Animal Camouflage: Disentangling Disruptive Coloration from Background Matching

Camouflage is ubiquitous in the natural world and provides adaptive benefits to both predators and their prey. In this study I test concepts of animal camouflage using the experimental paradigm of humans foraging for real and artificial moth targets on a computer screen and assessed camouflage efficacy by measuring detection rates. Chapter 1 outlines the questions and objectives of this doctoral thesis. In Chapter 2 I introduce the phenomenon of disruptive coloration, followed by a brief-review of the visual mechanisms contributing to visual search. Chapter 3 tested if non-random orientation behaviour of moths in the field could be explained by behaviourally-mediated camouflage. I showed that the preferred field-orientations of moths were associated with lower detection rates in the lab, and that the relative orientation of the moth to the tree was the key driver. Chapter 4 tested the fundamental assumption that disruptive coloration functions by impairing shape perception. It was predicted that if edge-intersecting patches are disruptive, then altering the shape of a target would interact with edge coloration. Artificial moth-like targets did show an interaction between edge coloration and target shape, which explained detectability. These findings suggest that effectiveness of camouflage due to edge markings is dependent on target shape, which further supports the hypothesis that edge markings function as disruptive coloration. Chapter 5 took a similar approach to chapter 4 but tested if there was an interaction between edge coloration and target boundary visibility, which could explain detectability of moth-like targets. Results from Chapters 4 and 5 suggest that shape and boundary properties play a role in disruptive function of edge markings. Chapter 6 tested how this might occur. It is thought that edge-intersecting patches impair object recognition. It was predicted that moth-like targets with more edge-intersecting patches would be harder to recognise. Recognition was characterised by human foveal vision, monitored by eye-tracking. Indeed, targets with a larger number of edge-intersecting patches were associated with being difficult to recognise, and reduced detectability even at the expense of background matching.