Phenotypic integration mediated by hormones: associations among digit ratios, body size and testosterone during tadpole development

Background: Developmental associations often explain phenotypic integration. The intersected hormonal regulation of ontogenetic processes fosters predictions of steroid-mediated phenotypic integration among sexually dimorphic traits, a statement defied by associations between classical dimorphism predictors (e.g. body size) and traits that apparently lack sex-specific functions (e.g. ratios between the lengths of Digits II and IV - 2D:4D). Developmental bases of female-biased 2D:4D have been identified, but these remain unclear for taxa presenting male-biased 2D:4D (e.g. anura). Here we propose two alternative hypotheses to investigate evolution of male-biased 2D:4D associated with sexually dimorphic body size using Leptodactylus frogs:I)'hypothesis of sex-specific digit responses' - Digit IV would be reactive to testosterone but exhibit responses in the opposite direction of those observed in female-biased 2D:4D lineages, so that Digit IV turns shorter in males; II) 'hypothesis of identity of the dimorphic digit' - Digit II would be the dimorphic digit. Results: We compiled the following databases using Leptodactylus frogs:1) adults of two species from natural populations and 2) testosterone-treated L. fuscus at post-metamorphic stage. Studied traits seem monomorphic in L. fuscus; L. podicipinus exhibits male-biased 2D:4D. When present, 2D:4D dimorphism was male-biased and associated with dimorphic body size; sex differences resided on Digit II instead of IV, corroborating our 'hypothesis of identity of the dimorphic digit'. Developmental steroid roles were validated:testosterone-treated L. fuscus frogs were smaller and exhibited masculinized 2D:4D, and Digit II was the digit that responded to testosterone. Conclusion: We propose a model where evolution of sexual dimorphism in 2D:4D first originates from the advent, in a given digit, of increased tissue sensitivity to steroids. Phenotypic integration with other sexually dimorphic traits would then occur through multi-trait hormonal effects during development. Such process of phenotypic integration seems fitness-independent in its origin and might explain several cases of steroid-mediated integration among sexually dimorphic traits.