Warning signal plasticity in hibiscus harlequin bugs

Color variation in aposematic (conspicuous and defended) prey should be suppressed by frequency-based selection by predators. However selection of color traits is confounded by the fact that coloration also plays an important role in many biological processes, and warning coloration may be constrained by biotic or abotic factors. Temperature, in particular the importance of thermoregulation, has been suggested as the source of much of the geographical variation in warning coloration we see in natural populations. Differential selection in different thermal environments may lead to developmentally canalized or 'fixed' differences between populations. Conversely, inter-population differences may be due to phenotypic plasticity, wherein trait expression is modified by environmental conditions. The hibiscus harlequin bug Tectocoris diophthalmus (Heteroptera: Scutelleridae), is a shieldback bug, with iridescent patches that show size variation between individuals, as well as inter-population variation with geographic patterning. This study aimed to identify environmental factors that drive the expression of this variable trait, using surveys, modeling, and experimental approaches. Surveys were taken at sites throughout Australia in three climate regions (tropical, subtropical, and temperate) at different time periods, and results were modeled with a multilevel ordinal regression. We tested for correlations between colouration and several biotic (density, host plant) and abiotic (temperature, rain-fall) factors. We found strong phenotypic plasticity with respect to temperature and rainfall. Higher temperatures and increased rain-fall were related to suppressed iridescence. A factorial experiment with tropical and temperate bugs in two climate-typical temperature regimes confirmed phenotypic plasticity in response to temperature, likely due to temperature sensitivity in melanin expression. Tropical and temperate populations showed striking differences between plasticity reaction norms, suggesting local evolution on the shape of phenotypic plasticity. We suggest that studying both biotic and abiotic selection pressures is important for understanding the causes of inter-population variation in aposematic signals.