17:15 〜 19:15
[ACG46-P05] 作物モデルMATCROを用いたアジア地域の過去1世紀にわたるコメ収量復元
キーワード：コメ収量、作物モデル、パラメータ最適化、気候変動
Global food demand would keep increasing by the end of 2050 (Dijk et al., 2021; Godfray et al., 2010). On the other hand, recent studies have reported a stagnation in the rise of crop yields (Ray et al., 2012), and agricultural production would be affected by future climate change (Challinor et al., 2014). Rice is one of the staple crops feeding more than half of the global population, and Asia accounts for more than 90% of global rice production, being important regions for global food security (Nimala, 2018). Thus, investigating rice yield trends in Asia is essential for maintaining crop production and ensuring future food security.
Several studies have used crop yield statistics and crop models to examine the relationship between rice yield and climate change (Lobell et al., 2011; Xiong et al., 2012). Others have employed crop models to investigate crop traits that are suitable for future climate conditions (Paleari et al., 2022). However, these previous studies have focused only on the past few decades, making it difficult to detect clear changes in climate conditions and long-term trends since the early 1900s. In addition, they have typically addressed either environmental or genetical factors in isolation, failing to conduct a comprehensive analysis that considers both climate change and crop improvement.
Process-based crop models are useful for considering both climate change and crop breeding. MATCRO (Masutomi et al., 2016), a process-based crop growth model, simulates rice yield by calculating rice growth based on crop physiology. This model includes parameters that represent rice growth characteristics, such as photosynthesis ability and temperature tolerance. We hypothesized that if these parameters were optimized to allow the model to accurately reconstruct past rice yield, the optimized parameters would provide insights into historical crop growth characteristics. Therefore, this study aims to optimize parameters in MATCRO to reconstruct rice yield in Asia over the past century (1906-2015) and to quantify the relative impacts of past climate change and breeding on rice yield increase with optimized parameters.
First, we optimized six parameters related to carbon allocation, morphology, temperature resistance, and phenology in MATCRO. Optimization was conducted at the point scale for four regions: Korea, Thailand, Indonesia, and Bangladesh. We assumed single cropping for Korea, double cropping for Thailand and Indonesia, and triple cropping for Bangladesh. The optimization window covered 20 years, beginning with 1996-2015 and moving backward in 10-year intervals to 1906-1925. To distinguish the contributions of past climate change and breeding, we ran MATCRO under various conditions, fixing environmental and genetic parameters at their 1906 values and quantifying the contribution of each fixed parameter to rice yield.
The optimization results showed a generally good reconstruction of observational data. Although MATCRO was struggled to accurately reconstruct regional yield variations, optimization significantly improved its ability to reproduce overall yield trends. While the optimized parameters in Korea, where single cropping was practiced, provided meaningful insights into phenology or carbon allocation, the optimized parameters in other regions produced unexpected results. This discrepancy likely arose because the optimization was conducted using only annual data. Moreover, the quantification of impacts revealed that rising atmospheric CO2, nitrogen fertilization, and temperature have significantly influenced rice yield.
Several studies have used crop yield statistics and crop models to examine the relationship between rice yield and climate change (Lobell et al., 2011; Xiong et al., 2012). Others have employed crop models to investigate crop traits that are suitable for future climate conditions (Paleari et al., 2022). However, these previous studies have focused only on the past few decades, making it difficult to detect clear changes in climate conditions and long-term trends since the early 1900s. In addition, they have typically addressed either environmental or genetical factors in isolation, failing to conduct a comprehensive analysis that considers both climate change and crop improvement.
Process-based crop models are useful for considering both climate change and crop breeding. MATCRO (Masutomi et al., 2016), a process-based crop growth model, simulates rice yield by calculating rice growth based on crop physiology. This model includes parameters that represent rice growth characteristics, such as photosynthesis ability and temperature tolerance. We hypothesized that if these parameters were optimized to allow the model to accurately reconstruct past rice yield, the optimized parameters would provide insights into historical crop growth characteristics. Therefore, this study aims to optimize parameters in MATCRO to reconstruct rice yield in Asia over the past century (1906-2015) and to quantify the relative impacts of past climate change and breeding on rice yield increase with optimized parameters.
First, we optimized six parameters related to carbon allocation, morphology, temperature resistance, and phenology in MATCRO. Optimization was conducted at the point scale for four regions: Korea, Thailand, Indonesia, and Bangladesh. We assumed single cropping for Korea, double cropping for Thailand and Indonesia, and triple cropping for Bangladesh. The optimization window covered 20 years, beginning with 1996-2015 and moving backward in 10-year intervals to 1906-1925. To distinguish the contributions of past climate change and breeding, we ran MATCRO under various conditions, fixing environmental and genetic parameters at their 1906 values and quantifying the contribution of each fixed parameter to rice yield.
The optimization results showed a generally good reconstruction of observational data. Although MATCRO was struggled to accurately reconstruct regional yield variations, optimization significantly improved its ability to reproduce overall yield trends. While the optimized parameters in Korea, where single cropping was practiced, provided meaningful insights into phenology or carbon allocation, the optimized parameters in other regions produced unexpected results. This discrepancy likely arose because the optimization was conducted using only annual data. Moreover, the quantification of impacts revealed that rising atmospheric CO2, nitrogen fertilization, and temperature have significantly influenced rice yield.