Evaluation of pre-emergence herbicides for weed management and rice yield in direct-seeded rice in Cambodian lowland ecosystems

Poor weed management in direct-seeded rice (DSR) at crop establishment stage has led to the need to explore different pre-emergence herbicides to reduce weed biomass and increase DSR yield in Cambodian lowland ecosystems. This study evaluated weed and yield responses to pre-emergence herbicides (pendimethalin, pretilachlor, butachlor, oxadiazon and no application as the control) under different weed management options during the 2018 and 2019 early wet seasons in farmers’ rice fields. All pre-emergence herbicides reduced weed flora and pendimethalin was most effective in controlling the population of mainly grassy weeds. Pendimethalin significantly reduced weed dry biomass by 36.2% (21.3 ​g ​m−2) in 2018 compared with the untreated control. In 2019, weed dry biomass was significantly reduced by 46% (17.9 ​g ​m−2) with butachlor compared with the untreated plot. Pendimethalin increased toxicity in rice plants, which decreased SPAD (leaf greenness) by 44.8% and tiller numbers by 39.3% at 7 days after herbicide application (DAPH) compared with butachlor, but plants recovered from pendimethalin toxicity at 14 DAPH. In 2018, oxadiazon herbicide significantly increased grain yield by 61% (4.6 ​t ​ha −1) compared with the untreated control, but oxadiazon produced similar grain yields as pendimethalin. In 2019, there was no significant effect of pre-emergence herbicides on yield but butachlor produced the highest rice yield (3.2 ​t ​ha−1) followed by pendimethalin (2.8 ​t ​ha−1) and oxadiazon (2.7 ​t ​ha−1). Our study demonstrated that pre-emergent herbicides function differently under varying soil-water conditions in 2018 and 2019, therefore rice growers can use butachlor to control weeds in dryer soil conditions and potentially use pendimethalin to control weeds in good soil moisture conditions. This information is crucial and safe in developing appropriate strategies to manage weeds and enhancing DSR productivity in lowland ecosystems.