Untangling Coevolutionary History

Abstract

Organisms in nature exist in complex environments where they interact with a diversity of other species. Just as interactions between species are fundamental to ecological studies, they also are integral to understanding many macroevolutionary patterns of diversification. Chief among these are the coevolutionary patterns apparent between closely interacting groups, such as hosts and parasites, or mutualists. A variety of historical processes play a role in the diversification of interacting taxa. These processes occur on a variety of scales, ranging from the intracellular to the community level, and across a diverse array of taxa, ranging from viruses to vertebrates. Recent phylogenetic studies reveal that the phylogenies of interacting taxa are often very similar, sometimes identical. This similarity is most often generated by repeated instances of cospeciation, the simultaneous speciation of interacting taxa. Alternatively, a number of processes can reduce the congruence between phylogenies of closely interacting taxa. These processes include switching, duplication, sorting events, and failure of one lineage to speciate in response to speciation in the other lineage. Recent cophylogenetic studies have revealed a diversity of patterns, ranging from complete congruence to no significant similarity between phylogenies of interacting taxa (Page, 2003). Coevolutionary biologists can help provide explanations for this diversity by integrating cophylogenetic studies with knowledge of the ecological and physiological mechanisms underlying species interactions. Research on coevolution often merges ecology, systematics, population genetics, physiology, and molecular biology. At the 2002 joint Society of Systematic Biologists (SSB)/Society for the Study of Evolution (SSE) annual meeting we had an opportunity to bring together coevolutionary biologists working on a diverse array of taxa. These workers have summarized their presentations in this issue of Systematic Biology. We are grateful to the SSB for providing support for the symposium and its publication. Our goal has been to inspire comparisons across studies to understand the processes responsible for different patterns of coevolutionary history. As background for this compilation of empirical studies, we review here the processes believed to be important in coevolutionary history (see also Page and Charleston, 1998; Page, 2003). For ease of description, we use hosts and parasites to illustrate these processes, but any pair of interacting taxa could be used. We provide specific examples from a single host-parasite system consisting of pigeons and doves (Aves: Columbiformes) and their parasitic lice (Insecta: Phthiraptera). The first process, cospeciation, is the simultaneous speciation of hosts and parasites. Repeated instances of cospeciation generally increase the congruence between host and parasite phylogenies; cospeciation is usually inferred directly from such congruence. For tightly interacting organisms, some workers consider cospeciation to be a null model of sorts, because forces isolating host populations also tend to isolate their parasite populations. An example of cospeciation can be seen in body lice (Physconelloides) found on New World doves (Fig. 1). One of the processes that reduces the congruence between host and parasite phylogenies is host switching. Incomplete host switching is simply an expansion of a parasite's host range, such as when a parasite colonizes an additional species of host (Fig. 2). Incomplete host switching reduces the parasite's level of host specificity; where a parasite species once occurred on a single host species, it now occurs on two host species. Host switching can be completed either by extinction of the parasite on the original host, speciation between the parasite populations on the new host and those on the original host, or both. Incomplete host switching has occurred in body lice (Physconelloides) of small New World ground doves (Fig. 2). Sorting events are another group of factors that, in combination with other processes, can reduce the similarity between host and parasite phylogenies. Sorting can involve extinction, in which a parasite lineage is lost from a host lineage, or the boat, in which a parasite lineage is missing from a host lineage during founder event speciation. An apparent sorting event has occurred in the wing lice (Columbicola) of the New Zealand pigeon (Hemiphaga novaeseelandiae). Wing lice have never been found on New Zealand pigeons despite intensive sampling (Paterson et al., 1999; Clayton, unpubl. data), whereas the closest relatives of the New Zealand pigeon do have wing lice (Adams, 2002). Yet another factor that can reduce phylogenetic congruence is duplication, which occurs when a parasite lineage speciates in the absence of host speciation. Duplication can occur via sympatric speciation of parasites or via allopatric speciation of parasite populations across the geographic range of a host. Although there are few examples of recent duplication events in lice (Johnson and Clayton, 2003), a duplication event seems to have occurred in the body lice of the Band-tailed