Residual P in river terrace soils of the West African humid uplands sustained satisfactory yields of Funk G-901 maize (Zea mays L.), Anethoda rice (Oryza sativa L.), and Mares groundnut (Arachis hypogaea) under a sequence of continuous cropping for 3 consecutive years. Up to 90 kg/ha soluble P applied at the inception of the cropping sequence failed to elicit a yield increase when other cultural parameters were maintained at levels recommended as adequate for satisfactory yields under these conditions.
At the start of the experiment the soil had already attained a residual P level of 55 kg/ha, the Ca + Mg concentration averaged 0.5 meq/100 g soil, and fresh crop and cover residues contributed 18 metric tons/ha of organic matter to the 15-cm soil depth each year.
Lime applied in the last year raised soil pH from 4.8 to 5.1, reduced Al base-saturation from 30% to 9%, and probably contributed to the increase in groundnut yield, although this effect was not experimentally tested. In the 3rd year a second application of the P treatments after liming did not induce a significant increase in yield over the check (0-P) or in those plots which had received only the P treatment initially.
In soils where exchangeable Al comprised 30% or less of the base saturation, the conversion of soluble P into insoluble Al and Fe phosphates apparently was not a yield-limiting reaction. Phosphorus fixation, particularly as Al phosphates, is a recalcitrant problem for soils in the humid tropics, but this is not true for soils under sustained cropping. By using the same crop varieties and soils similar to those used in this study, subsistence farmers could intensify their operations on selected sites utilizing residual P for extended periods without soil P becoming a yield-limiting factor.