RESUMO
Since the classic work of E.B. Ford, explanations for eyespot variation in the Meadow Brown butterfly have focused on the role of genetic polymorphism. The potential role of thermal plasticity in this classic example of natural selection has therefore been overlooked. Here, we use large daily field collections of butterflies from three sites, over multiple years, to examine whether field temperature is correlated with eyespot variation, using the same presence/absence scoring as Ford. We show that higher developmental temperature in the field leads to the disappearance of the spots visible while the butterfly is at rest, explaining the historical observation that hindwing spotting declines across the season. Strikingly, females developing at 11°C have a median of six spots and those developing at 15°C only have three. In contrast, the large forewing eyespot is always present and scales with forewing length. Furthermore, in contrast to the smaller spots, the size of the large forewing spot is best explained by calendar date (days since 1st March) rather than the temperature at pupation. As this large forewing spot is involved in startling predators and/or sexual selection, its constant presence is therefore likely required for defence, whereas the disappearance of the smaller spots over the season may help with female crypsis. We model annual total spot variation with phenological data from the UK and derive predictions as to how spot patterns will continue to change, predicting that female spotting will decrease year on year as our climate warms.