INTRODUCTION For more than half of the world's population, rice (Oryza sativa L.) is one of the most important staple foods. The rice production area in India as of the year 2019-20 is estimated to be about 43.78 million hectares with a production of 122.4 million tonnes (Anonymous, 2021). Direct seeded rice (DSR) has a higher weed population than transplanted rice. Weed infestation is one of the leading causes of low productivity, with 50-60% output loss due to crop and weed seeds germinating at the same time (Pinjari et al., 2016). Weed control in conservation agriculture (CA) is more difficult than in conventional agriculture because tillage activities do not bury weed seeds. Crop residue when retained in ZT practice, suppress weed growth, maintain soil temperature, and control air pollution caused due to residue burning (Sharma et al., 2012). Tillage, crop establishment method, machinery, agronomic practices etc. play crucial role in weed management in conservation agriculture (Laford et al., 2009). High labour expenses and timely unavailability have increased the use of herbicides in CA. Because a single herbicide cannot control a wide range of weeds, conservation agriculture can benefit from integrated weed management. In a conservation agriculture-based DSR-yellow sarson-greengram cropping system, the current experiment was done to investigate the effects of tillage and weed management measures on weed growth and rice productivity. MATERIAL AND METHODS A long term experiment on DSR-yellow sarson- greengram cropping system was started in2015-16. The present experiment on DSR was conducted during 2018 and 2019 under the same cropping system (4th and 5th year of the system) at Agriculture Farm, Institute of Agriculture, Visva-Bharati, Sriniketan, West Bengal. The texture of the soil in the experimental field was sandy loam. With three replications, the experiment was set up as a strip plot. The horizontal strip was divided into four tillage practises, including conventional tillage (CT) (DSR) — CT (yellow sarson) — CT (greengram), CT (DSR) — zero tillage (ZT) (yellow sarson) — ZT (greengram), ZT (DSR) — ZT (yellow sarson) — ZT (greengram), ZT + residue (R) (DSR) and three weed management practices, viz. recommended herbicides (RH) (pendimethalin at 1.0 kg ha-1 followed by bispyribac-sodium at 25 g ha-1 in DSR, pendimethalin at 0.75 kg ha-1 each in yellow sarson and greengram), Recommended herbicides + hand weeding (HW) at 35 days after sowing (DAS), Unweeded control were assigned to the vertical strip. Rice, yellow sarson, and greengram were grown using crop varieties 'MTU-1010,' 'B-9,' and 'Samrat,' respectively. The zero till ferti-seed drill machine was used to sow direct seeded rice and greengram. The seed rate for direct seeded rice was 50 kg ha-1, 5 kg ha-1 for yellow sarson, and 25 kg ha-1 for greengram. The recommended fertiliser doses of N, P2O5, and K2O were applied: 80:40:40 kg ha-1 in DSR, 80:40:40 kg ha-1 in yellow sarson, and 20:40:40 kg ha-1 in greengram. Herbicides were sprayed using a hand operated knapsack sprayer. All other agronomic recommendations were followed, and plant protection measures were used as needed. Weed counts were recorded by inserting 50 cm × 50 cm quadrats from the specified sampling area of 1.0 m2 in each plot, then drying them in a hot air oven at 70o C to determine weed dry weight. At harvest, rice yield attributes and grain yield were recorded and statistically assessed at a 5% level of significance. RESULTS AND DISCUSSION A. Effect on weeds Digitaria sanguinalis and Echinochloa colona were the most common grassy weeds in DSR during all tillage techniques in 2018 and 2019. Cynodondactylon, Setaria glauca, and Panicum repens were present as major grassy weeds under conservation tillage (ZT+R-ZT+R-ZT+R) and zero tillage (ZT-ZT-ZT). Ludwigia parviflora was the most common broadleaved weed detected under all tillage regimes for both the years. In addition to these Eclipta alba and Cleome viscose were the predominant broadleaved weeds in zero tillage and conservation tillage and in conventional tillage (CT-CT-CT) and conventional fb zero tillage (CT-ZT-ZT), Lindernia crustacea, Alternanthera sessilis and Spilanthes acmella were present as pre-dominant broadleaved weeds. Fimbristylis miliacea and Cyperus iria were the most predominant sedges in DSR under all the tillage practices during 2018 and 2019. Chakraborti et al. (2015 and 2017); Duary et al. (2016) both found similar weed flora in DSR. Conventional tillage recorded lower total density and dry weight of weed than other tillage practices at 60 DAS and it was at par with conventional fb zero tillage (Table 1 and Fig. 1). Conventional tillage (CT-CT-CT) recorded 25.85 and 21.62 % lower total weed dry weight as compared to zero tillage (ZT-ZT-ZT) in 2018 and 2019, respectively. Higher weed density and dry weight was recorded under low soil-disturbance systems as a result of zero tillage is likely to leave considerable amount of the weed seeds on the soil surface, where light encourages seed germination, resulting in more seedling emergence than high soil-disturbance system (Chauhan and Johnson 2009). The recommended herbicide in combination with one-hand weeding was found to be superior to other weed management approaches. Application of pendimethalin at 1.0 kg ha-1 fb bispyribac sodium at 25 g ha-1 at 20 DAS fb one hand weeding at 35 DAS recorded 84.92 and 82.48% lower total weed dry weight at 60 DAS than unweeded control in 2018 and 2019, respectively. The interaction effect was significant, indicating that under conventional tillage (CT-CT-CT and CT-ZT-ZT), the recommended herbicide combined with one hand weeding resulted in the lowest total weed density and dry weight at 60 DAS than conservation tillage (ZT-R+ZT+R-ZT+R and ZT-ZT-ZT) in both the years. In rice, efficient weed control was performed under pendimethalin fb manual weeding and pendimethalin fb bispyribac-Na fb manual weeding at all the stages of observations (Pinjari et al., 2016). Echinochloa sp. and Digitaria sanguinalis were well controlled by the sequential application of pendimethalin and bispyribac sodium, whereas Eragrostis spp. and Leptochloa chinensis were poorly controlled (Brar and Bhullar, 2012) .Pavithra et al. (2021) found that pendimethalin 1.0 kg ha-1 at 3 days after seeding (DAS) and fb bispyribac sodium 30 g ha-1 at 30 DAS were the most effective weed control treatments, with the lowest weed density and biomass of rice. B. Effect on crop During both years, tillage and weed management strategies had a considerable impact on rice grain yield. Among the tillage practises, conventional tillage (CT-CT-CT) produced the highest yield in both the years and it was at par with conventional fb zero tillage (CT-ZT-ZT).In 2018, conventional tillage (CT-CT-CT) yielded 30.90 and 22.67% more grain yield than zero tillage (ZT-ZT-ZT) and conservation tillage (ZT-R+ZT+R-ZT+R). In 2019, conventional tillage (CT-CT-CT) enhanced grain yield by 34.60 and 24.71% above zero tillage (ZT-ZT-ZT) and conservation tillage (ZT+R-ZT+R-ZT+R), respectively. Similar results have been reported by Govindan and Chinnusamy (2014); Paliwal et al. (2017); Surin et al. (2019). In a rice–wheat system, Upasani et al. (2014) found that continuous conventional tillage yielded 39.84 percent more than continuous zero tillage. Among the weed management practices, recommended herbicide combined with one hand weeding at 35 DAS registered highest grain yield over other treatments during both the years. Grain yield was reduced by 67.42 and 70.19 % under unweeded control as compared to recommended herbicide with one hand weeding in 2018 and 2019, respectively. During both years, the interaction effect of tillage and weed control practises on grain yield was substantial, indicating that conventional tillage (CT-CT-CT) combined with prescribed herbicide and one hand weeding produced the highest grain yield and it was at par same treatment imposed under conventional fb zero tillage (CT-ZT-ZT).