Landscape structure quantification using remote sensing derived tree-cover datasets can explain secondary forest recovery across the Neotropics’

Organisator Laboratory of Geo-information Science and Remote Sensing

do 19 november 2015 10:15 tot 10:45

Locatie Atlas, building number 104
Droevendaalsesteeg 4
6708 PB Wageningen
Zaal/kamer 3

By Jorn Dallinga (The Netherlands)

Tropical forests play an important role in providing ecosystem services, contain a high biodiversity and have an important role in the global carbon cycle. Major disturbances on tropical forests by human-induced activities, such as deforestation, result in degraded, depleted and secondary tropical forests. That more than half of the worlds tropical forests have been identified as degraded or secondary forests, has led to a rise in demand for indicators on drivers that lead to the recovery of tropical forest and its resilience. The surrounding landscape structure has been identified as a key driver in the recovery of tropical forests through seed dispersal. Our goal was to determine and quantify the impact of landscape structure by assessing two recent developed remote sensing derived tree-cover datasets on secondary forest recovery. We evaluated secondary forest recovery through Above Ground Biomass (AGB) and species richness data from a large secondary forest network of chronosequence sites. Our study area covers the Neo-tropics, ranging from southern Mexico to Brazil, including 42 chronosequence sites covering various tropical forest types. We related AGB and species richness recovery through standardized multiple linear regression to environmental drivers (precipitation, climatic water deficit, soil fertility and land-use intensity) in combination with tree-cover derived landscape metrics. Landscape metrics quantify the surrounding landscape matrix by assessing the connectivity, landscape continuity and the isolation of forests patches. For AGB, precipitation and Landscape Shape Index (LSI) (edge length and edge density) showed the highest significant (P < .05) explained variance (R2= 0.51) and for species richness the climatic water deficit, LSI and soil fertility showed the highest explained variance (R2= 0.65) with (P < .1). Both AGB and species richness showed the highest explained variances at a 500m buffer radius. Additionally, landscape metrics had an increasingly lesser impact on explaining AGB and species richness when the spatial scales were increased to 1000m and 5000m of surrounding buffer sizes. Based on these results, we conclude that landscape structure within 500m of secondary forest sites are a significant driver on the initial recovery stage of secondary forests. Forest management and conservation strategies can use this new information in assessing tropical forest resilience and by possibly integrating landscape structure in monitoring, reporting and verification activities, which can benefit initiatives such as REDD+.

Keywords: Tropical forest resilience; secondary forest; recovery drivers; tree-cover datasets; landscape metrics; REDD+; Neo-tropics