Interannual variability of NDVI and species richness in Kenya

Abstract

A central task in community ecology is to determine biophysical controls of species richness patterns. Most multivariate models relating environmental characteristics to species richness were developed with sparse point sampling from meteorological stations. However, satellite remote sensing provides an alternative, and in some ways potentially superior, means of estimating appropriate environmental factors and thereby improving predictions of species richness. The frequently used surrogate measures for species richness patterns are net primary productivity (NPP) and actual evapotranspiration (AET). Local spatial variability of NPP and AET, which indicates spatial heterogeneity, is hypothesized as another influence on species richness. The Advanced Very High Resolution Radiometer (AVHRR) derived Normalized Difference Vegetation Index (NDVI) has been shown to be related to NPP and AET in many vegetation types. We examined the relationship between interannual NDVI parameters and species richness of vascular plants, large mammals and birds. Statistical analyses revealed that at small spatial scales (10 x 10 km) higher average NDVI results in lower species richness of mammals and plants, whereas standard deviation of maximum NDVI and coefficient of variation correlated positively with species richness. Conversely, at large spatial scales (55 x 55 km) bird species richness increases as average NDVI increases, whereas higher standard deviation and coefficient of variation result in lower species richness. Thus, NDVI parameters appear to represent environmental factors influencing species richness. In other words, by utilizing remote sensing, our understanding of the spatial variability of species richness has been improved.