Integrated weed management in dry-seeded rice using stale seedbeds and post sowing herbicides

Abstract

Dry-seeded rice (DSR) grown with alternate wetting and drying water management (AWD) has recently been introduced in northwest India as an alternative to conventional puddled hand-transplanted rice which is labour, water and energy intensive. The aerobic seedbed of DSR can be extremely susceptible to invasion by diverse weed flora, and if weeds are not controlled effectively, yield losses can be very high. This study was undertaken to investigate the impacts of stale seedbed techniques on the soil weed seedbank and weed infestation in DSR, and to determine the influence of integration of the stale seedbed methods with post sowing herbicides on weed control and rice grain yield. The study, conducted in 2014 and 2015, comprised three seedbed treatments in main plots: without stale seedbed-conventional method, stale seedbed with glyphosate 1 kg ha⁻¹ and stale seedbed with shallow (5 cm) tillage, and four post sowing herbicide treatments in sub plots: unsprayed check, pendimethalin 0.75 kg ha⁻¹ (pre-emergence), bispyribac-sodium 0.025 kg ha⁻¹ (post-emergence) and pendimethalin followed by bispyribac-sodium. The two stale seedbed treatments included one additional irrigation prior to sowing which increased weed seedling emergence prior to sowing by 1.9–2.2-fold; weeds in the stale seedbed treatments were then killed with the application of glyphosate or shallow tillage. At sowing, both stale seedbed treatments significantly decreased the viable seedbank of Echinochloa colona and Dactyloctenium aegyptium to 25–30% of that without a stale seedbed. After rice harvest, both stale seedbed treatments had a significantly lower seedbank than without a stale seedbed, by 13–33%; the stale seedbed with tillage had significantly lower seedbank at harvest than the stale seedbed with glyphosate in the second year. The sequential application of pendimethalin and bispyribac resulted in a significantly lower seedbank of both these grass weed species at harvest. At 20 DAS, both stale seedbed methods had 22–51% lower density of Cyperus rotundus and 42–67% less grass weeds than rice sown without a stale seedbed. There was more than a 2-fold increase in C. rotundus density from 2014 to 2015 without a stale seedbed and with the stale seedbed with glyphosate, and a 1.6-fold increase in the stale seedbed with tillage. In the absence of post sowing herbicides, the stale seedbed with tillage increased grain yield from 0.7–1.0 t ha⁻¹ to 2.1–2.5 t ha⁻¹, while the stale seedbed with glyphosate only increased grain yield in 2015. The combination of the stale seedbed with tillage, pendimethalin and bispyribac had the highest rice grain yield (7.3 t ha⁻¹) and the highest economic returns ($ 1310 ha⁻¹); the returns in this treatment were $ 260 ha⁻¹ higher than using the same herbicides used without a stale seedbed. The results indicate that integrated use of a stale seedbed with shallow tillage followed by the sequential application of post sowing herbicides has potential to control the complex weed flora in dry-seeded rice. The reasons for greater consistency in weed control with the stale seedbed with tillage than glyphosate are unclear and need further investigation in dry seeded rice, as do the long term effects of use of stale seedbeds.