Integrating quantity–intensity relationships and machine learning to assess potassium dynamics and plant uptake in calcareous soils of India
摘要
The amount of potassium (K⁺) that plants can use might not be fully shown by exchangeable potassium numbers, since this method doesn't take into account the effect of non-exchangeable potassium (NEK). K⁺ availability and release in soils can be assessed through quantity-intensity (Q/I) relationships. A study was conducted in the calcareous regions of Muzaffarpur district, Bihar, focusing on the potassium concentrations in rice roots, shoots, and grains, as well as various soil characteristics. 92 samples were analyzed to determine the different K⁺ forms present in the soil as well as assessed NEK reserves and Q/I isotherms. The potential buffering capacity of Zone 1 (24.87 cmol kg−1 (mol L−1)−1/2) was higher than Zone 2 (21.67 cmol kg−1 (mol L−1)−1/2). Zone 1 exhibits an elevated equilibrium activity ratio (ARe0K) than Zone 2. The free energy values suggest that soil from both zones has moderate to significant K+ deficiencies. A positive correlation was observed between the exchangeable and NEK forms of K+ and Step-K and CR-K. AReK exhibited a positive correlation with K+ saturation, K0, -ΔG, KL, KV, and KKDO. The potassium concentration in rice is greatest in the grains, followed by the shoots, and least in the roots. Zone 1 soil exhibited the highest availability of potassium. Random Forest models accurately predict potassium availability and uptake, thereby enhancing soil fertility and precision agriculture, which in turn leads to improved crop yields and soil health. Consequently, comprehending the dynamics of potassium release and availability in calcareous soils, is essential for effective fertilizer management.