Resilience changes of carbon stocks to quantify the long‐term effects of ecological engineering projects in subtropical forests of China based on satellite‐derived net ecosystem production time series and inventory data

Abstract Understanding how resilient forests are to ecological engineering projects (EEPs) is essential to forest management and ecosystem health. Despite growing evidence that EEPs achieve increasing carbon stocks, whether such benefits can be sustainable and what are the consequences of EEPs on forest health remain unclear. This study aimed to investigate the long‐term effects of EEPs using forest resilience from aspects of resistance and recovery, by applying a change detection algorithm (breaks for additive seasonal and trend; BFAST) spatially on net ecosystem production (NEP) (proxy for carbon stocks) time series (1981–2019) in red soil hilly region (RSHR) of subtropical China. The spatial parameters (e.g., the number, magnitude, and time of changes) used to construct resilience metrics were generated based on BFAST‐derived breakpoints. These metrics were then utilized to analyze the dynamics of forest resilience in relation to EEPs factors in terms of plantation area, forest type, and stand age. Our results observed 92.77% of breakpoints in NEP after 2000, which corresponds well with the periods that multiple EEPs were conducted. NEP resilience showed great variability during 2001–2019, with a positive increasing trend in resistance ( R 2 = 0.72) and a continuous decline ( R 2 = 0.37) in recovery, indicating an unhealthy ecosystem in RSHR. Our findings revealed that forest resistance was strongly associated with plantation area ( R = 0.71), and the presence of monoculture and young coniferous forest may be the potential factors for the decline in recovery. This suggested that forest resilience in RSHR is mainly modulated by large‐scale EEPs.