Spatiotemporal pattern of vegetation water use efficiency between 2003 and 2017 and its coupling relationship with artificial carbon sequestration in the karst region of Southwestern China.

[Display omitted] • WUE as indicator used to enhance CS via safe-water use in karst area. • Ecologically critical zones were identified for karst CS with NbS. • DSI impact on WUE is nonlinear. WUE rose faster at DSI > 0.53 in negative area. • LAI-WUE correlation has pronounced thresholds for forests, shrubs. • Restricting access to mountains is effective way to enhance artificial CS. After extensive ecological restoration in the karst region of Southwestern China, a potential zone for achieving "carbon neutrality" has emerged, despite facing water scarcity. We analyzed the dynamics of water use efficiency (WUE) and its correlations with soil moisture (SM) and leaf area index (LAI) from 2003 to 2017 using PML-V2 data and multiple datasets for SM and LAI. Advantageous areas for artificial carbon sequestration (CS) were also identified. Key findings are as follows: (1) Temporally, WUE exhibited fluctuating growth at an annual rate of 0.01 gCmm−1H 2 O (P < 0.05). The advantage areas for CS accounted for 15.96%, with over 31.74% of regions needing management intervention. (2) Monthly WUE was highest in the peak forest plain (PFP) landform (2.88 gCmm−1H 2 O), while PFP experiencing an annual decrease of −0.0021 gCmm−1H 2 O. (3) Monthly WUE was highest in forests (2.88 gCmm−1H 2 O), followed by shrubs (2.49 gCmm−1H 2 O), farmland (2.32 gCmm−1H 2 O) and grassland (1.93 gCmm−1H 2 O), while farmland showing the highest rate of increase (0.02). (4) WUE and SM exhibited both positive (14.26%∼26.02%) and negative (14.19%∼30.98%) correlations. In positive areas, WUE decreased as the drought stress index (DSI) increased for all vegetation types. Clear DSI threshold observed for farmland (0.29 to 0.42) and grassland (0.19 to 0.30). However, the peak values for forests and shrubs were less pronounced. There was a transitional point in the negative impact of DSI on WUE, which increased at a faster rate for all vegetation types when DSI exceeded 0.53. (5) WUE and LAI exhibited both positive (27.09%∼30.25%) and negative (23.37%∼34.57%) correlations. In positive areas, WUE increased as LAI increased for all vegetation types. However, forests exhibited threshold LAI values of 1.85 based on MODIS data, 2.71 based on GLASS data, and 2.59 based on GEOV2 data, shrubs had values of 1.69, 2.66, and 2.23, respectively. While grassland displayed LAI values of 0.79, 0.70, and 0.72, respectively. In negative areas, WUE decreased as LAI increased for all vegetation types. The minimum WUE values were observed when the LAI of farmland was 3.14 based on MODIS data, 4.05 based on GLASS data, and for grassland, it was 1.04, 1.97, and 1.76 based on the GEOV2 data, respectively. This study helps us to identify areas with potential for enhancing CS. It also assists in making informed decisions regarding the implementation of ecological restoration initiatives by considering SM as the limiting factor for adjustment measures and utilizing LAI as the reference standard. [ABSTRACT FROM AUTHOR]