Book Reviews Ecology, 82(1), 2001, pp. 298±299 q 2001 by the Ecological Society of America ACRITICAL EVALUATION OF RESEARCH TECHNIQUES IN ANIMAL ECOLOGY In Chapter 4, David Garshelis provides the best review I Boitani, Luigi, and Todd K. Fuller, editors. 2000. Research have read on habitat evaluations. This is ``must reading'' for techniques in animal ecology: controversies and conse- all ecologists, especially wildlife managers and conservation quences. Methods and Cases in Conservation Science. Co- biologists. Garshelis makes a special effort to point out the lumbia University Press, New York. xxxii 1 442 p. $75.00 delusions in habitat evaluations by inappropriate measure- (cloth), ISBN: 0-231-11340-4 (alk. paper); $32.00 (paper), ments and use of the terms of ``use,'' ``selection,'' ``prefer- ISBN: 0-231-11341-2 (alk. paper). ence,'' and ``importance.'' Garshelis points out that habitat evaluations are often ¯awed because they do not take into Research techniques in animal ecology: controversies and consideration sex-age, social status, time of day, season, year, consequences is just that, a critical examination of some of and they often group animals rather than look at individuals. the most commonly used means of assessing animal popu- Considering all the ¯aws and problems in evaluating habitat lations. Each chapter identi®es limitations, common misuses, use by animals, I am skeptical that most habitat evaluations and possible solutions for gathering, analyzing, and inter- currently published have much relevance to ecology. The es- preting data in a variety of research areas. The edited volume sence of this chapter can be summarized by Garshelis's qoute contains 11 chapters and results from a workshop held at the of E. M. Kirsch: ``Unfortunately, proximate habitat features Ettore Majorana Centre for Scienti®c Culture in Erice, Sicily may not indicate habitat suitability, nor do they reveal the in December 1996. The preface starts out by reminding us possible selective pressures that in¯uence habitat selection in that ecology is rooted in the scienti®c method, but as Charles a system. One must measure components of ®tness, determine Krebs points out in the ®rst chapter, a paradigm that is often factors that in¯uence ®tness, and relate ®tness and factors overlooked. Krebs appropriately starts off with a simple, in¯uencing ®tness to habitats or habitat features.'' clear, and concise review of the scienti®c method and the role John Litvaitis starts Chapter 5 with the question ``Why of hypothesis testing in advancing our understanding of an- study food habits?'' By the end of the chapter, neither I nor imal ecology. Krebs de®nes and distinguishes among key the author know the answer. It seems that nearly every ®eld terms such as theory, hypothesis, models, experiments and ecologist studies food habits of their beast at one time or facts. We are left with several key recommendations such as another, but there is little evidence that food habits ever an- stating working (not null) hypotheses, removing value judge- swers any important ecological question. Supposedly food ments and biases in research design, using statistical esti- habits should explain something about population parameters, mation more than statistical inference, and addressing eco- competitive interactions, or optimal foraging, but results of logically signi®cant and not trivial problems. food habits studies have never contributed much to any of In Chapter 2, Dennis Murray and Mark Fuller cite over these. Hundreds of papers have been published on food habits 300 references to provide a critical review of the effects of of animals and probably an equal number on food addition marking on the biology of vertebrates. An assumption that is studies, yet the mammal, bird, and wildlife literature is in- often made in marking animals for identi®cation is that mark- undated with a plethora of more of the same every year. ing has no signi®cant effect on the animals, an assumption Because of all the problems and limitations of food habits that lacks rigor in testing for most marking techniques and analyses, the author reminds us that the utility of the food species. Fish tags, ®n mutilation, radiotransmitters, leg band habits analysis should be reconsidered before any such study color, neckbands, and toe clipping have been shown to have is conducted. signi®cant effects on animals. The authors provide guidance Population ecologists seek to explain why some animals on the best techniques to use and how to evaluate their ef- are rare whereas others are common, as well as what accounts fectiveness in achieving research goals. for observed changes in density. ``Detecting stability and Roger Powell (Chapter 3) provides a thorough analysis and causes of change in population density'' is the subject of critique of the various methods used in estimating home range Chapter 6 by Joseph Elkinton. Elkinton provides us with a size and space use in animals. The chapter focuses heavily simple and easy-to-follow summary of analysis of data for on techniques used to estimate how rather than why animals time series and mortality and survival, and detection of de- use space. Eleven pages are devoted to reviewing methods layed density dependence. As in all chapters in the book, the for estimating home ranges and only three pages to territories. author identi®es the limitations of various methods that ecol- In that I think territories are much more interesting than home ogists use to study dynamics of populations making this chap- ranges, I would prefer to see these priorities reversed. I was ter must reading before doing ®eld studies in population ecol- disappointed that Powell misrepresented my paradigm on ter- ogy. Elkinton's advice to those ``... embarking on such studies ritoriality in female mammals by citing an opposing example is to maintain a healthy skepticism of all the techniques and of territoriality in male birds. But other than that, Powell is to take a multipronged approach.'' one of the experts on evaluating methods for estimating home In my research design class, I teach students to avoid re- ranges and should be commended for his treatise of this sub- search that involves terms such as ``census,'' ``monitor,'' ject. ``count,'' and so forth. After reading James Gibbs' Chapter 298 January 2001 BOOK REVIEWS 299 7, ``Monitoring populations,'' now I know why. Like all chap- to ecology than most ecologists will need to know. Four pages ters in this book, Gibbs' starts with the grandiose expectation on grooming and four more pages on spacing in cats may be of solving some major dilemma in ecology. By the end of overkill. The chapter is extremely thorough, encompassing the chapter we are even more skeptical than when we started everything from ethograms and dominance to social structure, that monitoring populations tells us anything meaningful territoriality, and spacing. An excellent treatise is provided about population ecology. Monitoring of populations is some- on behavioral sampling and analysis of observational data. thing that almost all wildlife biologists do at one time or In the ®nal chapter, Fabio Corsi et al. describe how to model anotherÐthe utility of their efforts is often questionable. species distribution using GIS. This is currently a hot area Gibbs reviews the various methods used in monitoring pop- of research and application of a new technique that is not ulations, pointing out their weaknesses, of which there are without problems, but nevertheless useful. The authors review many, as well as their strengths. Considering all the problems the terminology and emphasize the importance of de®ning in assessing the validity of censuses, animals ecologists are habitat before providing a synopsis of what a GIS model is reminded to ``... establish explicit and well-reasoned moni- and how it is used, the biological and statistical assumptions, toring objectives before the initiation of any monitoring pro- use of scale, source of errors, and validity of assessment. The gram.'' authors conclude that GIS is a valuable tool and when used Chapter 8 ``Modeling predator±prey dynamics'' by Mark properly can contribute signi®cantly to our understanding of Boyce, and Chapter 9 ``Population viability analysis: data animal ecology. requirements and essential analyses'' by Gary White provide The book is produced in a comfortable format, easy to read, mathematical treatments and models of population parame- with good illustrations. The chapters are well referenced, with ters. Like Galileo, Boyce insists that the book of predator- several citations in 1999. The book is available in paperback prey dynamics be written in mathematical form. Boyce re- and affordable for students. The editors have done a remark- views the pros and cons of noninteractive models, true pred- able job at keeping uniformity among the chapters with each ator-prey models, and stochastic models, admitting that all following a similar outline of introduction and review, crit- models are wrong, but still may be useful. Perhaps the best icism, consequences, and recommendations. I highly rec- advice Boyce offers is for wildlife ecologists, regardless what ommend this book to seasoned as well as new researchers model they use, to focus their research on testing ecological working in any area of animal ecology. This would be an principles. When inconsistencies emerge between the models excellent book for a graduate seminar or a graduate orien- and ®eld observations, the solution is not to reject the model tation or research design course in ecology or ®eld methods. or modelers, but to ®x the models. White provides similar Because the book is wrought with skepticism of methods, advice in his assessment of population viability analysis techniques, procedures, models, and other assessments in (PVA) models. This chapter focuses on computer simulation ecology, it makes us wonder what we have actually accom- models to estimate population viability via numerical tech- plished in ecology over the past 50 years.