Weeds are a major constraint to upland rice production in west Africa. The objectives of this investigation were to study dierences in weed suppression ability among upland rice cultivars and to determine the morphological traits involved. Twelve contrasting cultivars, including West African indigenous Oryza glaberrima (Steudel) lines and traditional and improved O. sativa L., were cultivated under natural weed competition and low-input conditions in Coà te d'Ivoire in two seasons. Signi®cant dierences between cultivars were observed in weed biomass at 100 days after seeding (DAS) in 1994 and 1995, indicating dierences in their competitive ability. Weed biomass was negatively correlated with rice root growth at early growth stages and with rice shoot and root growth at later growth stages. Across cultivars, grain yields with a single handweeding were 60% in 1994 and 49% in 1995 of those in plots that were weeded regularly. High grain yields in weedy plots were associated with low weed biomass. IG 10, an O. glaberrima cultivar, was the most competitive against weeds.

No abstract

Field experiments were designed to investigate the effectiveness of integrated soil fertility management (ISFM), comparing fertilizer use efficiency and its impact on millet, cultivated close to the homestead (''infields'') and away from the homestead (''outfields''). Millet yields and response to N (0, 30, and 60 kg ha -1 ) and P (0, 15, and 30 kg ha -1 ) were determined on nine infields and nine outfields over a period of 3 years (from 1999 to 2001) in the southern Sahel of Niger. Rainfall was 650, 470, and 370 mm during the three successive years, interaction between decreasing rainfall and millet yield performance was also analyzed. While soil organic carbon (1.5 g kg -1 on outfields and 1.6 g kg -1 on infields) and pH-H 2 O (4.8 on outfields and 5.1 on infields) were comparable, total-N, plant available P (measured as P-Olsen and P-Bray), and exchangeable Ca, K, and Mg levels were higher on infields as compared to outfields. Without fertilizer, average grain yield (GY) and stover yield obtained on infields were three times as high as on outfields. GY across years and fertilizer treatments was higher on infields as compared to outfields (P \ 0.001). Average yield was 800 kg ha -1 on outfields and 1,360 kg ha -1 on infields (P \ 0.001). On outfields, average GY was stagnant over the 3-year experimental period. Despite declining rainfall, millet GY across all treatments gradually increased over time on infields (P \ 0.001). P fertilization alone resulted on both field types to steadily and substantial yield increases while yield response to N fertilization was only obvious when fertilizer P was applied. With no fertilizer applied, N uptake on infields (19 kg N ha -1 ) was more than twice as high as on outfields (7 kg ha -1 ), and P uptake was four times higher on infields (3 kg ha -1 ) than on outfields (0.8 kg ha -1 ). Indigenous soil N supply was on average 24 kg N ha -1 on outfields and 46 kg N ha -1 on infields. Average value for indigenous soil P supply was 4 kg P ha -1 on infields and 2 kg ha -1 on outfields. Apparent recovery of fertilizer N applied varied considerably among treatments and ranged from 17 to 23% on outfields and 34 to 37% on infields (P \ 0.001). Average apparent recovery of fertilizer P applied was significantly higher (P \ 0.001) on infields (31%) than on outfields (18%) over the 3-year growing period, illustrating ISFM-induced positive effect on millet

Maize is the primary food crop grown by farmers in the coastal savanna region of Togo and Benin on degraded (rhodic ferralsols), low in soil K-supplying capacity, and non-degraded (plinthic acrisols) soils. Agronomic trials were conducted during 1999-2002 in southern Togo on both soil types to investigate the impact of N and P fertilization and the introduction of a mucuna short fallow (MSF) on yield, indigenous N supply of the soil, N recovery fraction and internal efficiency of maize. In all plots, an annual basal dose of 100 kg K ha À1 was applied to the maize crop. Maize and mucuna crop residues were incorporated into the soil during land preparation. Treatment yields were primarily below 80% of CERES-MAIZE simulated weather-defined maize yield potentials, indicating that nutrients were more limiting than weather conditions. On degraded soil (DS), maize yields increased from 0.4 t ha À1 to 2.8 t ha À1 from 1999 to 2001, without N or P application, in the absence of MSF, with annual K application and incorporation of maize crop residues. Application of N and P mineral fertilizer resulted in yield gains of 1-1.5 t ha À1 . With MSF, additional yield gains of between 0.5 and 1.0 t ha À1 were obtained at low N application rates. N supply of the soil increased from 10 to 42 kg ha À1 from 1999 to 2001 and to 58 kg N ha À1 with MSF. Application of P resulted in significant improvements in N recovery fraction, and greatest gains were obtained with MSF and P application. MSF did not significantly affect internal N efficiency, which averaged 45 kg grain (kg N uptake) À1 . On non-degraded soils (NDS) and without N or P application, in the absence of MSF, maize yields were about 3 t ha À1 from 1999 to 2001, with N supply of the soil ranging from 55 to 110 kg N ha À1 . Application of 40 kg P ha À1 alone resulted in significant maize yield gains of between 1.0 (1999) and 1.5 (2001) t ha À1 . Inclusion of MSF did not significantly improve maize yields and even reduced N recovery fraction as determined in the third cropping year (2001). Results illustrate the importance of site-specific integrated soil fertility management recommendations for the southern regions of Togo and Benin that consider indigenous soil nutrient-supplying capacity and yield potential. On DS, the main nutrients limiting maize growth were N and probably K. On NDS, nutrients limiting growth were mainly N and P. Even on DS rapid gains in productivity can be obtained, with MSF serving as a means to allow farmers with limited financial means to restore the fertility of such soils. MSF cannot be recommended on relatively fertile NDS.

No abstract