Machine learning the abiotic stressor impacts on nitrogen availability and photo energy use in dryland forage systems under different tillage and green manuring practices

Sorghum sudangrass (SSG), a vigorous hybrid C4 plant grown as forage or bioenergy crop during the summer has deeply distributed fibrous root system beneficial to dryland agriculture. Despite these advantages, the crop faces challenges from abiotic stressors, particularly during severe drought seasons. Thus, the goal of planting SSG as a forage is to generate ample biomass with minimal inputs, minimize toxicity, identify stress factors, and implement suitable management strategies with negligible environmental impact. This study focused on machine learning the intricate soil–plant-water-atmosphere relationships associated with abiotic stresses and agroenvironmental variables to enhance energy-efficiency and alleviate nitrogen (N) stress. A long-term simulation study involving SSG with four N treatments (spring-fallow and N fertilization, and three spring-time cover with 100% residue addition as green manures) under no-till (NT) and conventional (CT) tillage systems was unsuccessful to provide distinct effects of N strategies. Machine learning (ML) was employed to determine how abiotic-stress factors were varied, specifically under influence of green manures such as oat (OG), field peas (FP) and grass peas (GP) under conventional and conservation systems. Results indicated significantly higher use-efficiency of photosynthetically-active radiation (PAR) and the lowest N stress in NT + FP contrasting with the lowest PAR and the highest N-stress in conventional system. ML Regression models highlighted significant impact of average daily temperatures on water (R2 of 78%) and N-stress (R2 of 79%). Multi-regression ML models, incorporating various agro-environmental variables identified tillage as the most influential factor during wet seasons, water-stress during normal seasons, and temperature as the predominant factor during drought seasons influencing N stress in SSG forage-system. Conspicuously, drought exhibited a more pronounced influence from green manuring compared to wet or normal seasons in our study. Thus, farmers should be encouraged to adapt the practice of using cover and green-manures despite the concerns about water reserves for summertime cash crops.