Use of numerical methods in determining soil fertility affinities in three contiguous sand dune fields of NW Nigeria
1995
Raji, B.A. | Chude, V.O. (Ahmadu Bello Univ., Zaria (Nigeria). Dept of Soil Science) | Esu, I.E.
Sixty sand dune surface soils from twelve sites were classified both by a non-numerical classification (Soil Taxonomy) and by three numerical classifications, based on (1) all 21 properties which were measured in the field and laboratory, (2) on properties measured its the field (morphology) only, and (3) on laboratory properties related to soil fertility. Two broad groups, Haplustults and Quartzipsamments, were obtained using Soil Taxonomy. However, using the numerical classification, four main groups and 2 unattached soil sites were delineated at the lowest level and these show a pattern related to the geographical distribution, meso-relief thickness and the underlying geology of the dunes. Comparison of the different classifications shows that there is broad agreement between them and between classifications based on different suite of properties. Even classification based on field properties only, showed close similarity to others. Results of this study also indicate a possibility of using colour value of dry soil as an index of soil ferlility in the dune soils. The close similarity between the numerical and the non-numerical systems of classification may be attributed to the fact that properties which relate largely to nutrient status and exchange capacity of the soils, are those on which the numerical classification was based and these properties are nearby, the same as those on which the non-numerical classification is also based. The non-numerical classification however, gave broad groupings and therefore seems not to be totally satisfactory in defining fertility affinity in closely related soil series especially when absolute values of soil properties are below critical limits.
Mostrar más [+] Menos [-]Palabras clave de AGROVOC
Información bibliográfica
Este registro bibliográfico ha sido proporcionado por University of Liège