[A-CG46] Biogeochemical Cycles in Land Ecosystem

[ACG46-P04] Tracing a century of harvests: how has climate shaped the rise, stagnation, and shifts in Thailand’s rice, maize, and soybean yields?

*Praeploy Kongsurakan¹, Tomomichi Kato¹, Li Xinli¹, Azusa Suzuki¹, Tatsuki Nakagawa² (1.Research Faculty of Agriculture, Hokkaido University, 2.Graduate school of Global Food Resources, Hokkaido University)

Keywords:yield trend, yield stagnation, dynamic linear model, teleconnection, climate variation

Agriculture is the cornerstone of a stable and sufficient food supply, particularly given the growing global population. Studying historical yield trends provides valuable insights into long-term changes and the key factors driving yield variability, which are essential for strengthening food security and promoting sustainable agricultural development. This research investigates a century of historical yields to assess the extent to which observed yield stagnation and growth in rice, maize, and soybeans in Thailand are linked to climate variability. We applied both polynomial detrending and a dynamic linear model (DLM) approach to reconstruct the annual production yields of the three crops—rice, maize, and soybeans—from 1918 to 2022. To detect significant periods of stagnation, we evaluated the estimated growth rates and confidence intervals derived from individual DLMs at the provincial scale. We further examined the links between climatic indices, including weather patterns and teleconnection oscillation systems, and crop yields. These analyses provide a basis for developing seasonal crop yield forecasting models, enabling more effective and dynamic adaptation to climate variability and change. Our results reveal that rice yield stagnation was initially widespread in central and southern Thailand (1932–1945) but became more fragmented over time, with pockets of stagnation persisting, particularly in the northern and northeastern regions. For maize, stagnation was most prominent in northern and central Thailand during the early 20th century (1918–1931) but gradually declined, with isolated areas experiencing stagnation in later periods. Soybean stagnation was less pronounced and more localized compared to maize and rice. Teleconnection variables explain, on average, 34% of inter-annual rice yield variability, ranging from 8% to 71% across regions. For grain maize, the average explained variability is 11%, ranging from 3% to 28%. For soybeans, teleconnection variables account for an average of 7% of yield variability, ranging from 4% to 12%. Some teleconnection variables, such as El Niño-Southern Oscillation (ENSO) and Pacific Decadal Oscillation (PDO), show strong associations with rice and maize yield variations, particularly during certain periods. Correlations appear more pronounced in the wet season (May–October) for rice, while maize and soybeans exhibit more variability throughout the year. Regarding weather variables, rice yield demonstrates higher sensitivity to precipitation, potential evapotranspiration, and vapor pressure, while maize is more sensitive to diurnal temperature and precipitation. In contrast, the effects of weather on soybean yield are less pronounced. This study highlights the historical influence of climate on yield stagnation and variability, providing a foundation for understanding past agricultural trends in Thailand. Additionally, these findings contribute to the development of a Southeast Asia database for long-term agricultural monitoring and climate adaptation strategies.