Statistical analysis of soil variability in a humid forest landscape of Central Cameroon

Abstract

The quantitative evaluation of key factors of soil variability together with the spatial pattern of underlying variables are important steps in understanding the functioning of soilscapes. Fifteen physical and chemical variables describing 24 pedons were used to elucidate processes of soil differentiation in a humid forest environment in Southern Cameroon. Landscape positions delineated on the basis of remote-sensed imageries included flat convex summits (FCS), linear valley slopes (LVS), colluvial inland valleys (CIV), and alluvial floodplain. Ordination by principal component analysis (PCA) resulted in a three-factor model, which efficiently summarized the dataset explaining 81.4% of total soil variance (TSV). The pedogenetic factor of ferralitization emerged as the main factor accounting for 43.7% of TSV. Identified underlying processes included basic cation leaching, solum acidification, and in situ clay build up. Cation exchange properties as influenced by organic constituents and soil available P were the second and third most important factors associated with 26.7 and 11.4% of TSV respectively. Organic constituents and related biological processes had strong control on nutrient availability and cation-exchange properties, while soil available phosphorus (AP) showed a strong dependence on potential acidity and organic matter mineralization status. The analysis of variance highlighted significant differences for most soil properties across landscape position and soil depth except for organic carbon (OC), total nitrogen (TN), and total phosphorus. All the soils, however, have low pH, low basic cation saturation, and correlatively high exchangeable Al irrespective of their position in the landscape.