Hot whodunnit leaves you cold

Australian rainforests have an unusual distribution. They occupy a large latitude range but exist throughout that range only as patches within tracts of eucalyptus forest. There have been some excellent books describing Australia's rainforests, and David Bowman's book does not seek to rework the material. Instead, it has a clear, single focus. It seeks to account for this widespread, fragmented distribution. This might seem to be rather too specialised a subject to hold out much promise of a good read. However, this is no dull bio-geographical treatise, but an intriguing whodunnit in which Bowman tackles his subject with the enthusiasm of a Belgian detective.

As with all good detectives, his first task is to identify the body. Defining a rainforest is a surprisingly sensitive subject in Australia. Australian ecology is highly politicised, and the word rainforest has been interpreted with the same sort of latitude that politicians use to manipulate the word freedom. To the environmental movement, rainforests have become an icon for their efforts to protect natural ecosystems and their definition is as all enveloping as maternal arms. State forestry departments, which have been bludgeoned into agreeing not to log rainforest, seek a definition that excludes anything they plan to exploit. All this squabbling bodes well for a gripping tale of forensic science but, disappointingly, Bowman pulls the pall back over the subject without giving his professional judgement. It is not, he declares, his aim to champion any view. Instead, he uses the definitions provided by the authors of the research that he selects to review. This is not, it strikes me, a very sound basis for finding general explanations, but a recipe for circular reasoning.

Bowman then sets about examining why rainforest in Australia occurs as scattered patches with abrupt boundaries. The subject is important. These patches harbour large numbers of endemic species, and their conservation presents a significant challenge. Chapter by chapter, Bowman exposes to scrutiny the evidence underpinning theories advanced to explain the distribution of Australian rainforest. Hypotheses promoting soil fertility, water stress, climate, light and temperature are interrogated with scholarly thoroughness and shown to be inadequate general explanations. Using his uniquely Australian conception of what constitutes a rainforest, Bowman shows that they occur across a range of soil types and have wide tolerance of drought, temperature and light regimes.

The denouement comes in chapter eight, where Bowman reveals the true culprit. The evidence that is presented is compelling. Bowman concludes that fire is the most important factor controlling rainforest boundaries across an enormous latitude and altitude range. At this point in most whodunnits there is a surprise when it is discovered that the most obvious explanation for the mystery is far from the truth. How deflating then to discover that, in this case, the explanation for the abrupt boundary between eucalyptus-dominated forest and rainforest is indeed the most obvious one. Fire has certainly been the most widely supported explanation for savanna-rainforest boundaries for the past 20 years. It is a pity that such scholarly effort seems to have been spent building up and then dismantling ideas that have had little life in them for decades. Certainly, any ecologist familiar with recent publications on gallery forest in Brazil or forest boundaries in West Africa will find little to surprise them. Although Australians may have a unique conception of rainforest, there is a wealth of highly relevant modern literature on forests elsewhere in the world that would throw important light on the issues that Bowman discusses.

Although this book is parochial in its perspective, I shall use it in my undergraduate teaching as a fine example of how to make science readable. Bowman writes extremely well, and tells a good story.

Nick Brown is lecturer in forest ecology, University of Oxford.