Legume-maize rotation and nitrogen effects on weed performance in the humid and subhumid tropics of West Africa

A long-term trial was established in 1998. The objectives among others were to assess the impact of nitrogen, natural bush fallow, and legume-maize rotations on weed growth and species composition in the forest/savanna transition zone (Ibadan) and northern Guinea savanna (Zaria) in Nigeria. At both locations, the experiment was arranged as a split-plot design. The main plots were green manure legumes (Pueraria phaseoloides [Roxb.] Benth in Ibadan and Aeschynomene histrix Poir. in Zaria), a forage legume (Stylosanthes guianensis [Aubl.] Sw.), grain legumes (cowpea [Vigna unguicidata {L.} Walp 'IT-90K-284-2' or soybean [Glycine max {L.} Merr 'TGx 1864 and TGx 1485-1 D'), natural bush fallow, and continuous cropping with maize [Zea mays L.]/cassava [Manihot esculenta Crantz] '82/00058' at Ibadan and maize at Zaria. Grain legumes were either double-cropped in one season by growing short-duration soybean with cowpea varieties sequentially, or by growing long-duration cultivars within one growing season. Nitrogen levels (0 and 30 kg N ha(-1)) were the subplot treatments. Weed data were collected in 2000 and 2003 after 2 years of continuous cropping or rotation with natural fallow or legumes. Weed density was significantly higher at Zaria (177 plants m(-2)) than at Ibadan (149plants m(-2)). Weed dry matter was higher at Ibadan than at Zaria. Weed density and dry matter were lower in the green manure and forage legume treatments in both years at Ibadan. At Zaria, the forage legume treatment had the lowest weed density and dry matter in 2003 only. In 2003 at Ibadan, per capita weed population growth rate (WPGR) decreased in the green manure and forage legume treatments. There was an increase in WPGR in continuous cropping, double-cropping with cowpea and soybean, and natural bush fallow treatments. At Zaria, WPGR was not affected by fallow type or nitrogen (P>0.05). Redundancy analysis showed significant differences in species composition among the various treatments. Nitrogen level did not affect density, dry matter, or the composition of weeds. (C) 2007 Published by Elsevier Ltd.