In a bid to bring more direction to a somewhat chaotic field, the US National Science Foundation (NSF) has launched an ambitious plan to increase biologists' use of mathematical and statistical techniques.
The agency wants researchers to build these techniques into their grant proposals. The scheme's initial budget will be at least $3 million, say officials, and future budget requests could boost agency-wide support to $100 million.
It is hoped that the initiative will provide tools for analysing the large data sets that will come from research programmes such as the planned National Ecological Observatory Network, a system of long-term monitoring facilities. The eventual goal is to find definitive answers to questions on population dynamics, dispersal of species over large areas, and the impact of ecological variables.
NSF officials say the fund will initially target neurobiology, animal behaviour, ecology and evolutionary physiology. New numerical techniques could be applied at every level from genes to ecosystems.
The NSF has sent out notices about the initiative, known as Quantitative Environmental and Integrative Biology. The agency's biological sciences directorate and its mathematics and physical-sciences directorate will use their core programmes to encourage the funding of research incorporating the new approach.
The notice says that the time is ripe to use existing mathematical tools and develop new approaches to speed up progress towards understanding and predicting environmental and integrative biology.
Principal investigators are being encouraged to train undergraduates and graduates in the new techniques, with a view to them becoming involved in the research. NSF officials see this as a way to create a much-needed generation of scientists skilled in both biology and mathematics.
Earlier this month, at an NSF-sponsored workshop at the University of California at San Diego's Supercomputer Center, mathematicians, statisticians and biologists came together to give the agency guidance on the types of project it should fund.
According to Margaret Palmer, director of the NSF's division of environmental biology and an aquatic ecologist at the University of Maryland, the consensus was that the initial targets should be population genetics, ecological system restoration, and animal physiology and behaviour.
People were excited about the potential for progress at the interface of environmental biology and mathematics, says Alan Hastings, an ecological mathematician at the University of California at Davis, who organized the workshop. The most interest, Hastings said, surrounded the issues of randomness and variability.
Scientists have had some success at developing models to analyse population data, for example, with the flour beetle, voles and Dungeness crabs, predicting population parameters and working out what determines species populations.
But difficult challenges remain, particularly in studying species spread over large areas. Many serious ecological questions involve space, such as the spread of invading species and genetically modified traits through native populations, says Tony Ives, a University of Wisconsin zoologist who participated in the workshop.
For wide-ranging populations, data at any one location are likely to be sparse, leading to statistical problems with characterizing populations. One attempt to address this involved a team of mathematicians and ecologists who worked with laboratory populations of the flour beetle Tribolium castaneum .
This produced a seminal paper (see Nature 375, 227–230; 1995) showing how a mathematical model could be used to predict changes in a species where cannibalism leads to wildly differing population growth rates.
At the NSF workshop, members of the team — Brian Dennis of the University of Idaho, Robert Desharnais of California State University at Los Angeles and Shandelle Henson of the College of William and Mary in Williamsburg, Virginia — explained how their techniques can be applied to species in the wild, translating biology into mathematical language.