The Bactra Review: Occasional and eclectic book reviews by Cosma Shalizi   84

Population Genetics

A Concise Guide

by John H. Gillespie

Baltimore, Maryland: Johns Hopkins University Press, 1998

Darwin's Equations (II)

``Population geneticists spend most of their time doing one of two things: describing the genetic structure of populations or theorizing on the evolutionary forces acting on populations. On a good day, these two activities mesh and true insights emerge.'' So Gillespie begins his guide to what's turned up on the best days since the 1920s, when the works of R. A. Fisher, J. B. S. Haldane and Sewall Wright founded the field. There've been many good days in the last seventy years, and we now know a lot more about evolution at the genetic level, particularly about what happens in small populations and about ``neutral'' mutations, those without effect on fitness. Since (as constant readers may recall) I've already babbled about the Deep Issues touching on population genetics while reviewing Haldane's The Causes of Evolution, and this is a textbook, I'll stick to mundane details here.

Gillespie builds up population genetics in a fairly traditional order, starting with Hardy-Weinberg equilibrium (essentially a result saying that, in the absence of mutation, selection, or finite-size effects, gene frequencies do not change). Next (ch. 2) he introduces genetic drift --- in finite populations, gene frequencies can change merely due to the luck of the draw, individuals becoming parents more or less by chance, without any selective pressures. After that (ch. 3) comes the keystone, selection in randomly-mating populations, including Fisher's ``fundamental theorem of natural selection'' (which almost never applies). Chapter four drops the assumption of random mating, which makes the life of the geneticist harder; out of all the kinds of non-random mating which must be considered, Gillespie picks out inbreeding and semi-isolated sub-populations. Chapter five deals with quantitative genetics, on the assumption that quantitative traits (e.g. beak-length, chest-width of Scots army recruits) are controlled by numerous, independently-inherited and additive Mendelian genes. (The three qualifiers, together with the Central Limit Theorem, give us the approximately Gaussian distribution of such traits, which so impressed nineteenth century writers.) A final chapter considers the much-debated subject of the evolutionary advantage of sex.

When Gillespie says ``concise,'' he means it; the text of the book, including problems and solutions but excluding appendices, is just a hundred and fifty pages. (Many interesting topics, like inclusive fitness and evolutionarily stable strategies, are only briefly touched upon.) Despite this, he covers all the essentials, and demands very little in the way of prior background of his readers --- some knowledge of biological jargon (what a gene is, what Mendel's laws are, what ``heterozygote'' and ``gamete'' mean), competence in ordinary algebra, a little familiarity with statistics. (There is an appendix on the necessary probability and statistics, which probably will work best as a refresher.) Demonstrations are not technically rigorous, but no worse than those found in, e.g., physics textbooks at a similar level, and Gillespie is always clear, precise and readable, in places almost warm. This is, in a word, a pedagogic jewel, worthwhile for anyone interested in evolution with a firm grasp of high school math. (I wonder if the publisher knew what it was doing in making the ratio of cover price to cost of xeroxing so high.)


Disclaimer: I asked for a review copy of this book from the JHU Press, but I didn't get one. I'd like to think this didn't bias me against it.
xvi + 174 pp., black-and-white graphs, bibliography, index of names, subjects and mathematical symbols
Evolution / Genetics
Currently in print as a paperback, US$19.95, ISBN 0-8018-5755-4, and as a hardback, US$45.00, ISBN 0-8018-5754-6. LoC QH455 G565
1 May 1999
Thanks to Nigel Snoad.