From rice-like plants to plants liking rice: A review of research on weeds and their management in African rice systems

Competition from weeds is the most important yield reducing factor in African rice production systems. Generally important weed management practices in rice are controlled flooding and the use of herbicides. Smallholder rice farmers in Africa however often lack the necessary water management infrastructure, access to affordable, good quality herbicide products and knowledge and equipment for their safe and effective application. Against this challenging backdrop, effective and affordable weed management strategies are highly needed. The literature on weed ecology and management in African rice systems is systematically reviewed to assess achievements in the last quarter of the past 50 years of international rice research endeavours, the period since the last comprehensive review (2009), and to propose the way forward for research and development. Most published studies are from West Africa and focussed on rainfed upland (43% of all relevant studies) or rainfed lowland (32%) rice. Grasses are the most frequently studied weed types, closely followed by parasitic weeds and broadleaved weeds (ex aequo). Most research (75% of published studies) focussed on weed management, mostly referring to or including curative measures (e.g., chemical, manual) or preventive weed management options that improve weed competitiveness of the crop (e.g., crop establishment, cultivars), while less attention was observed for preventive measures aiming at reduced weed recruitment or seed bank sizes (e.g., crop rotations, intercropping, mulches) or integrated weed management approaches. Future research should invest more in developing integrated weed management strategies that achieve (1) reduced weed recruitment, (2) reduced weed seed bank sizes and (3) improved crop competitiveness and that are compatible with farmer's production resources, fairly independent of (agrochemical) industries and markets and benign to the environment and human health. We recommend research on parasitic weeds to focus on a further broadening of the range of currently available management options, with a particular focus on the role of soil fertility and more efficient fertiliser technologies that simultaneously improve crop productivity and quality. For research to contribute to the development of meaningful weed management strategies for African rice systems in the future, we believe it would be best to identify and focus on target-location specific weed-communities, and to reconcile field level weed management strategies with the preconditions set at higher system levels (e.g., farming and agricultural systems) and anticipated scenarios regarding changing demographics and biophysical and institutional environments.