Does using species abundance data improve estimates of species diversity from remotely sensed spectral heterogeneity?

Abstract

Abstract Different approaches for the assessment of biodiversity by means of remote sensing were developed over the last decades. A new approach, based on the spectral variation hypothesis, proposes that the spectral heterogeneity of a remotely sensed image is correlated with landscape structure and complexity which also reflects habitat heterogeneity which itself is known to enhance species diversity. In this context, previous studies only applied species richness as a measure of diversity. The aim of this paper was to analyze the relationship of richness and abundance-based diversity measures with spectral variability and compare the results at two scales. At three different test sites in Central Namibia, measures of vascular plant diversity was sampled at two scales – 100 m2 and 1000 m2. Hyperspectral remote sensing data were collected for the study sites and spectral variability, was calculated at plot level. Ordinary least square regression was used to test the relationship between species richness and the abundance-based Shannon Index and spectral variability. We found that Shannon Index permanently achieved better results at all test sites especially at 1000 m2, Even when all sites where pooled together, Shannon Index was still significantly related with spectral variability at 1000 m2. We suggest incorporating abundance-based diversity measures in studies of relationships between ecological and spectral variability. The contribution made by the high spectral and spatial resolution of the hyperspectral sensor is discussed.