Spatial and temporal variability of future ecosystem services in an agricultural landscape

Context Sustaining ecosystem services requires enhanced understanding of their spatial–temporal dynamics and responses to drivers. To date, the majority of research has focused on snapshots of ecosystem services, and their spatial–temporal variability has seldom been studied. Objectives We aimed to address: (i) How is variability in ecosystem services partitioned among ‘space’ and ‘time’ components? (ii) Which ecosystem services are spatially/temporally coherent, and which are space–time incoherent? (iii) Are there consistent patterns in ecosystem service variability between urban- and rural-dominated landscapes? Methods Biophysical modeling was used to quantify food, water, and biogeochemical-related services from 2011 to 2070 under future scenarios. Linear mixed-effects models and variance partitioning were used to analyze spatial and temporal variability. Results Food production, water quality and flood regulation services were overall more variable than climate regulation and freshwater supply. ‘Space’ contributed to a majority of variations across most services, highlighting dominant importance of location-specific factors for service supply. Significant space–time interactions existed for water quality and soil carbon storage, indicating interactive effects between location- and time-specific factors. Variation in the relative controls of ‘space’ vs. ‘time’ factors between urban- and rural-dominated subwatersheds suggests that targeting different key drivers is needed for successful management of ecosystem services in urban vs. rural landscapes. Conclusions Our research reveals relative importance of underlying ‘space’ and ‘time’ controls for diverse ecosystem services. Our study presents a framework to investigate spatial–temporal variability of ecosystem services, and provides theoretical and practical implications for anticipating and managing the dynamics of future ecosystem services at the watershed scale.