Modeling the effects of alkalinity on morphological and photosynthetic traits of two spinach cultivars using artificial intelligence

Spinacia oleracea L. is a cool-season vegetable of high nutritional and economic value. Soil alkalinity is a major constraint to spinach productivity, and understanding genotype-specific tolerance is essential for improving resilience. This study assessed morpho-physiological responses of Iranian and Danish cultivars under three alkalinity levels (0, 15, 25 mM NaHCO3) using a completely randomized factorial design. Root length, chlorophyll fluorescence parameters (Fv/Fm, PI), SPAD index, and photosynthetic pigments were measured. Nonlinear regression and machine learning models were used to extend predictions across 0–50 mM. Increasing alkalinity caused measurable changes in all traits. Root length increased by 18–32% at 25 mM in both cultivars, with the Danish showing more stable responses. Fv/Fm declined from 0.82 to 0.71 in the Iranian cultivar and from 0.83 to 0.75 in the Danish one. SPAD values decreased by 22% in the Iranian and 14% in the Danish. Chlorophyll a declined by 28% and 17%, respectively. AI-based modeling predicted trait responses up to 50 mM with < 10% deviation. The complementary tolerance patterns of the two cultivars, together with the accurate AI-based predictions, demonstrate the potential of integrating physiological measurements with artificial intelligence to support breeding for alkalinity-tolerant spinach.”