In this book, Sagarin and Pauchard argue for the value of observation-based ecology, which they define as "ecology that relies on observations of systems that have not been manipulated for scientific purposes" (pg. 2). Drawing extensively on their own research, the authors argue that ecological science is undergoing a methodological change away from experimentation and theory, and argue that this change will facilitate application of ecology in environmental problem solving.
The book is organized into four parts. The first outlines the historical and current roles observation plays in ecology, the second describes methodological approaches to observation, the third discusses some of the issues and challenges with taking a strictly observational approach, and [End Page 340] the final part explores how observation-based ecology can bridge the gap between scientists and the broader public. In addition to the authors' perspectives, 1-2 page boxes scattered throughout the book present those of a dozen other investigators.
In Part 1, the authors describe the transition in ecology from natural history observation to experimental and theoretical science through the 20th century. They then argue that the heightened emphasis on experimentation and theory in recent decades has been of limited use for addressing current environmental challenges because the clearest experimental results are obtained at small scales and divorced from the environment. This argument may be valid for many laboratory studies—their example of the current emphasis on funding single-species lab experiments to explore ocean acidification is particularly appropriate and timely—but it overlooks the compelling contributions of large-scale field experiments to several of our most pressing environmental concerns. Examples include the whole-watershed harvesting experiments and lake manipulations that were enormously influential in recognizing and implementing policies on eutrophication, forestry practices, fisheries, and acid precipitation (e.g., Likens et al. 1970; Schindler 1974, 1988; Carpenter and Kitchell 1996); these experiments in turn were motivated by the results of small-scale experiments and field observations.
Part II considers some of the approaches to observation currently available. The first chapter emphasizes the importance of classical natural history observations and discusses ways that multiple senses can be incorporated to promote richer observation. Extending this sensory theme, the second chapter highlights how technology is expanding our observation abilities at both the macro and micro scales, and argues that technology is in part responsible for the increasing reliance of ecology on observational studies. Prominent technologies include remote and automated sensing, molecular methods, and electronic access to data and literature that promote meta-analysis. The authors also mention some of the shortcomings of depending on technology: the lack of uniform access that creates uneven coverage of observations around the globe, the potential ephemerality of collection methods, software capabilities, and data stored on rapidly evolving electronic media, limited access to collected data arising from both institutional barriers and the need for accessible "data translation" software, and the separation of researchers from direct interaction with the ecosystems being studied. I agree with the authors that the last aspect is particularly problematic and creates a curious tension between field and technological observations when trying to carry out effective ecological studies. I was also left wondering, however, why the authors finger field experiments as a leading culprit in the demise of field based observation, given that the best experimental practitioners often log thousands of field hours and make numerous observations, whereas that the emphasis on technology-based observation necessarily divorces ecologists from direct observation of ecosystems.
Part III considers the potential challenges of observational studies from several angles. One issue arising from technologically facilitated observation is the massive amount of data that can be created, and the authors highlight several difficulties that have not fully been solved. These include the simple logistics of storing and accessing the data for analysis. Massive data may also pose the problem of being too powerful, in that influences of variables may be detected statistically, even if they explain fractions of a percentage of the variation of interest, which may distract from identifying core processes. Yet...