Vegetation Description and Data Analysis
Rating: 
In writing reviews it is best to keep some professional distance; to report what the book tries to say and reflect on the quality and suitability in light of the academic field and competing texts. This is the first and probably only time I will break this rule, just at the outset. This is the second edition of a remarkable text for which I had always hoped to own the original edition. Both the current author and I had the good fortune to have a lengthy friendship with the original co-author, Paddy Coker. It was he who steered my interests in ecology and developed my knowledge of quantitative work (although I doubt he ever agreed with my use of statistics!). Sadly he passed away just two days before our own collaboration was published. Thus it is with a sense of interest, both personal and academic, that I approach this book.
One of the key factors in the first edition was the way in which the work was focussed. There was no attempt to dazzle the audience, rather the opposite. The aim was to inform, to make the complex, simple and to engage the reader and instil in them the love of quantitative ecology that the authors clearly had. Above all, this book demonstrates the need for a practical ecology (witness the field safety guide right at the beginning). This practicality continues in the first chapter which not only provides an overview of the book’s aims and structure but also something of the ways in which the scientific method works and how this is adapted for modern quantitative ecology (from induction and deduction to Bayesian analysis). The chapter also reflects on the problems facing ecologists as they analyse their work. After all, the whole aim of this field of work is to discover what is where and use statistics to narrow the answer to the question, why.
At this point, many texts move to describing methods and the mathematical background. In this case, the author has moved on to the logical step of describing some of the theory of plant distribution by way of environmental gradients and vegetation dynamics. A range of models is presented and critiqued. Although it might appear as a digression into theory it has profound implications for the work we do. The argument is put clearly that unless we understand how vegetation interacts with the environment (and other plants!) then we will not be able to interpret the data we have gathered. I have always liked the debate over the actual existence of ecosystems as a concept but here we see just how important such points are in understanding the distribution of plants. From this point, it makes logical sense to demonstrate how we can record vegetation in the field and the range of techniques that can be used. Once the theory and practice have been understood, we are moved on to considering the statistical techniques that can be used.
As one might expect from earlier chapters, we start not with a discussion about statistical packages, but a look at the basic techniques; a grounding in multivariate analysis. Three main methods are described – confirmatory, exploratory and Bayesian analyses – worked examples show the results that can be obtained using a number of common tests e.g. Spearman and Simpson. From here we move to ordination, starting with an overview of the method to a description of the many techniques that can be used. Each one has an example and an outline of the ways it can be used. Often, the text breaks to consider the issues that might arise from using any given method. This is a substantial chapter as befits a set of techniques so central to the idea of vegetation analysis. Ordination is a statistical analysis but there are other ways of investigating the relationships in vegetation patterns. Phytosociology has has an interesting history with the work of Braun-Blanquet central to this. Braun-Blanquet’s method could be described as subjective – the ordering of quadrats to find similarities. However, it is also possible to use objective or numerical methods such as cluster analysis, and dendrograms. The two, linked, chapters thus describe and critique the range of possible methods in this area.
When the first edition was written, much of the work had to done by hand or with some ‘primitive’ calculators. today, we have computers and software that make all the tasks described here a matter of data input. The penultimate chapter describes some of the programmes and the techniques they can help analyse. Of course, the main point is not using a programme (because anyone can input data) but understanding the data that come out which may well be why this chapter appears so late in the text! A final, brief,chapter discusses where this field might be heading.
It’s easy to be biased here. I knew one of the co-authors as a friend for over 30 years. I knew what he thought about ecology and about statistics and how the two should be put together (and often what he thought when the two were badly put together!). I also heard much about Martin Kent and how their work meshed. However, I put that aside and read the book as I do many others to see what I can learn more about ecology. I come to ecology from the abiotic rather than biotic side and such ideas as vegetation analysis find me good on theory but lighter in practice. I recognise the structure of the text as one designed to help people such as myself get a better grasp of the topic. It is to the current author’s credit that I understood virtually all and appreciated they way in which the layout and tone of the text helped me along. So many ecology texts focus on the nature of the subject in hand without consideration of the people actually sitting in front of the book. As you read this you can see how vegetation analysis should be done – from theory and consideration of method to practical application and limitations. In these days of electronic texts it’s easy to forget the needs of the reader but here it is made clear. Here is an expert ecologist slowly outlining his work to a new generation of students in the hope they get it right. It is a superb text – it should be a standard text in any course on vegetation analysis – it deserves nothing less than the widest readership.
