Effect of Selected Weed Management Practices on the Growth and Yield Components of Finger Millet (Eleusine coracana) in Western Region of Kenya

Abstract
Finger millet (Elusine coracana) accounts for 8% of the total area and 11% of the millet production worldwide. It is grown on over 4 million ha globally, mainly for food purposes. Millions of people in the dry lands of Central and East Africa, and South of India depend on finger millet as an important source of food to them (CGIAR, 2001). Finger millet is one of the most neglected and underutilized crops. Additionally, the crop has received limited research attention compared to wheat, rice, and maize (FAO, 2011). Therefore, production challenges such as those caused by weeds like goose grass Elusine indica remain at large. Manual weeding is the commonly employed weed control method in finger millet production, but is expensive and labour intensive. The current study was set to evaluate the influence of weed management practices on finger millet growth and yield components. The experiment was laid out in a Randomized Complete Block Design (RCBD) with three replicates. The plots measured 2 by 2m with a border width of 1m. The treatments included Pendimethalin, Dimethyl amine, Metolachlor, Metribuzin, Atrazine (at three rates each 1.0, 1.5 and 2.0 l/ha), No weeding and Hand weeding. Data was collected on the number of weed species, weed biomass, number of basal tillers, herbicide phytotoxicity, weed and crop heights, number of panicles, weight of panicles, weight of un-threshed and threshed grains and 1000 grain weight. Data was analyzed using one-way ANOVA using GenStat version 15.1. Application of Pendimethalin at 1.5 and 2.0 L/ha Active Ingredient (AI), resulted in weed optimal control and least phytotoxicity. Results also indicated that the height of finger millet was significantly (p<0.001) higher where the herbicides were applied. Lower weed biomass was also positively correlated with higher crop height, more panicles, high unthreshed and threshed weights and a 1000 grain weight. Application of 2,4D at rates of 1.5L and 2.0L resulted in significantly taller plants 33.00 cm, than the other weed management methods. Finger millet under Pendimethalin 1.5 L gave the highest number of 86 panicles while Atrazine 2.0L and Pendimethalin1.5L methods of weed control, had significantly higher weight compared to all the other treatments. The 1000 seed mass across the treatments averaged 2.31 g while the on the untreated treatments had an average of 1.54 g. Weed control using pre emergence herbicides significantly (p<0.001) increased the yields of finger millet.