Abstract
An effective and economically efficient phosphorus (P) management strategy requires accounting for the residual value of P fertilizer. Despite a great deal of research on soil P much remains unknown about the residual value of fertilizer P to upland rice in the humid tropics of Africa. A field experiment was conducted for six years (1993–1998) to determine the response of four promising upland rice cultivars to fertilizer P applied at 0, 45, 90, 135, and 180 kg P ha−1 as triple superphosphate (TSP) only once in 1993, and its residual value in 1994, 1995, 1996, and 1998. The experimental site was located on an Ultisol, low in available P, in the humid forest zone of Cote d'Ivoire. Grain yields of the rice cultivars were significantly increased by fertilizer P in 1993, and by the fertilizer P residues in the subsequent years although the magnitude of response decreased rapidly with time since the fertilizer was applied. The levels of Bray 1 extractable P in the soil fell below the critical value in the 1994 season in all treatments except those which received 135 and 180 kg P ha−1 in 1993, and fell below the critical value in all P treatments in the 1995 season. A decrease in available P coincided with decreased residual value of P fertilizer and the soil P test (Bray 1 P) appeared to predict well the crop response. Our results show that there was a rapid loss of available P in the Ultisol and it is suggested that annual application of light rates of soluble P is a better option than heavy application at one time.