ASSESSING IRRIGATION WATER USE EFFECIENCY FOR RICE CULTIVATION USING EXPERIMENTAL AND MODELING APPROACH

Cambodia is characterized by a tropical climate influenced by monsoon winds, causing 6 months dry and 6 months wet. The agricultural sector, of which the main product is rice, accounts for 24.4% of Cambodia's GDP. However, rice consumes a lot of water, which ranges from 1100 to 2500 mm per growing season, whereas the average annual precipitation of Cambodia was only between 1400 mm and 4000 mm. Insufficient irrigation limits the ability of Cambodian farmers to grow rice in the dry season. As a result, out of 3.58 million paddy land, only 19% were covered by dry-season or irrigated rice. The remaining rice cultivation areas are only possible for the rainfed rice.
Under these conditions, increasing water production in irrigated areas for dry-season rice was considered a more successful approach to improving food and water security. In addition, rice is very vulnerable to climate change. A dramatic temperature change is likely to affect agricultural productivity. For example, under heat stress, rice grain yield will decline by 10% for each 1oC increase in growing-season minimum (night) temperature in the dry season. Rice yield in Cambodia was projected for a moderate decrease (-10.5%) under RCP 8.5 by the mid-century.
Alternative wetting and drying (AWD) is one of the water-saving methods. However, more research is needed to examine AWD in depth for different varieties and soil texture. In addition, under potential climate change conditions, rice undergoing the AWD cycle may face drought alongside extreme heat. Investigating this hypothesis to confirm the viability of using the AWD method for dry-season rice in Cambodia as an adaptive technique is essential.
In this research, we first evaluate the possibility of saving water in dry rice production in Cambodia by quantifying the effects of AWD on rice yield and water use efficiency (WUE) with varying varieties and soil properties. We tested AWD at two threshold levels: safe AWD (AWD15) and moderate AWD (AWD20). Five field experiments were conducted from 2021 to 2023. Associative causal relationships between measurable and latent variables of rice grown under CF and AWD were tested using partial least squares structural equation modeling (PLS-SEM) and an analysis of variance (ANOVA).
Secondly, we assess AWD's potential to mitigate climate change's impact on short-cycle rice varieties with varying genotypes and maturity. We applied the AquaCrop Model and projected for climate change impact. We used two shared socioeconomic pathways scenarios (SSP3-7.0 and SSP5-8.5) developed within the Coupled Model Intercomparison Project Phase 6 (CMIP6) and projected the rice growth during 2040-2070. Data from field experiment 2023 was used to calibrate and validate the model. The results underscore the potential of AWD to mitigate the impact of climate change, providing hope for the future of rice production in Cambodia.
Our results showed that safe and mild AWD did not significantly affect grain yield, yield components, harvest index (HI), and root growth compared to conventional flooding (CF). Despite similar yields, AWD significantly reduced total water inputs by 10–30% in AWD15 and 22-24% in AWD20 compared to CF. Among the AWD treatments, AWD15 exhibited the highest WUE. Our study found that AWD15 and AWD20 treatments improved WUE over CF on sandy loam soil, and clay soil's WUE was higher than that of sandy soil. However, clay soil's WUE in AWD did not improve over CF. Our findings demonstrate that implementing safe and mild AWD has significant potential for rice growing on sandy loam.

Furthermore, the simulation results underscored the adaptability of AWD to different rice varieties. The effectiveness of AquaCrop in capturing crop development across treatments and varieties was demonstrated. The model's accuracy in simulating canopy cover (nRMSE = 14 - 32.5%), time series biomass (nRMSE = 22 - 42.5%), grain yield (Pd = 4.36 - 24.38%), and total biomass (nRMSE = 0.39 - 18.98%) was generally acceptable. The study found that varieties with growing periods longer than 93 days after transplanting (DAT), such as CAR15 and Sen Kra Ob, were most impacted by heat stress conditions, leading to reduced yield, harvest index (HI), and water use efficiency (WUE). In our case, CAR15 and Sen Kra Ob grain yields were reduced by 53% and 8%, respectively. However, AWD maintains superior WUE compared to CF regardless of the type of varieties, suggesting this technique is a robust and adaptable drought-adaptive strategy.
After all, this study underscores the need for further research to validate findings across a broader range of rice varieties and soil types, as it was conducted at a field experimental level. We also suggest a thorough investigation into the effect of heat stress on rice varieties and the water stress coefficient during severe AWD as this study did not address these aspects.