Preventive weed management in direct-seeded rice: targeting the weed seedbank.
In Asia, direct-seeded rice (DSR) is becoming popular as an alternative to puddled transplanted rice (PTR) due to its potential to save scarce resources (labor, water, and energy), reduce greenhouse gas emissions, improve soil physical properties, and increase yields in rotational crops. However, weed management in DSR is more difficult because the initial size differential between weeds and rice is small, reducing crop competitiveness and limiting opportunities for selective control measures including flooding. In this context, preventive approaches - those which focus primarily on limiting dispersal and persistence of weed propagules - may play a critical role in complementing the current reliance on curative tactics such as herbicides. Greater understanding and integration of preventive approaches in DSR may reduce the risks of herbicide resistance development, limit adverse effects of herbicides on human health and the environment, and lower the overall weed management costs. However, information on preventive weed management in DSR is relatively limited. Therefore, the central objectives of this review are to: (i) summarize existing knowledge regarding preventive strategies; (ii) discuss key integrated preventive weed management approaches that have the greatest potential for practical application in DSR systems; and (iii) identify knowledge gaps that limit our ability to optimize preventive approaches. Based on an extensive review of existing literature, we conclude that (i) Minimizing weed seed production in the field is critically important for managing weed seedbanks in DSR, but that given seed dispersal in both time and space, prevention of seed production from neighboring bunds, rice-fallow land and irrigation channels bordering DSR areas may be equally important; (ii) Minimizing dispersal of weed seeds into DSR fields may be a practical approach for species that are dispersed primarily by humans (e.g., as contaminants in crop seeds or through irrigation canals), but not for species that are dispersed primarily by other means (wind and birds); (iii) Promotion of seed predation may be a useful strategy in managing certain weed species in DSR - especially where zero-tillage is used - but more research is needed on the identity of seed predators and management factors that promote their activity; (iv) available evidence suggests that the potential for promotion of seed decay is limited in scope but may be valuable for the management of certain relatively nonpersistent weeds in some cropping systems; (v) strategies that stimulate fatal germination of weed seeds (e.g., stale seedbed) appear to be one of the most promising means of prevention in DSR, but increased information on the mechanisms and timing of dormancy release for key species is needed to optimize and enhance the value of this approach; (vi) Prevention of weed germination and emergence in DSR through mulching - especially in zero-till systems - has proven benefits, but its widespread applicability is limited by the economic tradeoffs associated with using mulch as a source of livestock feed; and (vii) development of anaerobic germination (AG)-tolerant rice cultivars and complementary flooding strategies which can tolerate anaerobic conditions/flooding hold great potential for the suppression of weeds in DSR. Successful integration of preventive approaches for managing weeds in DSR will depend on the development of multidisciplinary approaches which are biologically effective, economically feasible, and socially acceptable. Preventive weed control measures alone are unlikely to be sufficient for the effective and economical management of weeds in DSR systems, but their integration with curative approaches should reduce weed management costs and increase both the likelihood of adoption of DSR and the realization of its benefits for food security.