LAUSR

Abstract Environmental pollution and climate change can bias the sex ratios of animal populations in which sexual development is sensitive to environmental contaminants and temperature. Investigating these effects in field studies and ecotoxicological experiments is important but difficult when males and females cannot be distinguished without sacrificing them or applying expensive, specialized sexing methods. In this study, we examined the utility of skin coloration as a non-invasive sex marker in juvenile common toads (Bufo bufo) that appear sexually monomorphic. We raised toadlets from eggs in the lab, and exposed them during larval development to one of six treatments: two concentrations of two endocrine disruptor chemicals each (a glyphosate-based herbicide and 17α-ethinylestradiol, a contraceptive) and two controls. Before the first hibernation, we took a photograph of each toadlet’s back, then sexed them by inspecting their gonads, and measured the hue, saturation and brightness of their dorsal skin coloration from the photographs. We found significant sexual dichromatism with males being yellower-greener (less red) and brighter than females; 34% of males and 85% of females could be categorized correctly based on objective colour measurements from photographs. The ratio of greenish and reddish individuals as categorized subjectively by human vision correlated strongly with the sex ratio of treatment groups. Treatment with 1 µg/L 17α-ethinylestradiol resulted in 100% females, with similar coloration as normal females. Intersex individuals occurred in treatment groups with 3 µg/L glyphosate and 1 ng/L 17α-ethinylestradiol; these animals were less saturated and darker reddish-brown compared to normal individuals. These results suggest that skin coloration can indicate phenotypic sex and gonadal abnormalities in common toads. Although skin colour measurement in itself is insufficient for identifying an individual's sex or the sex ratio of a single group, it can be useful for qualitative comparisons of sex ratios between groups when no other means of phenotypic sexing is possible. We propose that counting the number of greenish and reddish individuals as seen by the human eye is a cheap and fast non-invasive method for identifying natural populations or experimental groups that may have skewed sex ratios compared to other groups, and this approach is worth testing in other species to help conservation practices and non-destructive ecotoxicological experiments.