Methodology Of Fertilizer Recommendation Based On Yieldresponse And Agronomic Efficiency For Rice And Maize

Low nutrient use efficiency caused by irrational fertilization, and difficulties of soil test based fertilizer recommendation for smallholder farms due to constraints such as access, cost and timeliness in multiple cropping systems drive the great need to establish a fertilizer recommendation method based on yield response and agronomic efficiency. The Quantitative Evaluation of the Fertility of Tropical Soil(QUEFTS) model was adopted to analyze the relationship between grain yield and nutrient uptake to obtain the balanced nutrient uptake for different rice plant types in term of season based on large datasets from multiple years and sites. Yield response, agronomic efficiency, indigenous nutrient supply, and yield gap under different nutrient management practices were analyzed as well. Through the above analyses, nutrient management principles and fertilizer recommendation methodology based on yield response and agronomic efficiency was formed for rice. By using computer information technology, a user-easily and friendly Fertilizer Decision Support System- Nutrient Expert for Rice(NE), though containing science-based fertilizer recommendation principles was built, and field validation was conducted to further test and optimize parameters related to fertilizer recommendation. Meanwhile, multi-site field experiments were conducted to validate the effects of Nutrient Expert for Hybrid Maize from agronomic, economic and environmental aspects. The main results are as follows:(1) Totally 2,218 on-farm experiments were collected in major rice production regions of China from2000 to 2013, to determine the relationship between grain yield and nutrient uptake in theabove-ground plant dry matter using the QUEFTS model. The QUEFTS model predicted a linearncrease in grain yield if nutrients were taken up in balanced amounts until yield reached about60–70% of yield potential. To produce 1000 kg of grain for single-season rice, 14.8 kg N, 3.8 kg P,and 15.0 kg K were required in the above-ground plant dry matter, and the corresponding internalefficiencies(IEs) were 67.6 kg grain per kg N, 263.2 kg grain per kg P, and 66.7 kg grain per kg K.Correspondingly, to produce 1000 kg of grain for early/middle/late rice, the amount of 17.1 kg N,3.4 kg P, and 18.4 kg K were required in the above-ground plant dry matter, and corresponding IEswere 58.5 kg grain per kg N, 294.1 kg grain per kg P, and 54.3 kg grain per kg K. The N:P:K ratiosfor the linear part of the relationship were 3.89:1:3.95 for single-season rice, and 5.03:1:5.41 forearly/middle/late rice.(2) Nutrients uptake by grain was also simulated by QUEFTS model. Nutrient requirements in 1000kg grain were similar for both single-season rice and early/middle/late rice. The requirement of N,P, and K were 10.6, 2.6, and 3.2 kg for single-season rice, 10.6, 2.7, and 3.1 kg forearly/middle/late rice, respectively. When the target yield reached 80% of potential yield, grainuptake of N, P, and K accounted for 72%, 69%, and 22% of total above-ground uptake forsingle-season rice, and 62%, 80%, and 17% for early/middle/late rice.(3) On average, yield responses to N, P and K fertilizer application were 2.4, 0.9 and 1.0 t/ha,indigenous N, P and K supplies were 91.3, 27.5 and 139.1 kg/ha, agronomic efficiency to N, P andK application were 13.0, 12.7 and 8.4 kg/kg, and relative yield to N, P and K were 0.71, 0.89 and0.89, respectively. Relative yield was used to classify the soil indigenous nutrient supply. The low,median and high classes of soil indigenous supplies were 0.63, 0.71 and 0.81 for N, 0.84, 0.91 and0.96 for P, and 0.85, 0.91 and 0.95 for K, respectively. A significant negative linear correlation wasobserved between yield response and relative yield, and significant quadratic relationship wasexisted between yield response and agronomic efficiency.(4) Principles and methodology of fertilizer recommendation for rice based on yield response andagronomic efficiency was established. Results from field validation indicated that grain yieldincreased by 4.2% and 2.6%, and net profit increased by 859 and 585 yuan/ha respectively byusing NE, as compared with farmers’ practices(FP) and soil test. Moreover, the recovery efficiencyof N(REN) improved by percentage point of 9.2 and 6.2, the agronomic efficiency of N(AEN)improved by 2.8 and 1.8 kg/kg, and the partial factor productivity of N(PFPN) increased by 4.8and 3.0 kg/kg according.(5) The parameters related to Nutrient Expert for Hybrid Maize including the attainable yield, yieldresponse and relative yield were updated. The relative yield used to classify the low, medium andhigh soil indigenous nutrient supply were 0.70, 0.82 and 0.90 for N, 0.86, 0.91 and 0.96 for P, and0.83, 0.89 and 0.95 for K, respectively.(6) The results from field validation across seven maize production provinces showed that, ascompared to FP, the NE treatment achieved up to 6.1% more grain yield and 137-1711 yuan/hamore profit with 30.4% of less N and 11.3% of less P fertilizer, and more K fertilizer. The NEimproved AEN by 3.9 kg/kg, REN by percentage point of 10.2, and PFPN by 17.2 kg/kg,respectively. More importantly, NE decreased apparent N loss by 35.6 kg/ha as compared with FP.Three-year fixed-site experiment for maize showed that, as compared with FP, the NE treatmentobtained 0.6 t/ha more yield and 1421 yuan/ha more profit with 66 kg/ha of less N fertilizer. Theyield gain by the NE treatment shared 91.3% of total profitability. The NE improved RE of N, Pand K by percentage point of 11.3, 7.6 and 7.7, AE of N, P and K by 5.4, 9.3 and 4.8 kg/kg, andlowered soil residual N and N loss significantly. Field validation also suggested that, NE had thecomparable yield and 328 yuan/ha more profit than OPTS, with 21.2%, 6.9% and 7.9% of less N, Pand K fertilizers. With this, the NE improved REN, AEN and PFPN by percentage point of 6.0, 2.5kg/kg and 12.9 kg/kg, respectively.In summary, the fertilizer recommendation method based on yield response and agronomic efficiency was proved to be a science-based, reliable and feasible method and can be used to make fertilizer recommendation for both rice and maize.