A Multiscale Method for Selecting Indicator Species and Priority Conservation Areas: a Case Study for Broadleaved Forests in Lombardy, Italy

Abstract

Our purpose was to define a quantitative and expeditious method to analyze the effects of processes that influence species distribution and abundance at different organizational scales. We considered habitat loss, the breaking apart of habitat patches, and habitat structural alteration critical processes that affect species distribution and abundance. We evaluated the effects of these processes by considering the response of selected indicator species to isolation (landscape scale), patch size and edge effect (patch scale), and habitat structure (plot scale). We used broadleaf forests as our case-study ecosystem and birds as indicator species. Faunal data came from a georeferenced, long-term, breeding-bird database, and environmental data were obtained from field surveys and land-use digital cartography Birds, grouped according to their sensitivity to patch isolation, patch size, edge effect, and habitat structure, indicated how environmental conditions affected species abundance. The most sensitive group to the above-mentioned processes included the Marsh Tit (Parus palustris), Nuthatch (Sitta curopaea), and Short-toed Treecreeper (Certhia brachydactyla). We evaluated the umbrella effect of these species for the conservation of the other co-occurring birds and found that their effectiveness varied according to the criterion used to select sites for protection. Site selection based on indicator species identified forest patches with greater species abundance than an alternative criterion that chose sites based on patch isolation and size, edge effect, and.forest structure. The alternative criterion, although not as efficient as indicator species, may nevertheless be useful and effective for conservation when faunal data are lacking and guidelines for habitat management or restoration are needed. Our method is applicable to other ecosystems and taxa because the processes we considered occur in many ecosystems and may have significant effects on species from all taxonomic groups.

Usefulness of coarse grain data on forest management to improve bird abundance models

“…It was measured by the FRAGSTATS Proximity Index (McGarigal & Marks 1995), modified by Bani et al (2006). The FRAGSTATS Proximity Index of a patch equals zero if there are no other patches of the same category (in our case broadleaf forest patches) within the proximity radius and increases as the area occupied by surrounding patches increases and/or their distances from the focal patch decrease.…”

“…The FRAGSTATS Proximity Index of a patch equals zero if there are no other patches of the same category (in our case broadleaf forest patches) within the proximity radius and increases as the area occupied by surrounding patches increases and/or their distances from the focal patch decrease. The Modified Proximity Index (MPI; Bani et al 2006) for a patch is defined as the product of its Proximity Index and area. In our case study we used a 2500-m proximity radius, which avoided considering neighbouring forest patches separated by the highest mountain ridges, a potentially important barrier during animal dispersal.…”

“…Moreover, occurrence of many species can be influenced by environmental characteristics or processes that are difficult to measure, or at least have not been measured adequately. In the case of forest species, the environmental characteristics influencing their presence and abundance can be the amount of forest cover, size and isolation of forest patches, edge effect, and age and structure of stands (Villard et al 1999;Hobson & Bayne 2000;Giese & Cuthbert 2003;Kurosawa & Askins 2003;Batári & Báldi 2004;Herrmann et al 2005;Venier & Pearce 2005;Bani et al 2006). While forest cover and landscape and patch metrics can be obtained or derived from land-use digital maps, data on forest structure are rarely available over large areas.…”

“…They all are typical of broadleaf forests and sensitive to fragmentation, but show different selectivity to the forest structure (Bani et al 2006). In our study area the marsh tit, wood nuthatch and short-toed treecreeper need mature forests, the blue tit can also use intermediate seral stages, while the long-tailed tit may breed in forests of any age (Bani et al 2006). …”

“…The need to produce reliable habitat suitability models for bird species prompted us to test the use of coarse grain data (available only at province level) on the percentage of forest management systems, which are the major determinant of habitat structure in our forests, that are all subjected to silvicultural practices. We also used an index of forest patch isolation (Bani et al 2006) derived from the proximity index calculated by the FRAGSTATS program (McGarigal & Marks 1995).…”