Securing fruit trees future: AI-driven early warning and predictive systems for abiotic stress in changing climate

Abstract

• Artificial intelligence (AI) and machine learning enhance early detection, predictive analytics, and remote sensing for abiotic stress mitigation in horticultural crops • Data accessibility, algorithmic biases, high implementation costs, and training requirements hinder AI integration in abiotic stress management • Advanced machine learning models like Random Forest and Gradient Boosting offer effective solutions for managing complex abiotic stress data • AI-powered tools optimize irrigation, nutrient management, and phenotyping, improving crop resilience under climate change • Cross-crop variability, poor physiological alignment, and socio-economic hurdles hinder real-world adoption of AI-driven stress management in horticulture • Implementing AI-based systems can drive carbon-neutral precision farming, ensuring food security and climate resilience in horticultural fruit crops Horticultural fruit crops are important product for global food security as they provide important macro and micronutrients. Emerging abiotic stresses—drought, salinity, temperature extremes, and waterlogging—threaten crop productivity by disrupting key physiological processes, challenging sustainable agriculture under climate change. Traditionally, these challenges are tackled through physiological strategies and conventional breeding methods; which lacks scalability, adaptability, and real-time feedback. Advancement in technology, distinctly artificial intelligence (AI), and machine learning have revolutionized precision farming in agriculture. AI integrated approaches such as stress prediction, irrigation optimization, and image-based phenotyping have enhanced agriculture, while machine learning models like Random Forest and Gradient Boosting improve stress management. However, the acquiring of such AI based system in horticultural crops still faces numerous challenges. Specifically, multi-omics, data accessibility, algorithmic biases, the cost of implementation and requirements for robust training programs need to integrate for sustainable agriculture. Addressing these barriers is essential to enable the effective integration of AI technologies for resilient, carbon-neutral fruit production under climate stress. This review highlights recent developments in AI-driven stress detection, predictive modeling, remote sensing, and machine learning frameworks that support precision horticulture and climate-smart agriculture.