Objective Changes in landscape pattern often bring about alterations in the structure, process, and function of ecosystem. In the context of rapid urbanization, drastic changes in land use/land cover (LULC) significantly disturb the structure and function of rural landscapes, leading to a series of ecological and environmental issues in rural areas such as farmland transformation, landscape fragmentation, and habitat degradation. These issues, in turn, pose threats to the stability of ecosystem. Therefore, understanding the impact of rural landscape pattern on ecosystem service provision is fundamental for managing and planning rural ecosystem under rapid urbanization. In light of this, this research takes Hangzhou as an example to reveal the relationship between rural landscape pattern and ecosystem service under rapid urbanization, in hope of providing a reference value for local decision-makers in terms of land use, and a basis for planning and layout in the central and western regions of China that are about to undergo urbanization.

Methods This research focuses on a typical area undergoing rapid urbanization — the rural area of Hangzhou, Zhejiang Province. This area has experienced dramatic changes in land use/land cover over the past 20 years, resulting in significant ecological issues such as habitat degradation and landscape fragmentation. The research employs models such as InVEST and Fragstats to assess ecosystem services (including food production, water yield, carbon sequestration, habitat support, soil conservation, and cultural services) in relation to rural landscape patterns. Spatial relationships between ecosystem service provision and rural landscape pattern within the research area are identified using OLS (ordinary least squares), GWR (geographically weighted regression), and MGWR (multiscale geographically weighted regression) models. Additionally, the performance of the MGWR model is compared with other global or local regression models.

Results 1) The spatial pattern of ecosystem service function in the rural area of Hangzhou is closely related to land use type. Overall, woodland and grassland have a positive impact on ecosystem service, while arable land, residential area, and construction land have a negative impact. Additionally, water yield service is influenced not only by land use/cover type but also by factors such as climate and watershed runoff. 2) The rural landscape of Hangzhou is becoming increasingly fragmented and homogenized, with the most notable changes occurring in the plain area, while the mountainous and hilly areas are less affected by human interference. This is primarily due to the fact that human activities, such as urban expansion and scale-up of agricultural and forestry operations, have led to an increase in urban construction area and population density. These changes have altered the landscape pattern, contributing to a certain degree of fragmentation and complexity. 3) The impacts of rural landscape pattern on different ecosystem services vary, which is reflected in both landscape pattern indicators and their impact degrees. Landscape pattern indicators (slope degree, total landscape area, Shannon’s evenness index, edge density, and contagion) have a significant impact on ecosystem service, presenting a nonlinear relationship. However, the impacts of rural landscape pattern on different ecosystem services vary. Specifically, slope degree (SLOPE) has the most significant impact on ecosystem service, showing a strong positive correlation with the latter. This is followed by total landscape area (TA), Shannon’s evenness index (SHEI), edge density (ED), and contagion (CONTAG). 4) Here are optimization recommendations guided by the relationship between landscape pattern and ecosystem service. At the rural scale, ecosystem service can be optimized based on the influence strength of various landscape pattern indicators. For provisioning services, it is recommended to rationally optimize the layout of agricultural land to restrict the expansion of construction land and to promote high-level farmland protection. For regulating services, enhancing the connectivity of landscape patches can be achieved by restoring woodland, increasing water area, and strengthening vegetation cover. In terms of cultural services, planning should be optimized based on existing rural landscapes, and “agriculture+” and “ecology+” models should be explored to achieve a synergistic development of ecological and economic benefits.

Conclusion This research is not a simple overlay of rural landscape pattern and ecosystem service. Rather, it distinguishes between the ecological conservation directions of rural landscapes and the priorities for regional development. It can serve as a theoretical basis for policy-making concerning ecosystem service and for the planning and management of rural landscape, aiming to achieve healthy and sustainable management of rural ecosystem services. Admittedly, the research has limitations: It models the relationship between landscape pattern and ecosystem service with 2.5 km × 2.5 km grid as the basic geographic analysis unit. In practical applications, spatial planning may require adjustments, potentially utilizing existing planning management units, such as the rural scale, and incorporating more influencing factors of ecosystem service to build a more comprehensive and integrated assessment framework. Additionally, temporal scale considerations should be included to understand the evolution and mechanism of rural ecosystem service provision under the pressure of urbanization. This approach will accurately reveal the complex relationships between ecosystem service and landscape pattern, and further advance the methods for rural landscape planning and design to better promote the sustainable development of rural living environment and regional development.