Search

Understanding the effect of multiple infections is essential for the prediction (and eventual control) of virulence evolution. Some theoretical studies have considered the possibility that several strains coexist in the same host (coinfection), but few have taken their within-host dynamics explicitly into account. Here, we develop a nested approach based on a simple model for the interaction of parasite strains with their host's immune system. We study virulence evolution by linking the within-host dynamics to an epidemiological framework that incorporates multiple infections. Our model suggests that antigenically similar parasite strains cannot coexist in the long term inside a host. We also find that the optimal level of virulence increases with the efficiency of multiple infections. Finally, we notice that coinfections create heterogeneity in the host population (with susceptible hosts and infected hosts), which can lead to evolutionary branching in the parasite population and the emergence of a hypervirulent parasite strategy. We interpret this result as a parasite specialization to the infectious state of the hosts. Our study has experimental and theoretical implications in a virulence management perspective.

Offspring quality and quantity are subject to a trade-off. Depending on species, ant colonies can produce either numerous winged queens that start new colonies alone (independent colony foundation [ICF]) or one propagule consisting of reproductives helped by workers (fission). Some species do both. In the Rhytidoponera impressa group, colonies with winged queens use ICF, whereas colonies with mated workers use fission. We assessed the effect of latitude and environment on colonial reproduction by collecting 79 colonies along the Australian east coast. A population dynamical model predicted that from tropical north to temperate south, seasonal fluctuations and harsher conditions should lead to a decrease in ICF success (-48%), making fission a suitable alternative. Yet, our empirical data showed that ICF persists at a higher rate than expected, presumably because it allows aerial dispersal, unlike fission. Investigation of colony demography and measurement of lean and fat weights in virgin winged queens and workers showed that, from north to south, colonies increase quality of queens relative to workers (+50%) at the cost of quantity (-86%). This modification limits the decrease in ICF success (-34%). A tremendous range of offspring phenotypes associated with quality-quantity trade-offs make ants competitive in diverse habitats.

The occurrence of polygyny and polyandry in social insects has long puzzled evolutionary biologists. If cooperation requires genetic relatedness, how do we explain the occurrence and maintenance of mechanisms that reduce the degree of relatedness among colony members? A much‐discussed hypothesis states that genetically diverse colonies are more resistant to parasitism than homogenous colonies because genetic diversity reduces the spread of a disease within a colony. However, as we will argue in this note, a necessary condition for the parasite hypothesis is that genetically heterogeneous colonies have a larger suite of parasites that are capable of infecting them. This implicit relationship is important because it implies that even if the cost per infection is reduced, this may not be sufficient to offset the increased rate of acquiring infections. The advantages of genetic heterogeneity as a defense against parasites thus may not be as big as commonly thought.

Most models of virulence evolution assume that a parasite cannot raise its transmission rate without causing more harm to its host. However, the existence of such trade‐off relationships has recently been challenged. Here, we study how a trade‐off can emerge from a model that explicitly incorporates within‐host dynamics. We find that the existence and the convexity of the trade‐off are robust, which implies a definite level of evolutionarily stable virulence (ESV) for the parasite. However, we also show that the dependence of the ESV on parameter values may be very strong. One possible consequence of this sensitivity is that relationships between transmission and virulence observed across populations need not conform to the patterns expected on the basis of a common (fixed) trade‐off. We discuss possible experiments and implications of our results for the development of virulence management strategies.