Abstract

Two major hypotheses of host-parasite interactions have been proposed to explain cases where hosts do not defend themselves against parasites. The evolutionary lag hypothesis suggests that there has been insufficient time for a host response to evolve, whereas the evolutionary equilibrium hypothesis proposes that host defence does not evolve because it carries costs that outweigh the benefits. We tested potential benefits and costs of host defence in an unusual interaction, between a freshwater fish, the European bitterling, Rhodeus sericeus, and live unionid mussels that are used as hosts for the fish's eggs. We found a significant reduction in ventilation rate of mussels that were incubating bitterling embryos, which became more severe with an increasing number of embryos. We tested the hypothesis that the risk of ejecting a mussel's own larvae while ejecting bitterling embryos has constrained the evolution of a host response. This predicts that brooding female mussels would retain more bitterling larvae than males or nonbrooding females. This prediction was not supported: brooding female mussels contained as many larvae per accessible gill as, and fewer in total than, males or nonbrooding females. In summary, based on the costs and benefits of ejection that we measured, we found no evidence in support of the evolutionary equilibrium hypothesis. However, other differences between mussel species, such as gill structure as measured in this study, ventilation rates and differences in the distance eggs are lodged into the gills, may contribute to differences in egg ejection rates. (c) 2005 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.