An improved quality assessment framework to better inform large-scale forest restoration management

Abstract

Dynamic monitoring of forest ecosystem quality is necessary for restoration program evaluation but remains challenging for very large-scale programs. Current evaluation methods employ regional forest quality indicators that compare the quality status of targeted forests with benchmarks from remnant old-growth forest communities, however data availability usually limits the application of available methods to small scales. We constructed an improved framework, integrating forest site classification selection and local remnant old-growth forest community delimitation, to quantify and map forest quality using environmental data and remote sensing (RS) based approaches. A classification strength model was introduced to improve the accuracy of forest site classification. The remote-sensing-based method integrates species composition and forest biological productivity characteristics recognition to develop a practical tool for large-scale remnant old-growth forest community delimitation. The new assessment framework was tested across the entire spatially heterogeneous Yangtze River Basin, the largest watershed in China and showed high accuracy in forest quality assessment based on observed field data validation. The forest site classification was selected by considering spatial heterogeneities in climate, topography and soil type, with 37 forest sites classified. The native forest community groups with less human disturbance in each forest site used for forest quality baseline estimation were also selected as forests with top 10% of biomass in protected areas. The case study demonstrated that forest areas of low and poor quality accounted for 34.46% of the total forest area in 2015. Between 2000 and 2015, 55.72% of forest areas experienced increases in quality level, and 7.07% experienced decreases. The improved forest quality assessment framework enhances the scope and accuracy of forest restoration information and can be applied as an evaluation tool for forest restoration management.