Spectral variation versus species beta-diversity at different spatial scales: a test in African highland savannas

J Environ Monit. 2010 Apr;12(4):825-31. doi: 10.1039/b921835a. Epub 2010 Feb 16.

Abstract

Few studies exist that explicitly analyse the effect of grain, i.e. the sampling unit dimension, on vascular plant species turnover (beta-diversity) among sites. While high beta-diversity is often a result of high environmental heterogeneity, remotely sensed spectral distances among sampling units may be used as a proxy of environmental gradients which spatially shape the patterns of species turnover. In this communication, we aimed to (i) test the potential relation between spectral variation and species beta-diversity in a savanna environment and to (ii) investigate the effect of grain on the achieved patterns. Field data gathered by the BIOTA Southern Africa biodiversity monitoring programme were used to model the relation between spectral variation and species turnover at different spatial grains (10 m x 10 m and 20 m x 50 m). Our results indicate that the overall fit was greater at the larger grain size, confirming the theoretical assumption that using a lower grain size would generally lead to a higher noise in the calculation of species turnover. This communication represents one of the first attempts at relating beta-diversity to spectral variation, while incorporating the effects of grain size in the study. The results of this study could have significant implications for biodiversity research and conservation planning at a regional or even larger spatial scale.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Biodiversity*
  • Environmental Monitoring
  • Geography / methods*
  • Namibia
  • Plants*