New research identifies specific combinations of ecological traits that may put some mammal species in greater jeopardy of extinction.

by Steve Carr

As the human population continues to grow and resource demands soar, biodiversity conservation has never been more critical. Researchers Ana Davidson, Marcus Hamilton, Alison Boyer, and James Brown in the UNM Biology Department, and collaborator Gerardo Ceballos at the Instituto de Ecologia, Universidad Nacional Autonoma de Mexico (UNAM), have studied extinction in mammals through multiple ecological pathways and published the findings in the Proceedings of the National Academy of Sciences.

The research represents an important advance and is vital to understanding the causes of extinction risk in mammals. It also goes beyond previous analyses on extinction risk by identifying specific combinations of ecological traits that cause some species to be at greater risk than others.

"One-quarter of all mammals are in danger of extinction and over half of all mammal populations are in decline, making it critically important for scientists to identify the characteristics of species that make certain ones at greatest risk," says Davidson.

Using a new database of nearly 4,500 mammal species, out of a total of more than 5,400 known mammals, Davidson and colleagues are using a novel methodological approach--decision-trees--to determine different pathways to extinction in mammals and provide simple rules of thumb that can be used for guiding conservation practice. They are discovering that extinction risk varies widely across mammals and that all kinds of mammals, across all body sizes, can be at risk depending on their specific ecologies.

Decision trees predict outcomes of interest, in this case extinction or survival, based on the nested relationships between predictor variables. The models are designed to identify non-linear, context-dependent associations among a suite of correlated predictor values. They require fewer assumptions and don't assume a specific distribution of predictor values. "Decision tree models offer an alternative to traditional methods for modeling complex ecological data and they are often more accurate for predicting complex outcomes like extinction" says Boyer.

In the research, they discovered that although large mammals are well-known to be at risk, 40 percent of all smaller mammals below 5.5 kg are also at risk of extinction. This was a particularly significant finding as 75 percent of all mammals are smaller than 5.5 kg. Yet, conservation efforts worldwide tend to focus primarily on large, charismatic species, such as jaguars, pandas, elephants, and polar bears.

Mammals with certain ecological traits such as small geographic range, low population density, slow life history, and large body size are known to be at risk of extinction.Davidson and colleagues have taken this knowledge to the next step by identifying how these kinds of traits interact to make some species at greater risk than others. For example, species with small geographic ranges are known to have a proportionally greater risk than those with larger geographic ranges. Because it is not enough to estimate risk from a single variable, this study identifies how traits like the size of the geographic range interact with combinations of other traits such as reproductive rate, population density, and social group size to cause variation in risk.

Davidson and colleagues also have identified other traits not commonly thought to be important predictors of extinction risk, such as living in trees versus living below ground. These kinds of insights highlight the urgent need for more information on the basic natural history of most species, which is still poorly known even for the most well studied groups like mammals, yet, is essential to achieving conservation goals.

Unlike previous studies, Davidson and colleagues also identify what is "small," "large," "fast" and "slow" by providing a map of extinction pathways that include the ranges of trait values where species are at greatest risk. This map outlining the pathways to extinction in mammals provides tangible "rules of thumb" for conservation practice. Their model also provides lists of species predicted to be inherently at risk based on their ecologies, many of which are have not previously been identified as at risk.

"Since there is extremely limited funding for conservation, it is critical to provide tangible results that help conservationists prioritize their efforts, and that is a key goal of our research," says Davidson.

"This newer statistical approach is much better at extracting information than traditional linear techniques," adds Hamilton. "Many researchers are seeing the usefulness of this tool. It is less restrictive and more flexible, and allows ecologists to work with large, complex data sets."