Effect of Soil Properties on Tree Distribution across an Agricultural Landscape on a Tropical Mountain, Tanzania

Abstract

Distribution of tree species can result in many factors including environmental variables, biotic interaction and management. Better understanding of these interacting factors is crucial in planning management interventions especially in managed landscapes: this study on the effect of soil properties on tree species distribution in Kilimanjaro, Tanzania will aid in this. Standard tree inventory procedures and soil sampling techniques were used to survey 48 plots from altitudinal level of 680 to 1690 m a.s.l. along 25 km long transect. All trees ≥ 5 cm at DBH were recorded, while soils were sampled from top and subsoils (0 - 20 and 21 - 50 cm depths). Tree species distribution index was assessed through abundance and frequency, while species interaction with environmental variables was assessed using Detrended Correspondence Analysis. Distribution index indicated that 77% of tree species were categorized as rare, while 10% and 13% were categorized as occasional and abundant respectively. Soil organic carbon and moisture content have shown high correlation with tree species (r > 0.8, p < 0.01), while ExMg, soil pH, P, ExCa, ExK, ExNa and bulky density indicated less correlation (r < 0.2, p < 0.001). The DCA-1 axis explained nearly 70% of the relationships between soil properties and tree species distribution: suitability of tree species were influenced by soil properties across the land use systems which exhibited different soil types. Different tree species communities correspond differently with soil properties between the land use systems. Fewer tree species spread in the lowland, which is known to have saline soils. Therefore, despite intensive human management of the landscape, tree species indicated distribution patterns in line with the soil properties.