Understanding human impacts on habitat quality: Interoperability of categorical and continuous approaches to landscape and habitat quality analysis.

Representations of landscape, such as the patch-corridor-matrix (Forman 1995) and the variegated model (McIntyre and Hobbs 1999) based on categorical maps, allow for the quantification of the main alterations of land cover structure (composition and configuration) at patch, class and landscape levels. Such quantifications, which can be obtained through both “global” (patch based) and “local” (pixel based) analytical approaches, yield important information on the “structural quality” of the landscape and of its elements. Yet they are per se blind to habitats internal characteristics (McGarigal and Cushman 2005). These are determined by continuous attributes (e.g., physical, biotic) often related to resource availability, to which organism ultimately respond according to their specific perception limits. We present the results of a multi-phase investigation on the effects of habitat fragmentation on habitat quality across multi-taxa perception scales. We preliminarily analysed the complementarities among discrete methods (e.g. landscape indices, morphological spatial pattern analysis, landscape mosaic analysis) (Mairota et al 2013), and then explored the potentials for their interoperability with continuous approaches. Measures of spatial and patterned variance (sensu Wiens 2000) derived from continuous, mainly remote sensing data, proved useful as proxies of habitat quality characteristics in connection with biodiversity surrogates (Mairota et al. 2015a). They were also informative on the non-random nature of the focal habitat distribution within the landscape, and suitable to demonstrate the non-interchangeability of habitat patches (Mairota et al. 2015b). Field data analyses (ordination) further support these results as they allow for the (qualitative) observation of the role of landscape structure in determining species assemblages and distribution (Labadessa 2014).