Bird communities and vegetation structure: I. Correlations and comparisons of simple and diversity indices

Publisher Website
Google Scholar
Abstract
In order to investigate relations between bird community and vegetation structure indices, with a focus on methodological problems, 22 study plots ranging from grassland to old forests were selected. Breeding passerine birds were censused by means of the mapping method. Vegetation structure was assessed by measuring cover values at 12 different heights (0.25 to 32 m). Simple indices (e.g. number of bird species, NRSPEC, and number of layers with vegetation, NSTRAT) as well as diversity values (bird and plant species diversity, BSD and PSD, resp.; foliage height diversity, FHD, and other indices of structural diversity) were calculated. Vegetation structure diversity, but not floristic diversity PSD, was found to be correlated with BSD. However, vegetation structure indices differed in several respects. The much-discussed BSD/FHD correlation held only if structurally different plots (forests and low vegetation) were included in the analysis, but not if the evaluation was restricted to forests alone. The index DT, suggested by Blondel and Cuvillier (1977) proved to be more useful, being more highly correlated to BSD, and more robust as to study site selection. It also offers the advantage of discerning between a vertical (DV) and a horizontal (DH) component. Due to methodological divergencies, it was found virtually impossible to make detailed comparisons, in terms of biological concepts, of the results of other authors and those of the present study, the problem of comparability apparently deserving more discussion than it has received hitherto. The designation “FHD”, esp., is used for numerical values arrived at by quite divergent field methods and computational procedures. It is concluded that simple indices (e.g. NRSPEC and NSTRAT), which are demonstrated to be good predictors of more complex ones (BSD and FHD, resp.), should be preferred as they permit better standardization and easier, more direct interpretation.