Relating spectral and species diversity through rarefaction curves

Abstract

Rarefaction represents a powerful analytical approach in ecology for estimating the expected number of species within a given study area from local (α‐diversity) to regional (γ‐diversity) scales. From a landscape perspective, rarefaction curves are directly related to the environmental heterogeneity of the area sampled. The greater the landscape heterogeneity, the greater the expected species diversity. Therefore, remotely sensed images may potentially be used for predicting species diversity through the indirect method of analysing local spectral variation. The aim of this study was to test whether spectral variability can be used as a proxy for species diversity, from local to regional spatial scales. A total of 977 sampling units, each 50 m×50 m, were selected within the Asciano district (Central Italy) following a stratified random sampling. Each sampling unit was manually classified according to the first level of the Corine Land Cover classification legend. Data on plant species composition were collected in 10 m×10 m plots located within 98 random sampling units. The normalized difference vegetation index (NDVI) was calculated from a QuickBird image, and quantized into 8‐bit data (256 digital numbers, DNs) for building spectral rarefaction curves. Only those plots falling within the QuickBird image were used, which had the effect of reducing the thematic legend to two classes: crops and seminatural vegetation. Species and spectral rarefaction curves were then constructed for each land cover class. Rarefaction curves based on species and spectral properties showed similar results, that is a significantly different number of accumulated values given the same sampling effort for the two classes considered. The results of this study suggest that the shape of the spectral rarefaction curves may be an indirect indicator of environmental diversity, and thus may have potential for predicting biodiversity from local to landscape scales.

Acknowledgements

We thank three anonymous referees for insights on a previous version of this manuscript. The support of the Editor Timothy Warner was crucial during the reviewing process. Marco Ferretti, Federica Baffetta, Lorenzo Fattorini and Michela Marignani helped us in developing and managing the field sampling. Giovanni Bacaro provided useful comments.