Optimising maize sowing under irregular rainfall distribution in northern Thailand: A model approach for improved decision-making

Abstract

Rainfed maize production in northern Thailand is strongly affected by high rainfall variability and recent apparent shifts of the onset of the rainy season. Hence, decision-making on maize sowing time has become extremely difficult for farmers. This study thus aimed at (i) evaluating the Water, Nutrient and Light Capture in Agroforestry Systems (WaNuLCAS) model for predicting maize performance under rainfed conditions, (ii) assessing maize yield performance under five sowing date options, and (iii) identifying the best sowing option under irregular rainfall. A 2-yr-data set with various maize sowing dates from a field experiment in northern Thailand was used to calibrate and validate the model. WaNuLCAS was able to predict maize yield well (goodness-of-fit statistics: R²=0.83; modelling efficiency: −0.61; root mean square error: 0.14 kg m −2 ; maximum error: 0.16 kg m −2 ; coefficient of residual mass: 0.02; coefficient of determination: 0.56). An analysis of rainfall data (1970–2018) of the Phitsanulok province, Thailand, showed strong interannual variations: 27.1% of the years corresponded with the long-term mean and were moderately dry or moderately wet, while the remaining years were either very wet (10.4%) or very dry (8.3%). The standardized precipitation index indicated both, an earlier rain onset and frequent dry spells towards the end of the rainy season in recent years. Therefore, sowing earlier can be seen as a valuable coping strategy. Simulated sowing date options were: farmers’ practice (FP), 15, 30, and 45 days before FP, and a combination of them (staggered planting). Simulations revealed that under current rainfall conditions water was the most limiting factor for growth and yield of maize, while nutrient limitations had little impact under the current fertilization regime. Maize water uptake was significantly correlated with yield (R²=0.45). The simulation results suggest that sowing maize 30 days before FP or staggered planting are potent alternatives under irregular rainfall, the later particularly when distinct weather forecasts are not possible. Both options reduced the risk of crop failure while maintaining yields under these conditions. Simulations further suggested, under climate scenarios (RCP 4.5) for northern Thailand and the 21st century even earlier sowing seems plausible. • Validation of a model approach to improve decision-making in maize sowing. • Analysis of past precipitation patterns to understand the impact of rainfall variability on maize growth and yield • Testing coping strategies to better adapt to climate change effects on maize cropping. • Early sowing or staggered planting of maize are potent alternatives to cope with climate change projections for northern Thailand.