Weed community responses to rotations with cover crops in maize-based conservation agriculture systems of Zimbabwe

Weed management is a challenge to resource constrained smallholder agricultural production in sub Saharan Africa due to insufficient and inadequate management strategies. When weeds are controlled with herbicides or through rotations under conservation agriculture, shifts in weed spectra are expected, increasing the need to adapt to this new situation. Experiments were conducted at four experimental sites namely the University of Zimbabwe farm (UZ) (clay soil), Domboshava Training Centre (DTC) (sand soil) and two contrasting soil types at Henderson Research Station (HRS sand (s), HRS clay (c)) to investigate the responses of weed communities to crop rotations. The trial was carried out from the 2008-09 to the 2013-14 cropping season. Rotations consisted of maize (Zea mays L) rotating with a range of green manures, and the control treatment was maize monocropping. Herbicides were only applied in the maize phase at seeding, supplemented by hand weeding whenever weeds were 10 cm tall or 10 cm in diameter for weeds with a stoloniferous growth habit. Weed count data was collected between 2011 and 2014. Weed density, the Shannon-Weiner index and its components were used to explain weed community responses to rotations. There was a decrease in weed densities over time at all sites with a percentage decrease as high as 92% (i.e. from 357 to 30 weeds m(-2)) observed in maize-velvet bean (Mucuna pruriens (L) DC) rotation. At all sites, some maize green manure rotations were associated with high weed densities and these included maize-black sunnhemp (Crotalaria juncea L.) and maize-cowpea (Vigna unguiculata Walp) rotations. The two weeds Galinsoga parviflora Cav. and Ricardia scabra L. remained abundant throughout the study at all sites with densities reaching above 1000 plants m(-2) per season in some plots. Shannon's E' index was highest at HRSs and HRSc sites in the maize-velvet bean rotation and maize-common rattle pod (Crotalaria grahamiana Wight & Arn .) rotations respectively suggesting that in these treatments dominant weeds were reduced in numbers. This suggests that rotations with cover crops such as velvet bean may reduce weed numbers and dominance of problematic weeds over time. This can potentially lead to a less intense weeding schedule, which is more cost effective and affordable for smallholder farmers. (C) 2014 Elsevier Ltd. All rights reserved.