Site-specific nutrient management for rice using soil properties to adjust phosphorus and potassium supply from compound NPK fertilizer

Abstract

Context or problem

Compound fertilizers containing nitrogen (N), phosphorus (P), and potassium (K) are often applied to rice (Oryza sativa L.) in Asia, but the fixed P:K ratio of these fertilizers can restrict the adjustment of P and K rates to match crop needs.

Objective or research question

We hypothesized that site-specific nutrient management (SSNM) using a locally available NPK fertilizer could increase rice yield and then that soil properties could be used to better adjust nutrient inputs to match crop needs.

Methods

Yield responses to added N, P, and K in on-farm nutrient omission plot technique (NOPT) trials across Indonesia were used to develop an SSNM practice, which was evaluated relative to farmer’s fertilizer practice (FFP) in 528 on-farm trials with lowland rice. Relationships of soil properties with yield responses were then examined across 38 additional trials on diverse soils.

Results

An SSNM practice with one NPK fertilizer supplying all P and K increased yield by 0.4 Mg ha⁻¹ and increased gross return above fertilizer cost by 173 USD ha⁻¹ compared to FFP. The SSNM practice used less N and P but slightly more K than FFP. Median relative yields — defined as the ratio of yield in a nutrient omission plot and yield in a full-fertilizer plot — were 0.75 without added N, 0.94 without added P, and 0.94 without added K for 252 NOPT trials. Based on NOPT trials, the SSNM practice could be improved by applying more K at some locations because the P:K ratio of 0.52 (P₂O₅:K₂O = 1) for the NPK fertilizer was higher than the estimated optimum to meet crop needs. Higher K requirement was associated with soil pH < 5.5, soil K saturation < 1.8 %, and soil base saturation < 40 %.

Conclusions

Farmer’s P use for rice can often be reduced with an upper limit for the P rate set to match the net removal of P by the crop. An NPK fertilizer with P:K ratio = 0.44–0.52 would meet P and K requirements at many locations, and a mid-season topdressing of additional K could be targeted to locations requiring more K. Higher K use could target areas with soil pH < 5.5, which could be approximated from soil maps and verified with soil test kits.

Implications or significance

The examination of associations between crop response to a nutrient and soil properties, other than soil nutrient status, merits use elsewhere to identify soil characteristics helpful in fine-tuning SSNM.