Sexual dimorphism in fin size and shape in North American Killifish

Sexual dimorphism is intriguing because it suggests that males and females differ in phenotypic optima for traits and that sex-specific trait values can evolve despite a shared genome. Differences in sexual dimorphism across populations or species suggest that the nature of sexual selection and/or genetic constraints differs among species. Here, we measured sexual dimorphism in 20 species of North American killifish (Fundulidae) in size and shape of dorsal, anal, and caudal fins. We observed profound sexual dimorphism in anal and dorsal fin size and shape across all species, suggesting a common direction of selection. Sexual dimorphism was also present in caudal fin size and shape but was much lower in magnitude, with several species not differing from a null expectation of zero. There was little evidence for a phylogenetic signal in the levels of sexual dimorphism in dorsal and anal fin traits. We also found a strong phylogenetic correlation between sexual dimorphism in anal and dorsal fin shape but no phylogenetic correlation between fin area, base length, or ray length across different fins. Our results indicate that there is pronounced sexual dimorphism in anal and dorsal fin size and shape across fundulids. Similar patterns of sexual dimorphism in anal and dorsal fin properties have been documented in other groups, including gars, bichirs, graylings, minnows, and many species in the Atherinomorpha, suggesting that this pattern may be common across Actinopterygii.

Methods
Here, we measured sexual dimorphism in 20 species of North American killifish (Fundulidae) in size and shape of dorsal, anal, and caudal fins. We took digital photos of 363 individuals across 20 species. We measured standard length and fin base length for the dorsal, anal, and caudal fins. For the dorsal and anal fins, we also measured the lengths of the first and last fin rays as well as the lengths of median, first quartile, and third quartile fin rays. For the caudal fin, we measured the median fin ray length and the length of the upper dorsal and lower ventral fin ray. From these data, we estimated fin area. We also used published genbank data to estimate a phylogeny for the group. We analyzed the data in R. Because there is pronounced variation in standard length (i.e., size) across these species, we first performed a linear regression of the log of the trait value on the log of standard length and used the residuals as a measure of size-corrected fin traits.