Response of Rice, Maize and Millet to Fertilizers in Mali

Abstract

Maize, rice and millet yield are very low in Mali due to low fertilizer use. The research results reported here are based on work carried out in the different agroecological zones (AEZs) in Mali for Africa, with a structure of treatments allowing the determination of the response functions for nitrogen (N), phosphorus (P) and potassium (K) on maize, rice and millet. A diagnostic treatment including Zn, Mg, and B was compared with the same NPK treatment to evaluate the effect of micronutrients. The results will be combined with research conducted under similar agro-ecological conditions to develop robust response functions for the different crops and ensure greater profit in fertilizer use decisions.

The fertilizer rates that were tested within the framework of the study showed different effects according to crops and AEZs. The combinations across the three agro-ecological zones showed that higher rates such as those blankets recommended would not be economically profitable. The same observation was made for conventional and hybrid maize. For maize, the best combinations across the three agro-ecological zones were: 90N 10P 0K, 90N 30P 0K, 30N 40P 0K, 120N 30P 0k, 60N 30P 10K, 60N 30P 30K.

Application of N, P, and K did not affect yields of upland rice, but there were significant differences in the yields of lowland rice and irrigated rice. Average yields were 2.482, 2.808 and 4.71 t/ha for upland, lowland and irrigated rice respectively. A diagnostic treatment including N, P, K, Mg, S, Zn, and B had a significant effect on lowland rice (rain fed and irrigated) but not on upland rice compared to N, P, and K alone. Nutrients effects were specific to each area and production conditions. The Fertilizer Optimization Tool (FOT) developed in the framework of the OFRA project allowed to propose economic rates lower than the current recommendations and for each type of rice cropping system depending on the agro-ecological zones.

The fertilizer rates that were tested showed significant differences between treatments, regardless of soil type. As sole crop, millet gave yields significantly higher than that usually found in farmers’ fields around the station (1500 average against 800 to 1000 kg/ha). It is concluded that the application of 22.5 P combined with 40 or 60N and that of 30P without N and K significantly improve millet grain yield. The best yields were achieved with P rates higher than that recommended (46P2O5) for millet fertilization.