产量(工程)
产量差距
播种
生长季节
作物
环境科学
农学
作物产量
作物管理
作物模拟模型
数学
生物
冶金
材料科学
作者
Maria Carolina da Silva Andréa,Kenneth J. Boote,Paulo César Sentelhas,Thiago Libório Romanelli
标识
DOI:10.1016/j.agsy.2018.07.004
摘要
Occurrence of staple crops' yield gaps is object of study worldwide. A theoretical approach, model and statistical-based, was carried out to assess the climate-induced variability of rainfed maize yields and yield gaps in different regions in Central-Southern Brazil in both main growing seasons. A crop simulation model was used to estimate potential (Yp) and water-limited (Yw) yields through thirty crop seasons. Based on observed local farmers' averages and simulated yields, yield gaps related to water deficit (WYg) and crop management (MYg) were determined for first (sowing starting in September) and second (sowing starting in January) typical maize growing seasons. Overall higher average values of Yp and Yw (15.3 and 13.1 t ha−1, respectively) were obtained in the first when compared to second growing season (10.3 and 9.2 t ha−1, respectively). Statistical approaches pointed to similar importance between water and temperature on local biophysical limits in the scenarios. Assessed regions showed greater gaps due to crop management, with absolute averages of 5.7 and 3.2 t ha−1 in the first and second growing seasons, than gaps due to water deficit, with 2.1 and 1.2 t ha−1 in the first and second growing seasons, respectively. Opportunities for increasing average yields by closing the gaps were found to be predominantly through crop management improvements, in higher and more variable absolute levels on first than on second growing season. However, this management must be aligned with local climate, since its variability can determine relatively large gaps, even at intensively managed cropping systems. This study was able to highlight the importance of combining management, climatic and regional characteristics to provide a full perspective on main constraints of maize production increases.
科研通智能强力驱动
Strongly Powered by AbleSci AI