Potential nutrient supply, nutrient utilization efficiencies, fertilizer recovery rates and maize yield in northern Nigeria
Tabi, Fritz Oben × Diels, Jan Ogunkunle, Ayo O Iwuafor, Emmanuel N. O Vanlauwe, Bernard Sanginga, N #
Nutrient cycling in agroecosystems vol:80 issue:2 pages:161-172
Potential N (SN) and P (SP) supplies, N and P utilization efficiencies and fertilizer recovery rates for the northern Guinea Savanna (NGS) agro-ecological zone of Nigeria were derived from data collected on farmers' fields, and used as input in the QUantitative Evaluation of the Fertility of Tropical Soils (QUEFTS) model. The potential N supply ranged from 7 to 56 kg N ha(-1), with a mean of 25 kg N ha(-1), while SP ranged from 2 to 12 kg P ha(-1) with a mean of 5 kg P ha(-1). Both SN (CV = 42%) and SP (CV = 57%) were highly variable between farmers' fields. Deriving potential nutrient supply from 'a' values gives lower estimates. The empirical equation in QUEFTS that estimates SN (SN = 1. 7 X OC X (pH - 3)) sufficiently predicted the SN of soils in the NGS (RMSE = 8.0 kg N ha(-1) index of agreement (IOA) = 0.81). The SP equation (SP = 0.35 X (1 - 0.5 X (pH - 6)(2)) X OC + 0.5 X OlsenP) predicted moderately potential P supply (RMSE = 6.80 kg P ha(-1), IOA = 0.54). When N or P is maximally accumulated in the plant (i. e., least efficiently utilized), the utilization efficiency was 21 kg grain kg(-1) N taken up and 97 kg grain kg(-1) P taken up. When these nutrients were maximally diluted in the plant (i. e., most efficiently utilized), the utilization efficiency was 70 kg grain kg(-1) N taken up and 600 kg grain kg(-1) P taken up. The range in N recovery fraction (NRF) of N fertilizer applied was from 0.30 to 0.57, with a mean of 0.39, while the P recovery fraction (PRF) ranged from 0.10 to 0.66 with a mean of 0.24. Although SP was moderately predicted, when QUEFTS model input parameters were adjusted for the NGS, the model sufficiently (IOA = 0.83, RMSE = 607 kg DM ha(-1)) estimated maize grain yield in the NGS of Nigeria. The original QUEFTS model however, gave better predictions of maize grain yield as reflected by the lower RMSE (IOA = 0.84, RMSE = 549 kg DM ha(-1)). Consequently, QUEFTS is a simple and efficient tool for making yield predictions in the NGS of northern Nigeria.