Weed dynamics and management studies in direct seeded rice (Oryza sativa L.)

Abstract

Weeds are a stringent biological constraint to direct seeded rice (DSR) growth and yield. Cultural practices can alter the patterns of crop growth and can modify dynamics of weed-crop competition, and hence need to be investigated as a tool for sustainable weed management in DSR. Studies on weed dynamics and their management in DSR sown at different times under different tillage systems were conducted at the Agronomic Research Area, Department of Agronomy, University of Agriculture, Faisalabad over two consecutive growing seasons during summer 2010 and 2011. In Experiment-I, a commonly grown rice (Oryza sativa L.) cultivar Super Basmati was sown on 15th June and 7th July under zero (ZT) and conventional tillage (CT) and was subjected to different durations of weed competition (10, 20, 30, 40, 50 days after sowing [DAS] and weed competition throughout the growing season). Weed free plots were maintained under each tillage system and sowing time for comparison. Experiment-II was also conducted under same tillage systems and sowing times as in Experiment-I, but here the weeds were managed by using pre- and post-emergence herbicides. Pendimethalin and bispyribac sodium at 1137 and 30 g a.i ha-1 were used as pre- or post-emergence application, respectively either alone or in sequence. A weedy check and weed free treatment were included for comparison. Data on weed dynamics (weed types, their density and dry biomass), and rice growth, grain yield and quality characteristics were recorded following standard procedures. Qualitative and quantitative differences were observed in response to two tillage systems, two sowing times, and six weed competition durations. The results showed Poaceae weeds to be more common under ZT while CT had high relative proportion of broad-leaved weeds in total weed density and dry biomass. The density and dry biomass of Cyperaceae weeds was higher in the late sown (on 7th July) crop than early (15th June) sowing. Shifting sowing time of DSR from mid-June to first week of July also reduced weed density but their dry biomass remained unaffected. Tillage systems had no effect on the total weed dry biomass in DSR. Plots subjected to season long weed competition had mostly Poaceae weeds with no broad-leaved weeds being observed at harvest. In the second year of study, the dominance of Poaceae weeds was increased under both tillage systems and sowing times. Significantly less biomass of Poaceae weeds was observed under CT than ZT in 2010; however in 2011, this effect was not observed. Horse purslane (Trianthema portulacastrum L.) and crowfoot grass [Dactyloctenium aegyptium (L.) Willd.] were the dominant broad-leaved and grassy weeds, respectively. Purple nutsedge (Cyperus rotundus L.) was the dominant sedge weed, especially in the July sown crop. Increasing the duration of weed competition increased the time to heading (from 99 to 102 days in 2010 and 98 to 103 days in 2011) while it curtailed time taken from heading to maturity (from 29 to 27 days in 2010 and 29 to 26 days in 2011). Certain growth attributes of DSR were more severely affected in 2011 than in 2010 possibly due to higher weed density and weed biomass. Root length of the DSR crop was decreased by 32% when sowing was delayed from 15th June to 7th July. The DSR crop sown under CT had more root length as well dry biomass than roots produced ZT. A significant interaction was noticed between tillage, sowing time and weed competition duration for the productive tillers during the first year (2010) but not in the second season (2011). The 1000-kernel weight was the same for weed competition durations of 20 to 50 DAS; however, season long weed competition significantly reduced 1000-kernel weight (18.46 to 17.30 g during 2010 and 16.03 to 13.17 g during 2011). Relative grain yield loss (RYL) ranged from 3-13% and 7-16% when weeds were allowed to compete only for 20 DAS. Under season long weed completion, the RYL ranged from 68 to 77% in 2010 and 74 to 80% in 2011. Kernel length and percentage of normal kernels, taken as measures of grain quality were not affected by tillage or sowing time, however, weed competition duration significantly reduced these quality attributes. In addition, increasing the duration of weed competition reduced nitrogen, phosphorus and potassium uptake by the DSR crop in both grains and straw, whereas weed depleted more quantity of these nutrients as the duration of weed competition was extended. In Experiment-II, the highest density of the Poaceae weeds was observed in weedy check plots of ZT or those sown on 15th June during 2010. A sole application of bispyribac sodium did not reduce the density of Poaceae weeds. Plots sown under ZT on 7th July and treated with pendimethalin recorded significantly higher weed density even greater than weedy check during both years. Under ZT, Cyperaceae weeds were at par with the weedy check in the pendimethalin treatment but higher than the weedy check in the July sown DSR crop treated with the same herbicide. Sequential herbicides applications were better than sole application in averting weed growth. For example, the ZT plots sown on 7th July recorded a 56% reduction in weed dry biomass under a sole application of pendimethalin whereas there was an 83% reduction when pendimethalin was followed by bispyribac sodium. Suppression in weed dry biomass achieved in CT plots sown on 15th June with sequential herbicide application was higher (78%) than that of a sole application of pendimethalin (61%) as compared to weedy check. A sole pendimethalin application increased the time taken from sowing to emergence (0.5 days) by the DSR crop. Early heading (1.5 days) was noticed in plots where sequential applications of pre- and post- emergence herbicides were made and these plots also headed at a similar time to the weed free plots. Sequential herbicide applications improved the rice growth attributes over that observed following the sole herbicide application. A sole application of pendimethalin recorded significantly lower grain yield than its sequential application with bispyribac sodium. The sole and sequential chemical weed control treatments increased rice grain and straw nutrient content. The upper and lower most limits for nutrient uptake were noticed in the weed free and weedy check plots, respectively. Sole or sequential herbicide applications were not significantly different from each other regarding nitrogen uptake by weeds yet these treatments were quite effective in preventing nitrogen removal by weeds as compared to weedy check plots. During both years, higher benefit cost ratio (1.91 and 1.92 during 2010 and 2011, respectively) and maximum marginal rate of return (4352 and 3951% during 2010 and 2011, respectively) was associated with the sole application of bispyribac sodium to ZT plots sown in June. Sequential herbicide application was the second most effective treatment in this regard. Weed competition even for 20 DAS reduced grain yields in DSR. Poaceae weeds were much higher under ZT system and were difficult to control with herbicides. Although ZT was cost-effective yet it can only be practiced where an effective chemical weed management strategy can be adopted. A pre-emergence application of pendimethalin with a following post-emergence application of bispyribac sodium was the most effective way to manage weeds in DSR. An early sowing time of 15th June was effective in minimizing weed proliferation and rectifying the yield penalty associated with a later sowing time of 7th July. Therefore, it is concluded that DSR should preferably be sown on 15th June under CT tillage system and weeds must be controlled before 20 DAS to avoid yield loss. For effective weed control within this time frame, sequential herbicide application program comprising of both pre- (pendimethalin) and post-emergence (bispyribac sodium) herbicides should be followed to sustain productivity of DSR. Successful adoption of direct seeding of rice by farmers in Pakistan will depend on whether farmers can control weeds and prevent shifts in weed population from intractable weeds to more difficult-to-control weeds as a consequence of rice direct seeding.