Weed Diversity Under Different Water Regimes in Irrigated Direct-Seeded Rice Cultivation

This study comprised two sets of experiment. The field trials were conducted at MARDI Bertam Rice Research Station in Seberang Perai, Penang, while the experiments on weed seed viability were done in a glass house and the Weed Biology Laboratory at Universiti Putra Malaysia. This study began in February 2004 and ended in June 2006. All field trials and weed seed viability studies were done in Off Season 2004 and Main Season 2004/2005. The objectives of the study were: a) to investigate the population and status of dominant weeds and their effect on rice yield under different water regime treatments in field condition; b) to determine the total weed seed reserve, species composition of the entire weed seed bank in the soils and their viability after being treated under different water regime treatments in the rice field. Five water regime treatments were used namely; T1 = continuous flooded condition until maturity, T2 = early flooding until 55 DAS (day after sowing) followed by saturated condition until maturity, T3 = early flooding until 30 DAS followed by saturated condition until maturity, T4 = continuous saturated condition until maturity, T5 = continuous field capacity condition throughout the experiment period. Results from field trials indicated that in unweeded plots (no weed control treatment involved), T1 and T2 showed the lowest number of weed m-2 and weed biomass m-2 compared to the other water regime treatments. Plots treated with T1 and T2 significantly suppressed weed population to approximately 18 – 58% and reduced weed biomass to 14 – 57% as compared to the highest in T5 at all sampling dates (30, 60 and 90 DAS) in both Off Season 2004 and Main Season 2004/2005. For weed composition across the water regime treatments, 11 weed species were recorded in Off Season 2004 and 10 weed species in Main Season 2004/2005. Broadleaved weeds mostly Monochoria vaginalis and Limnocharis flava were the most dominant weeds in most of the water regime treatments. The SDR values of broadleaved weeds were 48.7% (T2) > 46.4% (T1) > 44.2% (T3) > 40.7% (T5) > 35.8% (T4) in Off Season 2004. In Main Season 2004/2005, the SDR values for broadleaved weeds increased to 79.5% (T2) > 68.2% (T1) > 62.4% (T3) > 62.2% (T4) > 50.57% (T5). Sedges mainly Fimbristylis miliacea and Cyperus iria were dominant in Off Season 2004 with SDR value more than 34% in all water regime treatments, but decreased to less than 23% in Main Season 2004/2005. For grasses, mostly Echinochloa crus-galli, Echinochloa colona and Leptochloa chinensis, SDR value more than 20% were recorded in T4 and T5 in Off Season 2004 while in Main Season 2004/2005, SDR value between 21 – 34% were observed in T1, T3 and T5.Meanwhile rice plants treated with T1 significantly produced the highest rice grain yield, followed by T2 and T3, while T4 and T5 were the lowest in weeded plots in both planting seasons. However, rice yield in unweeded plots were not significantly different in all water regime treatments. The rice productions in weeded plots were in the range of 22 – 60% higher in Off Season 2004 and 56 – 78% higher in Main Season 2004/2005 than in unweeded plots. Weed control treatment evidently showed stronger influence than water regime treatments. The percentages of reduction in weeded plots from the highest rice yield in T1 to lowest in T5 were 40% in Off Season 2004 and 44.4% in Main Season 2004/2005. Meanwhile the reduction of rice yield in unweeded plots compared to weeded plots was much higher, which recorded the highest of 59.1% (in T1) in Off Season 2004 and 78.3% (in T4) in Main Season 2004/2005. The effect of different water regime treatments on viability of weed seeds in rice field was evaluated. Total weed seed reserves in soils taken from Off Season 2004 rice plots were dominated by broadleaved weeds (90625 seeds m-2) mainly Hedyotis corymbosa, Monochoria vaginalis and Ludwigia hyssopifolia, followed by sedges (34257 seeds m-2), mostly F. miliacea and Cyperus iria, while grasses (20647 seeds m-2), predominantly Leptochloa chinensis recorded the lowest number in all water regime treatments. In soils taken from Main Season 2004/2005 rice plots, sedges (53041 seeds m-2) mainly Fimbristylis miliacea along with broadleaved weeds (54624 seeds m-2), predominantly Monochoria vaginalis and Ceratopteris pteridoides dominated in most of the water regime treatments, while grasses, mainly Leptochloa chinensis and Panicum repens again recorded the lowest number (24935 seeds m-2). 10 weed species, which were not observed in the field trials, were recorded from the same soil in the weed seed bank experiment.Differences in water regime treatments did not significantly reduce the viability of weeds seeds in the soil in Off Season 2004. However, a small reduction in seed viability (approximately 8%) was observed in Main Season 2004/2005. As for the similarity between weed seed bank composition and aboveground weed flora composition, a moderate value (62.07 - 73.33%) was observed in Off Season 2004, while in Main season 2004/2005, the value was slightly lower (51.65 – 64.29%).

Preview PDF FP_2008_7.pdf Download (463kB)

Actions (login required)

View Item