Additive genetic variation, but not temperature, influences warning signal expression in Amata nigriceps moths (Lepidoptera: Arctiinae)

Many aposematic species show variation in their colour patterns even though selection by predators is expected to stabilise warning signals towards a common phenotype. Warning signal variability can be explained by trade-offs with other functions of colouration, such as thermoregulation, that may constrain warning signal expression by favouring darker individuals. Here, we investigated the effect of temperature on warning signal expression in aposematic Amata nigriceps moths that vary in their black and orange wing patterns. We sampled moths from two flight seasons that differed in the environmental temperatures and also reared different families under controlled conditions at three different temperatures. Against our prediction that lower developmental temperatures would reduce the warning signal size of the adult moths, we found no effect of temperature on warning signal expression in either wild or laboratory-reared moths. Instead, we found sex- and population-level differences in wing patterns. Our rearing experiment indicated that ~70% of the variability in the trait is genetic but understanding what signalling and non-signalling functions of wing colouration maintain the genetic variation requires further work. Our results emphasise the importance of considering both genetic and plastic components of warning signal expression when studying intraspecific variation in aposematic species.The first datasheet contains data from wild-caught individuals collected between years 2017 and 2020, which was used to analyse seasonal differences in warning signal expression in the wild population. The second datasheet contains individual data from the rearing experiment, which was used to analyse the effect of temperature on life-history traits and warning signal expression. The third datasheet contains records of larval survival in the rearing experiment (to analyse the effect of temperature on survival), and the fourth datasheet contains data to create a figure of larval survival in the rearing experiment (Fig. 3 in the published article). In addition, we have provided an R script for the analyses.Funding provided by: Australian Research CouncilCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000923Award Number: DP190101028Funding provided by: Jenny ja Antti Wihurin RahastoCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100004022Award Number:We collected data from a) wild Amata nigriceps moths that were captured from Macquarie Park (Sydney, Australia) during years 2017-2020, and from b) Amata nigriceps moths that were reared in the laboratory in three different temperatures (20C, 24C and 28C) during summer 2020-2021. We photographed the moths and quantified the proportion of orange in the wings using pavo package in R to investigate within-species variation in warning signal expression. We also measured wing length from the photographs using ImageJ. We recorded collection site, collection date and sex of the wild-caught individuals. In the rearing experiment, we also recorded larval survival, length of the larval period, pupal weight, time from pupa to eclosion and eclosion success