Measuring forest fragmentation using multitemporal remotely sensed data: three decades of change in the Dry Chaco

Abstract

We introduce an approach based on remotely sensed data to summarize forest fragmentation over time, which specifically accounts for the interdependencies between landscape composition and configuration changes. The proposed method consists of five steps: i) multitemporal landscape sampling, ii) calculation of selected landscape pattern indices, iii) statistical comparison, iv) construction of sampled-based relationship spaces, and v) trajectory analysis. To show how the proposed method works in practice we examined the multitemporal fragmentation of the Arid Chaco forest in central Argentina during the period 1979-2010 using forest maps derived from Landsat images. As shown by our results, the approach provides a consistent framework for the interpretation of landscape structural changes over time.