Explaining imperfect ant mimicry

Many species have evolved to mimic unprofitable models in order to avoid attack by predators. Mimics are most likely to deceive predators if they resemble their models very accurately, so the existence of many poor mimics is a puzzle. Currently, there is no shortage of hypotheses to explain the persistence of poor mimicry, but comparative evidence supporting them is largely lacking. In this study, I first review the literature regarding imperfect mimicry hypotheses and evidence, then address several hypotheses using ant mimics as study subjects. Ants are avoided by many predators, and ant mimics are widespread, abundant and diverse, making them effective subjects for mimicry studies. Utilising morphometric analysis to quantify body shapes and statistical analysis of walking trajectories as a measure of behavioural mimicry, the accuracy and variation of mimicry in ant-mimicking spiders is compared with that of ant-mimicking insects, and used to assess several imperfect mimicry hypotheses. We find some evidence that poor visual mimicry in spiders may result from developmental constraints, but no evidence that selection is relaxed for smaller mimics. Additionally, ant-like walking neither reinforces nor compensates for visual mimetic accuracy.