Assessment of productivity of red paddy soil based on soil fertilityand relative yield

Objectives  To study soil fertility improvement and yield based on the quantitative relationship between soil fertility index (SFI) and yield.

Methods  This study is based on the long-term fertilization experiment and county-scale sample surveys of red paddy fields in Jinxian County, Nanchang City, Jiangxi Province of China. The long-term field experiment was started from 1981, which included no fertilization (CK), nitrogen fertilizer only (N), phosphate fertilizer only (P), potassium fertilizer only (K), combined nitrogen and phosphorus fertilizers (NP), combined nitrogen and potassium fertilizers (NK), combined nitrogen, phosphorus and potassium fertilizers (NPK), double the amounts of nitrogen, phosphorus and potassium fertilizers (2NPK), combined organic and inorganic fertilizers (NPKM). Then, Soil pH, organic matter, available phosphorus, readily available potassium and rice yield were measured in 1981, 1985, 1989, 1995, 2002, 2005, 2012 and 2017. Meanwhile, the soil pH, organic matter, available P, readily available K and late rice yield were also analyzed from 58 representative sites of Jinxian County in 2017. SFIs of different treatments in a long-term fertilization experiment were evaluated by the Fuzzy method in a red paddy soil and the quantitative relationship between SFI and relative yield was established and validated.

Results  Different fertilization treatments affected rice yield and SFI significantly. Compared to CK, the average yield of NPKM, 2NPK and NPK were increased by 71. 6%, 59.1% and 36.5%, respectively. Their SFIs were increased by 64.1%, 47.3% and 27.7%, respectively. Compared with NPK treatment, the average yield and SFI in NPKM treatment were improved by 25.6% and 28. 5%. In contrast, the yield of NP and NK treatments and their SFIs decreased significantly. The average yield and SFI among N, P and K treatments were not significantly different from the CK. Furthermore, SFI and relative yield showed a linear relationship (P < 0.05), and the relative yield was improved by 7.2%−22. 6%, while the SFI was increased by 0.1. The validation experiment on the county scale showed that the estimated relative yield closely matched the measured values (R² = 0.5268, P < 0.01, RRMSE < 25%), suggesting that the SFI could indicate rice production on the county scale.

Conclusions  Application of organic and chemical fertilizers is a key technology for improving SFI and yield in the red paddy soil. Moreover, the quantitative relationship between SFI and relative yield could guide the management of soil fertility in the red soil.