By Alain F. Zuur, Elena N. Ieno and Graham M. Smith, New York, Springer, 2007, xxvi + 672 pp., £54.00 or US$84.95 (hardback), ISBN 978-0-387-45967-7
As stated in the preface, finding a suitable and satisfactory model to explain the underlying patterns in ecological data is always a big challenge. The book tackles this difficulty and provides a ‘down to earth’ guide to various data analysis techniques, which will address the statistical needs of the great majority of contemporary ecological and environmental researchers. The authors’ enthusiasm for their subjects shines through this book. Across a wide diversity of ecological disciplines, traditional statistical methods blended with interesting examples are used to illustrate problems and analytical approaches to practical data sets. The book aims to give readers sufficient information to apply statistical methodology in a correct and useful way.
The book contains 37 chapters, and can be broadly divided into two parts. Part I presents a collection of applied statistical techniques mainly of interest to ecologists. The progression of topics in Part I can generally be segregated into four categories. (1) The first four chapters provide essential background material for data management and exploration. (2) The regression chapters (5 to 9) start with the fundamentals of simple regression and clearly describe the meaning and application of the technique, then progressing naturally through to advanced regressions, such as generalised linear modelling, generalised additive modelling and mixed modelling. The issues related to the whole modelling framework are also presented, with emphasis on deciding on the amount of complexity to allow in the model and on selecting a suitable number of predictors. (3) Chapters 10–15 concentrate largely on what might loosely be described as multivariate analysis. It covers the most important multivariate methods in terms of current use, including principal component and redundancy analysis, correspondence analysis, discriminant analysis and principal coordinate analysis. (4) Chapters 16–19 give general discussion of time series analysis and spatial statistics. In general, each chapter deals with a particular technique by presenting the objectives and problems behind it, some definitions, a short overview of the method, technical aspects and illustrative examples.
Part II (the remaining chapters) presents 18 case studies covering a wide variety of ecological subjects. The data sets for most case studies are available on the web (www.highstat.com). This is not an overly technical part as the more technical details are mostly presented in Part I, so they can be referred to when required. Each case study is not only related to a single method, but also requires the use of various methods. Each case study generally provides a discussion of the pros and cons of a given type of analysis and furnishes the reader with pertinent information concerning the conditions under which a method is preferred, mentioning alternatives. These case studies are very nicely done, although at times they make the text a little disjointed, they are generally very helpful and are interesting in their own right.
The book is written with a refreshing style and manages to demonstrate the ideas behind statistical analysis and modelling without relying heavily on mathematical methodologies. Graphs and tables are used extensively to illustrate the presentation of data and interpretation. The absence of exercises may be considered as a disadvantage for those who wish to use the book for self-tuition, and it is slightly disappointing that recent developments in Bayesian statistics and high-dimensional regression methods are not covered, when these could be of great practical use in some ecological applications.
In summary, the book meets its aims, covering a wide range of statistical techniques and dealing with many situations that are encountered in ecological statistics. This is an excellent, nicely presented and very readable book. I would highly recommend it to numerate researchers and students interested in environmental and ecological data analysis.