There is substantial heterogeneity in immune response to parasitic infection. For example, in gastro-intestinal nematode infections of sheep, the parasite-specific IgA response is skewed with many animals producing relatively weak responses. As these responses are also highly heritable, genetic theory predicts that evolution will optimise immune responses for maximal fitness, so why many sheep produce weak responses is unclear. Traditional explanations implicate trade-offs between immunity and growth or immunity to different diseases but these are not supported by the data, which suggest that the sheep with strong responses and low parasite loads have the highest growth rates.

One potential explanation is that it may be evolutionarily advantageous for sheep to ‘cheat’ by mounting reduced immune responses and allow other sheep to control infection. We used a game-theoretic approach to explore the benefit of mounting a lower immune response as a function of the strategies of other individuals. Our model, parameterised using data on the sheep - Teladorsagia circumcinta system, encompasses dynamics on epidemiological and evolutionary timescales. The epidemiological dynamics determine the infection levels and growth rates of the flock for a given set of immune responses, whilst the evolutionary dynamics allow mutations to the strategies that determine the host densities and evolutionarily stable immune strategies.

Across a range of costs of infection burden and immune response, the evolutionary dynamics converge on an equilibrium with suboptimal growth rates. This has important consequences for domesticated sheep flocks in which breeding is managed. In contrast to the view that trade-offs may render selection detrimental, our results suggest that it may be possible to achieve higher growth rates in managed sheep flocks by deliberately selecting for enhanced immune responses. In principle, selection procedures that take into account the behaviour of other animals could reduce disease prevalence and enhance growth rates.