Generalities of vertebrate responses to landscape composition and configuration gradients in a highly heterogeneous Mediterranean region

Abstract

A B S T R A C TAim To examine the distributional patterns of vertebrates (including birds, bats, carnivores and lagomorphs) along landscape composition and configuration gradients to better understand the effects of landscape modification on occurrence patterns at both species and community level.Location The region of Alentejo, a forest-dominated area of southern Portugal.Methods The study area was framed using 1647 hexagonal plots, each of 259 ha in size. Composition and configuration gradients were obtained for each plot by integrating the proportions of the main land cover types and their configuration patterns using multivariate analyses. Species-specific vertebrate responses were investigated using data from 75 plots in which carnivores, bats and lagomorphs were sampled, and from 135 plots in the case of birds. Community-level responses were investigated through changes in species richness and beta-diversity in 57 plots where all vertebrate groups were simultaneously sampled. At the species-level, an information-theoretic approach was used to determine the effects of landscape gradients on species' responses. At the community level, Mantel tests were used to determine between-plot differences in species composition using the Sørensen dissimilarity index.Results We found that the occurrence patterns of most vertebrate species were best predicted by composition-related gradients, although configuration gradients were also frequently included in species-specific occurrence models. We also found a weak correlation between species richness and most landscape gradients suggesting a turnover in the identity of species, something that was corroborated by the stronger correlation between environmental gradients and beta-diversity measures. The amount of forest cover and landscape complexity (estimated as the heterogeneity in the size and number of land cover types) were the main composition and configuration gradients determining vertebrate responses at both species and community level.Main conclusions Our work contributes to a more refined understanding of the mechanisms underlying species distributional patterns in real-world human-modified landscapes. By uncovering generalities of species with multiple ecological requirements and by describing the entire landscape mosaic through landscape gradients, we also suggest that our work greatly helps to fill the gap between existing conceptual landscape models aimed to understand species distributional patterns in human-modified landscapes.

Exploring the links between social metabolism and biodiversity distribution across landscape gradients: A regional-scale contribution to the land-sharing versus land-sparing debate

“…On the other hand, the role of landscape structure and composition in biodiversity distribution in human-modified regions has started to beexplored. Herrera et al (2016) recently emphasized the importance of forest cover and landscape complexity in vertebrate responses at both species (i.e. abundance) and community (i.e.…”

A distribution-oriented approach to support landscape connectivity for ecologically distinct bird species

“…Most research aiming to infer landscape connectivity from resistance values computed by landscape characteristics have assumed that the availability of preferred land-cover type plays an essential (and often exclusive) role in shaping the spatial distribution patterns of species [ 44 , 45 ]. Nevertheless, because most landscapes, particularly human-modified landscapes, are often mosaics of various land-cover types that may offer a number of different suitable resources and environmental conditions [ 22 ], such an approach to landscape suitability analysis might be too simplistic and theoretically inadequate [ 46 ]. Landscapes are increasingly recognized continuous, multi-dimensional gradients of environmental suitability rather than discrete units of suitable versus non-suitable habitat [ 47 ].…”

“…This work was carried out in the region of Alentejo in Southern Portugal (centroid: 16271.45, -113395.21; EPSG: 3763-ETRS89 / Portugal TM06), in a study area of ca . 426,000 hectares ( Fig 1A ; see also Fig 1 in [ 22 ]). The owners of land properties gave permission to conduct sampling surveys.…”

“…The regional landscape was characterized using both composition and configuration metrics [ 22 , 24 ]. Landscape composition was quantified based on the proportions of the six land-cover types detailed above within individual cells.…”

“…the landscape matrix) [ 19 – 21 ]. This is because the landscape matrix may offer suitable resources and environmental conditions for multiple species [ 22 , 23 ]. However, most studies to date have focused principally on the characteristics of the presumed o estimated preferred land-cover type(s) rather than on the structural properties of the entire landscape mosaic.…”