Spatial genetic patterns systematically accelerate and bias drift-based genetic erosion

Abstract

Genetic erosion of an ecosystem's key species weakens the basis of ecosystem stability. In the absence of selection and migration genetic drift is the only factor influencing the degree of genetic erosion. Populations of sessile organisms represent a pattern of genetic information in space. In this paper, we show how spatial genetic patterns bias and accelerate the dynamics of drift-based genetic erosion. Using a Cellular Automaton (CA) as a modeling environment for discrete systems with local dynamics, we study the boundary conditions for such pattern-dependent genetic erosion. The system is designed as a one locus two allele model for haploid loci. Each cell in the CA represents one sessile individual of the simulated population. In order to analyze the behavior of the model we varied the following four variables, (1) the initial spatial distribution of haplotypes; (2) the magnitude of local gene-flow; (3) the noise in the initial pattern and; (4) the intensity of global (non-local) gene-flow. We show that for certain spatial genetic patterns genetic drift systematically leads to the fixation of one allele, if the size of the patterns and the dimension of local gene flow are of similar scale. Moreover, drift is substantially accelerated compared to the situation, where the two alleles are randomly distributed. These results are rather stable to noise in the initial pattern but external gene flow (EGF) has to be limited to a certain threshold to allow spatial patterns to drive genetic erosion.