Quantifying the direct and indirect effects of terrain, climate and human activity on the spatial pattern of kNDVI-based vegetation growth: A case study from the Minjiang River Basin, Southeast China

In the context of global change, it is vital to comprehensively understand the spatial pattern and driving mechanism of vegetation growth to maintain the stability of watershed ecosystems. Previous research has focused mainly on identifying the main drivers of vegetation growth, while the direct and indirect effects of climate, terrain, and human activity on vegetation growth have rarely been explored. This study used the Minjiang River Basin (MRB), an important ecological barrier and the largest watershed in southeastern China, as an example. The kernel normalized difference vegetation index (kNDVI) was calculated on the Google Earth Engine (GEE) platform to examine the spatial pattern and evolution characteristics of vegetation growth. The optimal parameter-based geographical detector (OPGD) and partial least squares structural equation modeling (PLS-SEM) were used to analyze how terrain, climate, and human activity influenced the spatial pattern of the kNDVI. (1) From 2001 to 2020, vegetation growth in the MRB was predominantly rated as excellent or good, and 88.93% of the area showed an increasing trend of vegetation growth. (2) The OPGD revealed that the primary drivers influencing the spatial distribution of the kNDVI in the MRB included population density, nighttime light, elevation and temperature, which explained >40% of the variation in the kNDVI. The interaction of all paired drivers enhanced the explanatory power of the kNDVI, among which the strongest interaction was between population density and elevation, and the second interaction was between population density and temperature. (3) PLS-SEM revealed that human activity had a direct negative effect on the kNDVI, while terrain and climate had direct and indirect positive effects on the kNDVI. Overall, the total effects of terrain, climate and human activity on the kNDVI were 0.594, 0.233 and − 0.495, respectively, indicating that the positive effect of terrain outweighed the negative effect of human activity on vegetation growth in the MRB. These findings not only provide scientific evidence for ecological conservation and management in the MRB but also offer a useful reference for other regions exploring the complex causes of spatial patterns of vegetation growth.