Effects of Continuously Varying Landscapes on the Evolution of Cryptic Polymorphism

Sammanfattning

In a changing heterogeneous landscape camouflage becomes a less effective defence for a mobile prey item because the prey needs to move across backgrounds where it is more visible to the predator. Often, some subhabitats host more resources than others, adding an additional complexity. The prey can thus either try to generally match many subhabitats or specialize to one habitat type, often resulting in polymorphic populations of many specialists. The selection for cryptic morphs happens through predation and the contribution of predator-prey interactions are already well studied, but previously heterogenous landscapes have rarely been modelled explicitly. A previous patch model suggested that a prey should specialize to the habitat which it prefers, but does this generalize to any spatial arrangement of habitats? Here I implement an Adaptive Dynamics framework for reaction-diffusion systems to study these questions for landscapes where background colour varies continuously between two extremes. Landscapes are randomly generated and evaluated by variance, skewness and Moran's Index which is a proxy for the landscape's `patchiness'. I study two scenarios: a) the prey finds resources uniformly throughout the landscape and b) the distribution of resources correlates with background colour making one habitat type more preferable. For robustness I also compare results to a two-patch model. In case a), variance is found to be the most decisive factor for predicting polymorphism. High variance significantly promotes polymorphism, whilst high patchiness and low skew slightly facilitate polymorphism. In b) none of these factors alone can predict polymorphism, but interactions skewness*patchiness and skewness*variance are decisive. The results mean, that an adaptation to resource poor patches is advantageous if richer patches are small and scattered, thereby giving rise to cryptic polymorphism. The model here finds that varying predation and the crypticity trade-off form polymorphism, but the effect varies in amplitude on different landscapes. Thereby, landscapes define the limits to the interplay between ecology and evolution. This work contributes to a better understanding of the landscape features which lead to different evolutionary outcomes and explain the sometimes puzzling differences in strategies for crypsis observed across species.