EVOLUTIONARY PATTERNS OF ALTERED BEHAVIOR AND SUSCEPTIBILITY IN PARASITIZED HOSTS.

Abstract

Adaptation is the usual context for interpreting parasite-host interactions. For example, altered host behavior is often interpreted as a parasite adaptation, because in some cases it enhances parasite transmission. Resistance to parasites also has obvious adaptive value for hosts. However, it is difficult to evaluate the adaptive significance of host-parasite interactions without considering the historical context in which these traits have evolved and if they can be predicted by host (or parasite) phylogeny. We examined the influence of host phylogeny on patterns of altered behavior and resistance to parasitism in a cockroach-acanthocephalan system. A consensus cladogram for cockroach subfamilies was produced from the morphological data of McKittrick. We used this cladogram to predict patterns of altered host behavior in seven cockroach host species. Each species was experimentally infected with a single species of acanthocephalan, Moniliformis moniliformis, a parasite that is transmitted when cockroaches are eaten by rodent final hosts. Activity patterns, substrate choices, and responses to light were measured for control and infected animals. These data were recoded into a behavioral matrix of discrete characters. We determined the most parsimonious distribution of the behavioral characters on the tree obtained from McKittrick's data. We then measured the concordance between the behavioral data and the cockroach cladogram with the consistency index (CI). We compared the observed CI to the expected value based on a randomization of observed character states. For three different models of evolutionary character change, there was no evidence of strong concordance (significantly large CI) between altered host behavior and host relationships. Parsimony analysis of the interior nodes of the phylogenetic reconstruction suggested that unaltered behavior was the ancestral state for most host behaviors. We also compared host phylogeny to a data set on the susceptibility of 29 cockroach species to infection with M. moniliformis. At the species level, there was a significant concordance between susceptibility and host phylogeny. This pattern was consistent with the finding that susceptibility of species varied significantly among different subfamilies. However, at the subfamily level, susceptibility was not strongly concordant with phylogeny. We predict that, given enough time, resistance should be lost in subfamilies that are currently resistant to parasitism. In spite of the potential importance of phylogeny in the evolution of behavior and susceptibility, we found little evidence for phylogenetic effects in this system. We conclude that changes in the behavioral responses of hosts to parasites and, to a lesser extent, changes in susceptibility are more frequent than cockroach speciation events in different cockroach lineages. This finding strengthens the assertion that at least some of the altered behaviors are adaptive for host and/or parasite.