不同管理模式对农田生态系统服务的影响模拟研究——以太湖流域为例

高强度的农业生产给生态环境带来诸多负面影响,如何权衡农田生态系统服务是农业可持续发展面临的重要问题之一。本文以太湖流域稻麦轮作农田生态系统为研究对象,利用基于0—10cm土壤温度、10—20cm土壤温度、水稻生物量和小麦生物量等为目标函数的PEST软件校准APSIM模型,遴选三类重要生态系统服务——粮食供给(粮食产量)、水质调节(氮流失)和气候调节(N_2O排放、固碳-土壤有机碳),针对不同秸秆还田(无秸秆还田、稻秆还田麦秆不还田、麦秆还田稻秆不还田、稻麦秆全还田)和施肥模式(0、-5%,-10%,-20%)等设定16种情景模拟1986—2015年太湖流域生态系统服务变化趋势,利用自组织映射法(SOM)分析其权衡协同关系变化,以期为农田生态系统管理提供科学依据。结果表明:(1)相对秸秆还田,粮食产量、N_2O排放、氮流失和土壤有机碳受施肥影响更为显著,均随施肥量减少而下降;(2)不同管理模式对固碳影响较弱,对N_2O排放和水质调节影响明显,稻麦秆全还田且减量施肥5%、10%或20%情景对N_2O排放抑制最为有效,减排量均超过30%,氮流失量减量大于7%;(3)粮食供给与固碳、N_2O排放均存在显著的协同关系,粮食供给与水质调节、固碳与水质调节、水质调节与N_2O排放呈显著权衡关系,固碳与N_2O排放呈弱权衡关系;(4)基于N_2O排放量和氮流失量均低于基准模拟且粮食单位产量减少低于5%的情景下,无秸秆还田且减量施肥5%(S0_F1)和稻麦秆全还田且减量施肥5%(S3_F1)两种模式相对有利于整体提升太湖流域农田生态生态系统服务。此外,APSIM模型有待更多的野外监测数据进一步校准改进以提高模型率定精度,同时需扩展不同模式影响下太湖流域农田生态系统服务综合效应评估,以获取更加科学、合理的优化模式与途径。

Simulating the effects of different management modes on the ecosystem services of agroecosystems: a case study of the Taihu Lake Basin, China

Ecosystem services (ES) of rice-wheat rotation agroecosystems play an important role in global ecosystem services. Some ES are facing large-scale degradation owing to changes in land use and agriculture management modes in the Taihu Lake Basin. This has resulted in an increasing awareness of sustainable agricultural development through optimizing and advancing ES of the agroecosystems. This study simulated the trends and trade-offs of agroecosystem ES in the Taihu Lake Basin from 1986 to 2015 using the APSIM model. The APSIM model was calibrated by PEST software, with observations on grain supply (grain field), climate regulations (N₂O emissions, carbon sequestration-soil organic carbon), and water quality regulations (N-losses). Recently, high utilization of chemical fertilizers and low-use efficiency of straw has led to many severe issues in the Taihu Lake Basin, such as agricultural non-point source pollution. Based on the combinations of differential straw-returns (non-straw returning, rice straw-returning and non-wheat straw-returning, wheat straw-returning and non-rice straw-returning, all straw-returning) and different fertilizer use (0, -5%, -10%, -20%), this study designed 16 scenarios to simulate and assess three key ES in the Taihu Lake Basin. The trade-offs were assessed by using a self-organizing map (SOM) to derive effective management measures to improve overall agroecosystem ES. The results show that:(1) grain yield, N₂O emissions, N losses, and soil organic carbon are more significantly affected by fertilizer use than are straw-returns, and the former are characterized by declines with reduced fertilizer utilization; (2) Different management modes have a weak impact on carbon sequestration, but significant effects on N₂O emissions and water quality regulation. All straw-returning and 5%, 10%, or 20% reductions of fertilizer utilization are more effective in suppressing N₂O emissions, and the reduction of N losses is greater than 7%. (3) There exists a strong synergistic relationship between grain supply and both carbon sequestration and N₂O emissions, respectively. Grain supply and water quality regulation, carbon sequestration and water quality regulation, and water quality regulation and N₂O emissions present significant trade-offs, respectively. Carbon sequestration and N₂O emissions show a weak trade-off. (4) Assuming water quality regulation (N losses) and climate regulations (N₂O emissions) are lower than those of the basic scenario and the reduction in grain yield is less than 5%, two scenarios-non-straw returning and 5% fertilizer reduction (S0_F1), and all straw-returning and 5% fertilizer reduction (S3_F1)-can help to advance overall agroecosystem services in the Taihu Lake Basin. It is necessary to improve calibration of the APSIM model because of the relatively low soil organic carbon. Additionally, the comprehensive effects of different modes on agroecosystem ES in the Taihu Lake Basin need to be further explored.