Abstract
Although pervasive, the impact of temporal environmental heterogeneity on coevolutionary processes is poorly understood. Productivity is a key temporally heterogeneous variable, and increasing productivity has been shown to increase rates of antagonistic arms race coevolution, and lead to the evolution of more broadly resistant hosts and more broadly infectious parasites. We investigated the effects of the grain of environmental heterogeneity, in terms of fluctuations in productivity, on bacteria–phage coevolution. Our findings demonstrate that environmental heterogeneity could constrain antagonistic coevolution, but that its effect was dependent upon the grain of heterogeneity, such that both the rate and extent of coevolution were most strongly limited in fine-grained, rapidly fluctuating heterogeneous environments. We further demonstrate that rapid environmental fluctuations were likely to have impeded selective sweeps of resistance alleles, which occurred over longer durations than the fastest, but not the slowest, frequency of fluctuations used. Taken together our results suggest that fine-grained environmental heterogeneity constrained the coevolutionary arms race by impeding selective sweeps.
Highlights
The importance of environmental heterogeneity for antagonistic species interactions was recognized over 50 years ago by the ‘disease triangle’ concept, which identifies host genotype, pathogen genotype and the environment as the primary determinants of infection outcome [1]
Until recently studies of antagonistic coevolution have often treated environmental heterogeneity as ‘noise’ that has been excluded from experimental work and theoretical models [2]
Our findings demonstrate that environmental heterogeneity, constituting fluctuation between high and low productivity, can act to decelerate antagonistic coevolution, and constrain evolutionary escalation in the breadth of bacterial resistance range and phage infectivity range
Summary
The importance of environmental heterogeneity for antagonistic species interactions was recognized over 50 years ago by the ‘disease triangle’ concept, which identifies host genotype, pathogen genotype and the environment as the primary determinants of infection outcome [1]. (Note that ‘time-shift’ was fitted as a linear covariate because we were interested here in detecting change in the rate of trait evolution in response to directional selection.) To test the effect of treatments on the extent of coevolution, resistance range and infectivity range data from the cross-infection assay were analysed in separate models fitting ‘treatment’, ‘timepoint’ and their interaction as fixed effects, and ‘founding clone’ and ‘population’ nested within ‘founding clone’ as random effects. To conform to model assumptions (i.e. normality, homogeneity of variance), infectivity and resistance data were arcsine-square-root-transformed, infectivity range data were arsine-transformed and resistance range data were square-root-transformed
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
