Rainfall and river networks prove accurate predictors of fish biodiversity

“It is the first study I am aware of that provides a real quantitative framework for the study of river biogeography,” D'Odorico said.

In their research, the authors merged different sets of existing data from the Mississippi-Missouri river basin, an extremely large region that covers more than half of the United States. This network of rivers springs from the Mississippi River, which cuts down the middle of the country. The triangle-shaped basin stretches from Minnesota to Louisiana and from Montana to New York.

Using one set of data, the researchers were able to identify 824 distinct sub-basins and establish how the rivers within each sub-basin were linked together. Another set of data identified 433 different species of fish living in those sub-basins. A third set of data identified each region’s average runoff, which is the amount of rainfall that ends up in rivers or streams as opposed to water that is soaked up by the ground.

Rachata Muneepeerakul, a postdoctoral researcher at Princeton who received his Ph.D. from Princeton in 2007, is the lead author of the Nature paper. Click here for more information.

The researchers combined all these data and came up with a computer model that accurately predicts how many different species of fish will inhabit any given sector of the river basin. Their research shows that the habitats richest in the diversity of species are areas where multiple streams are close to one another.

“This will help identify which parts of a river basin are ‘hot spots,’ meaning they have more species than others and therefore should receive special care,” said Rodríguez-Iturbe, senior author of the paper.

To create their model for the Nature paper, the researchers disregarded the biological features of the fish in question -- for example, which species might be tenacious predators or which might be well-suited to take advantage of available food in the area. The model tracks how many species will thrive in a given area but does not predict which species. It is what is known as a “neutral” model and thus treats each fish equally.

The lead author of the paper is Rachata Muneepeerakul, a postdoctoral researcher at Princeton who received his Ph.D. from the University in 2007. Co-authors are hydrologists Andrea Rinaldo and Enrico Bertuzzo of the Ecole Polytechnique Federale de Lausanne (EPFL) in Switzerland, and Heather Lynch and William Fagan of the Department of Biology at the University of Maryland. The study was funded by the James S. McDonnell Foundation.