COS 162-3 - The evolution of imperfect prudence

Thursday, August 9, 2012: 2:10 PM
E146, Oregon Convention Center
Doug Jackson, Southwest Fisheries Science Center, NOAA National Marine Fisheries Service, Santa Cruz, CA
Background/Question/Methods

A well-established theoretical foundation, coupled with recent experimental work, has shown that prudent pathogens can evolve in spatially-structured host-pathogen systems. Prudence is a form of cooperation in which individual pathogens reduce their rate of reproduction (e.g., transmissibility or infectivity), thereby preventing the overexploitation of the common resource (susceptible hosts). 

In a well-mixed population, prudence is evolutionarily unstable: cheaters that more voraciously exploit the common resource outcompete prudent pathogens, leading to a population dominated by rapacious pathogens. However, when the movement of pathogens and/or hosts is restricted, rapacious pathogens can be at a fitness disadvantage due to their overexploitation of their local environment.

Using a probabilistic cellular automata model, I ask whether prudence of hosts, in the form of reproductive restraint, can similarly evolve in a spatially-structured host-pathogen system. Hosts and pathogens reproduce locally with some probability. Host reproduction probability evolves via random mutation during host reproduction. Rapidly reproducing hosts should tend to create large, well-connected clusters, providing an ideal landscape for the spread of the pathogen, generating a conflict between what benefits the individual (maximizing reproduction) and what benefits the group (minimizing the size and connectivity of host patches).

Results/Conclusions

Prudent hosts, characterized by an intermediate host reproduction probability (g), evolved across a range of initial conditions: the average host reproduction probability of the model converged to an ESS of 0.2 for ginitial = [0, 0.65]. Above this range, the population evolved to a maximum reproduction probability, at which point the host population would invariably form a single large cluster that would subsequently be extinguished by the pathogen.

Group-level characteristics such as population size and variability were not optimized. At the ESS, relatively large clusters would form and then be eliminated by the pathogen, leading to large fluctuations in the total host population size. At lower host reproduction probabilities, the host landscape was characterized by very small, evenly distributed, isolated clusters, leading to much smaller epidemics and a larger average host population size.

These results demonstrate that it is theoretically possible for a low-connectivity host distribution to evolve in a spatially-structured host-pathogen system. More generally, they suggest that the spatial distribution of hosts could be a phenotype subject to evolutionary pressure. However, these results also suggest that we should consider multiple measures of performance when assessing the degree to which evolved behaviors benefit the population.