2 Simple remote sensing metrics predict forest bird richness across geographically diverse agricultural regions

Thursday, August 6, 2009: 1:50 PM
Grand Pavillion V, Hyatt
Jai Ranganathan , National Center for Ecological Analysis and Synthesis, Santa Barbara, CA
Background/Question/Methods The long-term survival of most tropical biodiversity largely depends on the capability of human-dominated agricultural areas to support native biota. However, the conditions that allow for species persistence in the tropical countryside remain mostly uncertain, especially at the local scales at which landholders make land-use choices. As a result, there is an acute need for inexpensive and quick methods for estimating the conservation value of agricultural landscapes across regions. We addressed this concern by asking whether two easily-calculated satellite-derived measures rarely used in conservation (wetness and brightness) correlated with avian species richness at local scales (< 100 ha) in widely disparate regions. We also tested whether a more commonly used metric, normalized difference vegetation index (NDVI), correlated with richness. We compared two highly heterogeneous agricultural landscapes composed of remnant tropical wet forest and complex crop mixtures: one in the Coto Brus Valley of southern Costa Rica and one in the Western Ghats Range of southwest India.

Results/Conclusions Despite the large differences in avian species composition, cropping patterns, and native vegetation between the two regions, we found very similar correlation patterns between them. Specifically, in both agricultural landscapes, the richness of forest-affiliated bird species strongly correlated with wetness and brightness, two metrics that themselves correlated with tree cover. Further, the strength of the correlation between species richness and the satellite-derived metrics increased similarly with area from 10 to 75 ha in both regions, suggesting that forest species communities are responding to their environments at similar scales. In contrast, NDVI, which is frequently associated with productivity or biomass, did not correlate with species richness. Our findings demonstrate the promise that wetness and brightness hold for evaluating the conservation value of tropical countryside, across a wide variety of regions, where traditional field assessments are impractical.