PS 50-151
Estimating plant diversity from space

Wednesday, August 7, 2013
Exhibit Hall B, Minneapolis Convention Center
Steven D. Warren, Grasslands, Shrublands and Deserts, US Forest Service, Provo, UT
Anke Jentsch, Disturbance Ecology, University of Bayreuth, Germany
Keith D. Olson, Center for Environmental Management of Military Lands, Colorado State University, Fort Collins, CO
Martin Alt, Umwelt- und Bodenchemie, Universität Koblenz - Landau, Landau, Germany

Fundamental to the fields of ecosystem management and ecological restoration is the question, “How many species are present, and why?” There is increasing evidence that heterogeneous disturbance facilitates habitat heterogeneity and, thus, biodiversity. To test the relationship, a project was conducted on a 16 km2 portion of an army training facility in Bavaria, Germany.  One hundred 1-ha plots were inventoried; different types of land disturbance were recorded and mapped, along with the resulting habitat patches. Plant species in each plot and habitat patch were also recorded.


Statistical analyses revealed that the number of plant species present per plot was significantly correlated with the number and diversity of disturbance types, and with the number of resulting habitat patches, i.e., habitat heterogeneity. Processing of an IKONOS satellite image (4m resolution) of the study area yielded 168 potential measures of spectral heterogeneity. Statistical analyses revealed that some measures of spectral heterogeneity were well-correlated with the heterogeneity of disturbances types, the number of resulting habitat patches, and the number of plant species present.  The results suggest the feasibility of roughly estimating plant species diversity using satellite imagery.  With that capability, it should be increasingly possible to (1) identify probable biodiversity hotspots, (2) compare plant species richness of different locations, and (3) monitor trends in plant biodiversity over time. The use of satellite technology may reduce dependence on, but not eliminate more expensive ground measures of biodiversity based on expert knowledge of species identity. It may also reduce the cost and help optimize the process of site selection for on-the-ground ecological restoration efforts.