Assessing 3D vs. 2D habitat metrics in a Mediterranean ecosystem for a wiser wildlife management

Abstract

Although many ecological studies recognize the importance of the three-dimensional complexity of the landscape, only a small number of investigations incorporate it into their methods. Through an empirical study we assess the extent to which habitat metric patterns vary measured by 3D versus 2D methods, and whether it is possible that variations are associated with scale, topographic complexity or specific animal behavior. We analyze areas and potential animals where the use of 3D approaches is most influential for reliable results in ecological studies. This study was carried out in southern Spain, focusing on game estates covering an area of about 2.241 km2. The topographical features of the areas were obtained from a digital terrain model (DTM) and a scalable terrain model (STM) generated from Lidar data. The estates were categorized according to a preliminary 3D vs. 2D surface increment index (SI index). Then an experimental repeated measures design was carried out to compare three nested scales from 1 to 4 km square plots. The influence of plot topography on the SI index was analyzed statistically. Plots were also compared pair-wise through a post hoc comparison analysis. In order to establish how 3D affects the metrics of the trajectories travelled by different animal species, we simulated movements of Oryctolagus cunniculus, Cervus elaphus, Canis lupus and Lynx pardinus, which move and behave differently in the study area. We simulated 15 replicate paths per specie, assuming 24 fixes per path. Variations in animal trajectories were analyzed by fitting general linear models (GLMs), which also provide a basic method for estimating the potential errors produced by a simple 2D analysis. We obtained significant differences between the projected and geometric surface areas, with an increase of up to 11.15% and an increase in animal paths length up to 5% using 3D approaches. Landscape complexity, rather than species type, is related to animal path metrics differences. However, in species such as deer and lynx the 3D spatial component is basic for an accurate estimate of path length, especially in heterogeneous landscapes. Altitude and slope range are determining factors for considering the 3D approaches as advisable. Our results confirm that ecological studies ignoring three-dimensional (3D) components run the risk of misinterpretation due to the fact that the bi-dimensional (2D) approach underestimates distances and landscape area. The use of 3D methods provides more accurate results than with 2D approaches, with far-reaching implications for a wiser wildlife management and conservation in Mediterranean ecosystems.