The latest report from the Intergovernmental Panel on Climate Change (IPCC) warns that 20-30 percent of plant and animal species are at increased risk of extinction if global average temperatures rise by more than 1.5 to 2.5 degrees Centigrade. Although most scientists agree that global warming presents a significant threat to biodiversity worldwide, some experts criticize that current modeling techniques lack the accuracy and consistency to make authoritative predictions. A recent article in Science magazine, "Predicting Oblivion: Are Existing Models Up to the Task?" explores the current drawbacks and future potentials of modeling climate-related biodiversity loss.
Climate Change Brings a New Challenges to Biodiversity Modeling
Over the past century, humans have changed ecosystems faster and more extensively than during any other period in human history, resulting in an extinction rate that is currently 1,000 times greater than the background rates typical of Earth's past. The World Conservation Union's Red List of Threatened Species is the most comprehensive database tracking the status of endangered species all over the world, but it has no mechanism to preemptively monitor species that are currently doing well, but have a high future risk of extinction due to forecast global warming. Including such species on the Red List would require firm predictions of how many and which species will be threatened, but current models involve high levels of uncertainty.
Climate-Envelope Models
The most widely used model to date is the "climate-envelope model," which analyzes all the places where a species has been recorded and looks for common climatic features in those areas, such as average rainfall or winter temperature. The model then estimates the physical size of future climate envelopes given a predicted increase in global temperature and assuming that smaller envelopes lead to increased extinction risk. This was the model used by an influential study published in Nature in 2004, which figured prominently in the latest IPCC report. It concluded that 15-37 percent of species will be "committed to extinction" based on a midrange warming scenario for 2050.
Room for Improvement
However, many researchers have found reasons to question the accuracy of these models. Some studies have shown that climate-envelope models often fail to make accurate predictions of the present when run with historical data, and others have found that the most frequently-used models yield dramatically different results when using the same baseline data. The question remains as to how to improve current modeling techniques. Some scientists argue that models must integrate more biological factors, such as how quickly a species disperses or how it interacts with other species. Still others insist that averaging the results of many climate-envelope models could improve accuracy, although this approach requires complex calculations beyond the capacity of existing software.
The future of climate-extinction modeling will ultimately depend on the overall goal of the modelers. Do we need to calculate an exact rate of extinction for a particular species, or would it be more useful to predict regional variations in vulnerability for certain species groups? Certainly both approaches have merits, but for the sake of existing conservation strategies, such as deciding how to alter existing protected areas or where to establish new ones, perhaps the latter approach would be most effective.
Full Citation:
Carl Zimmer (17 Aug 2007). "Predicting Oblivion: Are Existing Models Up to the Task?" Science 17(317), 892-893.RELATED LINKS:
Convention on Biological Diversity: Climate Change and Biodiversity
Climate Change and Biodiversity (IPCC Technical Paper V)
The Millennium Ecosystem Assessment
EarthTrends
Data Table: Biodiversity Overview 2005 (pdf)
Biodiversity and Protected Areas searchable database
