S-Space College of Agriculture and Life Sciences (농업생명과학대학) Dept. of Plant Science (식물생산과학부) Theses (Ph.D. / Sc.D._식물생산과학부)
Establishment of Weed Management System Based on Soybean-Weed Competition Model for Soybean Production
- 농업생명과학대학 식물생산과학부
- Issue Date
- 서울대학교 대학원
- 학위논문 (박사)-- 서울대학교 대학원 : 식물생산과학부, 2016. 2. 김도순.
- Primorsky-krai, located in far-eastern regions of Russia, is a core agricultural area for Koreas overseas farming operations to secure food crop supply. In the region, about 10 Korean farming firms have been operated for soybean production in a large scale farming system. Average soybean yield in this region was as low as 1.05 ton ha-1, much lower than 1.46 ton ha-1 of average soybean yield in Russia (Russian Statistical Yearbook, 2011). The main reason of such low soybean yield may be due to lack of adequate weed management system. Therefore, this study was conducted to investigate soybean-weed competition to determine economic thresholds for weed control under different nitrogen fertilizer levels, and to establish weed management system based on soybean-weed competition model and sequential application of PRE and POST herbicides. The competition effects of single weeds, Ambrosia artemisiifolia, Sonchus oleraceus, Chenopodium album, Echinochloa crus-galli, and Beckmannia syzigachne, on soybean yields were modeled using the rectangular hyperbolic model, resulting in weed compeititivities of 0.134, 0.083, 0.064, 0.151 and 0.076, respectively. The competition effects of multiple weed interferences on soybean yield were modeled using the multivariate rectangular hyperbolic model with density equivalents estimated from soybean and single weed competition experiments. Under different levels of nitrogen fertilizer, the rectangular hyperbolic and the multivariate rectangular hyperbolic models showed good description of soybean yields as influenced by single weed and multiple weed interferences, respectively. In the rectangular hyperbolic model, two parameters Y0 and β responded to nitrogen fertilizer level in the inverse quadratic function, suggesting that inverse quadratic functions for the two parameter can be incorporated to the rectangular hyperbolic model. As a result, the combined model well described both nitrogen fertilizer and weed competition effects on soybean yield under single and multiple weed interferences. Pre-emergence (PRE) and post-emergence (POST) herbicides tested alone or in a sequential combination showed that the sequential application of acetochlor followed by bentazon+acifluorfen or bentazon+imazamox controlled weeds most effectively and minimized soybean yield losses from weed competition. The acetochlor-based sequential application with bentazon mixtures showed greater soybean yield, resulting in better gross profit than the common practice of herbicide use in the region. Economic thresholds for weed management in soybean with acetochlor-based sequential herbicide application range from 0.39 to 0.75 density equivalents m-2 in soybean cultivation under multiple weed interference. In conclusion, the models developed for soybean-weed competition under different nitrogen fertilizer levels in either single or multiple weed interference are useful for predicting soybean yield as influenced by weed competition and nitrogen fertilizer and decision-making for weed management in this region. The herbicide-based weed management system based on sequential application of PRE and POST herbicides can provide effective and economic weed management for soybean cultivation in Primorsky-krai, Russia.