干旱区生态环境知识资源中心

The precipitation legacy effect, defined as the impact of historical precipitation (PPT) on extant ecosystem dynamics, has been recognized as an important driver in shaping the temporal variability of dryland aboveground net primary production (ANPP) and soil respiration. How the PPT legacy influences whole ecosystem-level carbon (C) fluxes has rarely been quantitatively assessed, particularly at longer temporal scales. We parameterized a process-based ecosystem model to a semiarid savanna ecosystem in the southwestern USA, calibrated and evaluated the model performance based on 7 years of eddy-covariance measurements, and conducted two sets of simulation experiments to assess inter-decadal and interannual PPT legacy effects over a 30-year simulation period. The results showed that decreasing the previous period/year PPT (dry legacy) always increased subsequent net ecosystem production (NEP) whereas increasing the previous period/year PPT (wet legacy) decreased NEP. The simulated dry-legacy impacts mostly increased subsequent gross ecosystem production (GEP) and reduced ecosystem respiration (R-e), but the wet legacy mostly reduced GEP and increased R-e. Although the direction and magnitude of GEP and R-e responses to the simulated dry and wet legacies were influenced by both the previous and current PPT conditions, the NEP responses were predominantly determined by the previous PPT characteristics including rainfall amount, seasonality and event size distribution. Larger PPT difference between periods/years resulted in larger legacy impacts, with dry legacies fostering more C sequestration and wet legacies more C release. The carryover of soil N between periods/years was mainly responsible for the GEP responses, while the carryovers of plant biomass, litter and soil organic matter were mainly responsible for the R-e responses. These simulation results suggest that previous PPT conditions can exert substantial legacy impacts on current ecosystem C balance, which should be taken into account while assessing the response of dryland ecosystem C dynamics to future PPT regime changes.