Figure 3

A spatial lattice model with varying r. Each row of figures corresponds to a different initial spatial structure organized by the average value of r. In each case the leftmost figure shows an example of the initial population state (independently generated for each simulation), where PCD+ genotypes are shown in red (light for A _p and dark for B _p ) and PCD− are in blue (light for A _n and dark for B _n ). The bar graphs show the number of times PCD+ genotypes win out of 100 competitions for different probabilities of PCD and disasters. In all cases, higher disaster probabilities allow PCD+ strains to win more often and can support strains with higher rates of PCD (comparing PCD = 0.1 at disaster probabilities of 0.01 and 0.1 is significant with p ≪ 0.01 using a Fisher’s exact test). The next two panels show the average value of r in simulations that PCD+ won (red) and PCD− won (blue) over time. Time is scaled by the length of the simulation. The disaster probability and PCD probability are both 0.05. The black line shows the mean across the simulations. When PCD+ wins the mean r value is high and stays close to 0.9. When PCD− wins, the r value is lower and fluctuates more (shown by the variation in the blue plots).