'You could plan a pretty picnic
but you can't predict the weather, Ms. Jackson...'
Ms. Jackson - Outkast
What will the future bring? This is always an important question to ask. Conservation policy makers tend to be interested in the future, as expressed in the term sustainability (Brundtland report, WCED 1987) which centres around our obligations to future generations.
However, the future is not always predictable. Our understanding of biodiversity and ecosystems is very limited. Most of the species on earth have not been described yet, and we know even less about their roles in ecosystems. Furthermore, although there are some exceptions (see Willis et al., 2007 for an overview of long-term data) most of the data used does not go back more than 60 years.
Mathematical models can be used for forecasting the effects on future biodiversity. In order to develop a model, scientists start by simplifying perceived reality through a number of assumptions, such as “all animals know the locations of all the food items” (an assumption behind the Ideal Free Distribution, a model from foraging ecology) or “all species are equally affected by climate change” (an assumption from Williams and Araujo, 2002). Using a number of predictor variables (such as precipitation, temperature or vegetation) and mathematical relationships between these variables and a response variable (such as species distribution) a model can then be used to predict the future state of the response variable. Subsequently, a model is ideally verified against existing observations of the response variable.
A successful model? Observed and predicted range of the red-backed shrike (Lanius collario) in the 1990’s using a species distribution model. Whatever might govern the distribution of this bird species, it is not quite covered in this model… Figure from Araujo et al., 2005
Although some mathematical models are simple and relatively straightforward (if the mathematical hoo-ha is removed), others are quite complex. For instance, whilst the Bioclim model, used for forecasting the effects of climate change on biodiversity, could be compared to a primary school counting exercise, the spectral models used by the IPCC require skilled mathematicians to translate them.
We think that biodiversity-response models are crucial to conservation planning. Policy makers require, at the very least, the best guess of the scientific community as to what the impact of different policy choices will be. Without these predictions, they would be flying-blind, making decisions on the basis of myth and dogma.
However, we also think it is important that biodiversity policy makers remain aware of the limitations inherent in modelling. We want to remind them that our knowledge of biodiversity is very limited, and that mathematical models are sometimes no more than a way to hide the deficiencies in our understanding. Models, especially the less transparent ones, need a thorough verification and review process. Those making policy decisions want simple answers, clear-cut choices, and in this context there is a tendency to attach too much certainty to uncertain findings through modelling. It is crucial that conservation scientists resist this temptation.
Wouter Langhout and Tim Hodgetts
Araújo, M.B., Whittaker, R.J., Ladle, R.J. and Erhard, M. (2005) Reducing uncertainty in extinction risk from climate change. Global Ecology and Biogeography, 14: 529-538.
WCED (1987) Our common future. Available online at http://www.un-documents.net/wced-ocf.htm
Willis, K.J., Bennett, K.D., Froyd, C. and Figueroa-Rangel, B. (2007) How can knowledge of the past help conserving the future? The need for a long-term perspective in biodiversity conservation Philosophical Transactions of the Royal Society, Series B, 362: 175-186
Williams, P.H. & Araújo, M.B. (2002) Apples, oranges and probabilities: integrating multiple factors into biodiversity conservation with consistency. Environmental Modeling and Assessment, 7: 139-151.