scinax genus, in which studies of social behaviour are sorely needed.