Climate change is seriously threatening forest ecosystems, affecting their structure and functioning. Rising temperatures and reduced precipitation lead to prolonged droughts, which weaken the natural defences of trees and affect their photosynthetic and stomatal activities, making them vulnerable to other stresses, including pest attacks. Identifying the most vulnerable forest areas and understanding the mechanisms of tree decline are therefore crucial to developing effective management strategies.
This research combined dendrochronological and isotopic analyses with remote sensing to detect the early signs of dieback in a population of Pinus pinea L. in southern Italy affected by Toumeyella parvicornis parasite outbreak. Moreover, a comparative analysis of the methods was conducted to identify the most effective data processing techniques for detecting forest decline.
The study showed that the pest outbreak, which started in 2014, caused progressive tree defoliation from 2015, reaching a critical point in 2020, leading to a severe tree carbon deficit and triggering a significant decline in growth. Despite attempts to rebalance carbon metabolism through stomatal opening, as indicated by the intrinsic water use efficiency (WUEi) data, the pine forest did not recover. Indeed, these trees, particularly stressed by the climatic conditions of the area, died in 2023.
Although all satellite indices tested were able to detect defoliation dynamics, EVI and EVI2 proved to be particularly sensitive to changes in canopy cover, more than NDVI. The integration of additional indices, such as NDMI, improved the monitoring of canopy moisture and provided valuable insights into decline dynamics. Finally, dendrochronological analyses showed that detrended growth chronologies (BAI and TRW-I) were more sensitive in detecting dieback signals than raw tree-ring chronologies.
In conclusion, the data collected in this study not only provide information on the dynamics of forest dieback, but also highlight the importance of integrating different detection methods for the effective management of forest ecosystems.
