Spatial-temporal dynamics and drivers of ecosystem service interactions along the Yellow River area in Shaanxi Province

Abstract

Optimizing ecosystem services (ESs) necessitates the identification of suitable management units and zoning strategies, which fundamentally depend on a comprehensive understanding of the complex interactions of ESs across scales. However, the absence of precise cross-scale research methodologies has impeded progress in this field. This study introduces a cross-scale framework employing the InVEST model, correlation analysis, Geographically Weighted Regression, Morlet wavelet, Self-Organizing Maps, and Geo-detector methods to evaluate and manage ESs along the Yellow River area in Shaanxi Province. We quantified four key ESs in this region, including annual water yield (WY), soil conservation (SC), carbon sequestration (CS), and food production (FP), and analyzed their trade-offs, synergies, and spatial distributions from 2000 to 2020. Our study identified two ES pairs (WY-FP and SC-CS) with strong synergies from 2000 to 2020, yet these diminished over time. Moreover, three ES pairs (CS-WY, CS-FP, and SC-FP) exhibited notable trade-offs across grid and sub-basin scales. The escalating trade-off intensity between CS-WY and CS-FP over time underscores growing conflicts, underscoring the urgency for targeted management strategies. Additionally, the spatial disparities of ESs were evident, with SC and CS more prevalent in the central Loess Hilly-Gully region, contrasting with the southern Guanzhong Plain where WY and FP were more pronounced. Moreover, a marked northward shift in high-value areas for these services was observed from 2000 to 2020. The Morlet wavelet function identified critical scales for ES interactions, notably between 25 and 225 km² (townships) and 400–900 km² (sub-basins). Our analysis of the spatial-temporal dynamics of ES bundles at these scales revealed that most ES bundles exhibit analogous spatial distributions, although certain combinations, such as the WY package at the township scale, manifest distinct scale-specific patterns. Climatic and topographic variables exerted a notable influence, yet the non-linear enhancement between anthropogenic and geographic elements exerted a more pronounced effect on the majority of ES interactions. This research presents a scientifically robust cross-scale framework for ES management, enhancing our understanding of scale-dependent ES dynamics and guiding the development of targeted policies and management strategies for sustainable regional development.