Measuring forest fragmentation using multitemporal forest cover maps: Forest loss and spatial pattern analysis in the Gran Chaco, central Argentina

Abstract

Forest fragmentation is a landscape-level process that consists of two interdependent components: forestloss and spatial pattern changes to which species respond differently. Efficient programs to conservenative biodiversity require a sound understanding of the relation between forest cover and the spatialpattern of forest fragments, but these issues remain almost unknown for subtropical ecosystems.We examine the forest fragmentation of the Gran Chaco in central Argentina over the last 30 years. Inparticular, we quantify forest loss and spatial pattern changes using random sampling techniques onmulti-temporal forest cover maps (1979, 1999 and 2010). We analyzed forest fragmentation accordingto the following steps: (i) selection of fragmentation pattern indices (PIs), (ii) sampling on forest covermaps and PIs calculation, (iii) statistical comparison by bootstrapping, and (iv) trajectory analysis. Duringthe last three decades, forest cover declined dramatically (∼90%) and the selected pattern metrics (MPS,PD, ED) varied significantly (p < 0.05). The results depict a devastating situation of Gran Chaco forestswith a progressive reduction to few small fragments during the last decades. Furthermore when forestloss exceeded the∼50% of the total land area, the temporal trajectories of the selected PIs underwentan abrupt change. Distinguishing habitat spatial pattern changes from forest loss over time supports theidentification of specific conservation actions and provide the basis to establish the lower threshold offorest cover and the more effective arrangement of fragments necessary to mitigate the fragmentationeffects.