Sensor-based assessment of fertilizer strategies in soybean: linking SPAD, NDVI, plant height, and thermal imaging with biomass accumulation

This study aimed to investigate the temporal effects of different fertilization strategies on the physiological, morphological, and biomass-related traits of soybean under controlled greenhouse conditions. Individual and combined applications of urea, zinc (Zn), and microbial inoculants were evaluated using a multi-sensor approach. Optical parameters (SPAD, NDVI), plant height, and thermal imaging were monitored across days after onset (DAO), and post-harvest biomass traits were measured to establish integrative relationships. The findings demonstrated that SPAD values increased by 18-27% and NDVI by up to 22%, with combined treatments (Urea + Microbial, Zn + Microbial) showing the most pronounced improvements. Urea + Microbial increased plant height by 15% and fresh biomass by 28% compared to the control. Thermal imaging revealed a 1.8-2.5 °C reduction in canopy temperature under combined treatments, indicating enhanced stomatal regulation and water-use efficiency. Strong positive correlations (r = 0.71-0.84) between SPAD/NDVI and post-harvest biomass confirmed the reliability of early-stage sensor measurements for predicting yield-related traits. Importantly, the integration of microbial inoculants with mineral fertilizers enhanced both physiological resilience and water-use efficiency, while the identification of tentative threshold values for SPAD (~ 35) and NDVI (~ 0.60) provides practical benchmarks for fertilization decisions and automation in precision agriculture. Overall, this study highlights the utility of combining optical and thermal sensing with morphological and biomass assessments to optimize fertilization strategies in soybeans. The results provide novel insights into the role of micronutrient (Zn) and microbial management in crop monitoring and underline the potential of sensor-based approaches to improve nutrient efficiency and support sustainable agricultural production.