Aposematic prey use conspicuous warning signals to advertise a secondary defense. To be effective, warning signals should be reliable throughout a prey population. However, warning signals vary considerably across and within populations, often for unknown causes. Here, the Australian diurnal moth Amata nigriceps, which has orange wing spots on a black background, was used to explore four ecogeographic hypotheses explaining warning signal variability: the thermal melanism, the photo-protection, the desiccation resistance, and the predator-loss hypotheses. We collected live moths from 29 populations across 746 km of coastal New South Wales, Australia, and sampled relevant museum collections from 1896 to 1982. We quantified the proportion of orange wing spots (warning signal size) relative to site-specific abiotic (temperature, precipitation, solar radiation) and biotic factors (diversity of bird predators).

In museum specimens, warning signal size varied within the decade of collection, without a discernible temporal trend. Recent collections also varied in signal size, but we found no support for the four hypotheses. Our results are in contrast with studies that have found significant relationships between warning colours and eco-geographic factors. Local phenotypic responses in Amata nigriceps together with relaxed predator selection and genetic effects may explain the persistence of warning signal variation.

Fresh Amata nigriceps moths were collected from 29 sites across eastern Australia, preserved, and imaged.

Historical specimens were imaged from 3 Australian museums, dating from 1896 to 1982.

Proportion of warning signal size (orange wing spots on a black background) was quantified, and this 'pSpot' value was used to compare changes in warning signal size over time in the historical collections, and compared with climate and bird diversity data in the fresh specimens.