Assessing the relationship between forest structural diversity and resilience in a warming climate

Ecosystem resilience represents the capacity of an ecosystem to withstand and recover from external perturbations, an increasingly important property for ecosystem function in an era of escalating climate extremes and anthropogenic pressures. Whilst recent studies have related forest resilience to natural factors such as climate and biomass, the link between forest diversity and resilience is not yet understood.

 

This study quantifies the sensitivity of ecosystem resilience on forest diversity in Europe over the period 2003-2021. Two commonly used resilience indicators are considered based on MODIS kNDVI (kernel Normalized Difference Vegetation Index) data acquired at high spatial and temporal resolution: the 1-lag temporal autocorrelation, relating to the ecosystem memory, and the standard deviation, relating to the ecosystem stability. Forest diversity is expressed in terms of horizontal and vertical structural heterogeneity metrics derived from GEDI (LiDAR) (Light Detection and Ranging) acquisitions. A Random Forest (RF) model is leveraged to isolate the interplay between forest resilience and diversity metrics by disentangling possible confounding environmental variables such as climate. The RF model is then applied to retrieve local sensitivities in terms of Individual Conditional Expectations.

 

The work first finds that European forests with a higher level of vertical and horizontal structural diversity are systematically associated with higher resilience levels. The relationship is coherent across bio-geographical regions in Europe. Importantly, the emerging relation between forest resilience and forest diversity is consistent under increasing temperature patterns. This suggests that forest management targeted to higher levels of forest heterogeneity has the potential to offset the decline in forest resilience associated with the projected climate warming scenarios and the consequent increasing disturbance regimes.