Multivariate analysis of soil productivity indicators in the northern guinea savana agro-ecology of Nigeria


J.O. Omeke*, K.C. Uzoma and J.F. Akpan

A trial was conducted at the Research Farm of the Institute for Agricultural Research, Ahmadu Bello University (IAR/ABU), Samaru, Zaria during the year 2011 and 2012 cropping seasons to determine the productivity of the Alfisol using multivariate data analysis. The result indicated that the factor loadings of the soil properties for the PCA analysis showed their contribution to soil productivity, which was explained by 55.38% of the proportion in PCI being higher than other PCs. Among the soil properties evaluated, ECEC, CEC, Ca, Mg, OC, TN, pH (chemical properties), bulk density and porosity (physical properties) dominated in PC1, which cumulatively contributed 55.38% of the total variation in soil productivity. However, negative loading was only observed in bulk density and soil pH. Generally, the factor loading of the soil properties indicated that individual contribution to soil productivity was in the order of: ECEC>CEC>Ca>OC>TN>pH>Porosity>BD>Mg>P in PC1 which was contrary under PC2 in terms of P. The results of stepwise multiple regression revealed that in step 1; N uptake (x3) was retained with R2=0.92, while other 5 variables were removed. A similar trend was observed in step 2 (x3 and x5; R2=0.97) and 3 (x2, x3 and x5; R2=0.97). But in step 4, four independent variables were retained (stover dry matter x1, harvest index x2, N uptake x3 and N utilization efficiency x5) with R2=0.98, which justified the maximum maize grain yield changes. This implies that integration of inoculated soybean in maize-based cropping systems in combination with minimum disturbance of the soil would enhance soil productivity.

Share this article