Tangled Trees: Phylogeny, Cospeciation, and Coevolution

Rating:
Summary: Creating relationships between trees
Review: Parasites are often quite choosy about the species that they infect. In some cases, even slightly different hosts have very different parasites - the connections can be very close. Hosts change, though. They migrate to new ranges and split, through evolution, into new species. The parasites that depend on those hosts must adapt or die.

This creates fascinating relationships between the family trees of the host and parasite species. That is the subject of this book: the study of the parasite's phylogeny, in terms of the host's. (The reasoning can be applied to symbionts or to species with other dependency relationships, as well.) The book is a series of mongraphs by different authors, all discussing various aspects of coevolution. The actual times and places of speciation events are lost to history, so much of the discussion centers on ways to create credible phylogentic histories, and to estimate the probability of any guess being accurate.

About 2/3 of the book is biologically oriented: case studies of various host/parasite relationships, actually families of hosts and related families of parasites. Some of these are very interesting. One uses presence of a single parasite to argue for a close relationship between two rodent clades. Others describe coevolution at different scales, subject to different forces, or in different biological systems. I have to admit, though, that the biology is tangential to my own interests.

The first third of the book is more mathematical, the material that's closer to my own needs. It starts with a basic discussion of the events - branching in the family tree, jumping between branches, and so on. This includes discussion of approaches to take for resolving two trees, scoring mechanisms for evaluating conjectured relationships, and more. It's also a logical next step in bioinformatics. Historically, solved problems turn into steps in larger problems. Once, just getting a sequence was an achievement. Next, the sequence was taken for granted and comparison of two was the achievement. Successively, comparisons became steps in tree-building algorithms, then individual trees became pieces of "consensus trees", statistically graded estimates of history. This study takes different trees as the inputs, and creates histories of coevolution between them.

This would be a good second or third book on coevolution and computational support for the study. Unfortunately, there does not currently seem to be a first book - this just leaps into advanced discussions that assume a large base of knowledge. The references are all to current journal articles, in biology, evolution, and mathematical graph theory. Well, that's a normal part of the progression from a specialist study to a widely-known body of practice. Introductory material may not exist for some years to come.

For now, this seems to be the only book around that describes computational analysis of coevolution at all. This is not for the cut-and-paste programmer, but serious software developers will find some good basics and references.

//wiredweird