Assessing the effect of scale on the ability of landscape structure metrics to discriminate landscape types in Mediterranean forest districts

  • C. García-Feced Universidad Politécnica de Madrid
  • S. Saura Universidad Politécnica de Madrid
  • R. Elena-Rosselló Universidad Politécnica de Madrid
Keywords: landscape patterns, grain size, minimum mapped unit, variability analysis, metric correlation, forest planning


Scale is a key concept in landscape ecology. Although several studies have analyzed the effect of scale on landscape structure metrics, there is still a need to focus on the ability of these metrics to discriminate between landscape types at different scales, particularly in Mediterranean forest landscapes. In this paper we assess the scaling behavior and correlation patterns of eight commonly-used landscape metrics in two Spanish forest districts (Pinares in Burgos and Soria, and Alto Tajo in Guadalajara) in order to detect at which grain sizes the landscape type differences are emphasized. This occurred in both districts at fine spatial resolutions for the metrics related to shape complexity and the amount of boundaries in the landscape, while a coarser spatial resolution was required for the landscape diversity and mixture metrics, suggesting that the differences in the spatial and compositional diversity of these landscape types are not so large locally (alpha diversity) but amplified at broader scales (gamma diversity) as a result of the high turnover rates characteristic of Mediterranean landscapes. The maximum variability for the fragmentation-related metrics did not appear at the same scale in both districts, because forest fragmentation in the Pinares district is mainly driven by harvesting treatments that operate at considerably different scales from those related to the topographic and climatic patterns that predominantly control the forest distribution in the more environmentally heterogeneous and less intensively managed district of Alto Tajo. Our methodology and results allow extracting the maximum information content from these metrics and adequately identifying and separately assessing those complex land cover mosaics that result from a similar set of biological and social forces and constraints. This should be valuable for an improved forest landscape planning and monitoring with a quantitative ecological basis in the Mediterranean and other temperate areas.


Download data is not yet available.
How to Cite
García-FecedC., SauraS., & Elena-RossellóR. (2010). Assessing the effect of scale on the ability of landscape structure metrics to discriminate landscape types in Mediterranean forest districts. Forest Systems, 19(2), 129-140.
Research Articles