Foraging for Survival: Yearling Baboons in Africa

Foraging for Survival: Yearling Baboons in Africa. STUART A. ALTMANN. The University of Chicago Press, Chicago, 1998, xii + 608 pp., cloth $70. (ISBN 0-226-01595-5.)

I have not so much reviewed this book as studied it carefully, it is so full of information. Altmann gives a detailed and understandable description, with examples, of the structure he uses for modeling optimal foraging. This thoroughness will be extremely useful for both students and researchers. He is careful to describe, explain and justify inescapable shortcomingsplaces where fudge factors were used because no real data exist. He has a good perspective on reality and helps the reader to keep perspective. However, this also makes the few topics on which he is a bit dogmatic stand out in clearer relief. Being a wildlife nutritionist, I am in a position to argue for lightening up where he is strict and tightening up where he has been more lenient.

The book has ten chapters. Altmann starts out with a description of his field methods and how to determine diet diversity, and then provides a thorough discussion of optimal diet calculations based on linear programming. He then compares real diets to ideal diets on a whole year basis, followed by a discussion of individual differences among his study subjects. Chapters 8, 9, and 10 are discussion: diets and fitness, selectivity, and why baboons have been such a successful species.

This book will be both inspiring and inhibitory to future, similar work. The demand that only gramsseen-eaten be used in the model imposes a real hardship and is not completely necessary. The only paper referenced by Altmann that had previously examined this question (Kurland and Gaulin, 1987) used a tiny data set gleaned from the literature. They admit in the paper that time spent feeding on leaves directly predicts grams consumed. They found that fruit intake is underestimated by 30%. I suspect that the fruit weights included the seeds. which normally pass straight through the animal and should not be included in the weight (seed feeding time itself might overestimate intake because of the varying handling time involved in extracting seeds). I do not dispute that feeding time overestimates the intake of insects; it is difficult to separate search time from feeding time. However, if one is studying an animal that picks a fruit and pops it into its mouth, as leaves are commonly eaten, time probably predicts grams of intake good enough.

Altmann admits that, whereas the intake of corms (time-consuming to harvest) and wood/bark (minor foods) is overestimated by time spent feeding, "Other major food classes ranked nearly the same by time as by mass" (p. 170). am not saying that it is not worth trying to determine actual grams of intake; if at all possible, it should be done. However, one should not abandon a nutritional ecology project just because time spent feeding is the only measure consistently possible.

Altmann also implies that there may not be statistical methods for comparing "nutrient intake levels that were estimated from them [bout durations and rates] via sums of products of means of random variables . ." (p. 235). I have to admit that in this situation human and animal nutritionists use simple t-tests and ANOVAs, using third-level interactions as the error term, despite its violating strict statistical rules. Altmann's decision to call intakes with nonoverlapping standard error intervals "distinguishable" instead of "significantly different" is a very conservative solution.

My third topic for comment is the nutrient requirements and nutrient analyses. The way Altmann "guestimated" nutrient requirements is as good as any other, given the current state of knowledge. The National Research Council is currently updating the book titled "Nutrient Requirements of Nonhuman Primates," but it will not be available until the end of this year.

The description of nutrient analyses performed on the food samples needs additional explanation-even the paper referenced is incomplete. In particular, the fiber analysis description is incomplete. I suspect the old crude fiber method was used because the fiber levels look low. Crude fiber only measures at most half of the true fiber present. It is difficult to compare Altmann's chemistry to other data sets without more details on the choice of methods used. His labs should have provided more complete descriptions.

Altmann's nutritional advisors also did not properly explain the carbohydrate fraction they gave him. The fraction determined by difference is total nonstructural carbohydrates (TNC), not the readily water soluble carbohydrates (WSC), though WSC is a subset of TNC. The WSC fraction is the sweet, monosaccharide sub-fraction, important in selectivity for some animal species (and the fraction reported by Hausfater and Bearce, 1976). The TNC fraction (the only fraction Altmann has) also includes the less readily soluble starches and gums, both potentially digestible but not necessarily sweet.