Can warning signals be honest? : wing colouration and the strength of chemical defence in the female wood tiger moth (Parasemia plataginis)

Abstract

The warning displays of aposematic organisms signal to predators that they possess a secondary defence and are unprofitable. Within species variation exists in the strength of the signal and defence. As natural selection is expected to favour higher levels of defence, variation in such traits requires explanation. One hypothesis is that the strength of primary and secondary defences are correlated as they reflect the condition of the signaller. This study explores if variation in individual conspicuousness is an honest signal of the level of defence in the wood tiger moth (Parasemia plantaginis). P. plantaginis has conspicuous hind wing warning colouration, which in females varies from yellow to red. I investigate if variation in female colour corresponds with the effectiveness of their secondary chemical defences. I tested the first assumption of honesty; that either the primary (colouration), or secondary (chemical) defence, is costly for its’ producer. Cost was manipulated via food deprivation in the larval stage. I expected food deprived individuals would possess weaker defences compared to control individuals, which were fed ad libitum. To measure the strength of the secondary defence I tested the response of a predator (blue tit, Cyanistes caeruleus) to the defence fluids of the adult moths, in the absence of any visual cues. As selection by predators is thought to shape the evolution of chemical defence in P. plantaginis, predator aversion should be an accurate reflection of defensive strength. To test the second assumption of honesty; correlation between the strength of warning signal and defence, I used the variation present in my sample population to test for a correlation between adult hind wing colour and the chemical defence strength. I expected that fluids from redder moths would induce greater aversion in the predators. Food manipulation had no significant effect on wing colouration or amount of defensive thoracic fluid produced. However, the odour quality of fluids was influenced by the larval treatment, as birds hesitated longer before attacking prey applied with fluids from control moths. Birds also showed more disgust behaviour (beak cleaning) in response to fluids from redder moths. However, the opposite was true for the proportion of the prey eaten; those with fluids from more conspicuous moths were eaten more, and amongst the palest moths those from the control were eaten more than the food deprived. My results suggest that secondary defences in female P. plantaginis are costly. However, it is less clear if the conspicuousness of the warning signal is an honest signal of unprofitability, suggesting that the chemical defences are multimodal and undergo different levels of investment at different phases of the predation sequence. The efficiency of chemical defence seems to depend on external factors, such as resource availability and predator community structure.