Evaluation and comparison of integrated, chemical and mechanical control of broadleaf weeds in seed alfalfaύ.

weeds of alfalfa fields which cause significant seed yield loss annually at the Safi abad agricultural research center of Dezful, Iran. The experimental design was a randomized complete block with strip plot arrangement of treatments. Herbicide treatment was as horizontal factor at five levels which in the initial year of study consisted of: application of bromoxynil at 2 L/ha in the 2 - leaf stage of alfalfa + metribuzin at 0.5 kg/ha after the cutting for seed, application of bentazon at 2 L/ha in the 6 - leaf stage of alfalfa + metribuzin at 0.75 kg/ha after the cutting for seed, application of bentazon at 2 L/ha in the 6 - leaf stage of alfalfa + bentazon at 3 L/ha after the cutting for seed, weed free and weedy checks. Cultivation was as vertical factor at three levels which in the initial year of study consisted of: reciprocal application of shovel cultivator when alfalfa plant had a height of 30 cm and after the cutting for seed, reciprocal application of rolling + shovel cultivator when alfalfa plant had a height of 30 cm and after the cutting for seed and noncultivated plot. In the second and third year of experiment application of herbicide treatments on the same plots after the cutting for seed were as follows: application of metribuzin at 0.5 kg/ha , metribuzin at 0.75 kg/ha, bentazon at 3 L/ha, weed free and weedy checks, and cultivation treatments on the same plots after the cutting for seed were: reciprocal application of shovel cultivator, rolling + shovel cultivator and noncultivated plot. In the first year of the experiment, compared to the weedy check, total number of broadleaved weeds were reduced by the application of bromoxynil + metribuzin, bentazon + metribuzin and bentazon + bentazon by 77, 81 and 56% , respectively, and by the application of each of the shovel and rolling + shovel cultivators reduced by 59% compared to the noncultivated treatment. In the second year of the experiment, compared to the weedy check, total number of broadleaved weeds were reduced by the application of metribuzin at 0.5 and 0.75 kg/ha and bentazon by 83, 95 and 83%, respectively, and in the third year of the experiment were reduced by 68, 87 and 65%, respectively , by the aforementioned herbicide treatments. Although, effect of herbicide treatments on alfalfa seed yield was not significant but compared to the weedy check, alfalfa seed yield was increased by the application of metribuzin at 0.5 and 0.75 kg/ha, bentazon and weed free check by 51, 11, 83 and 37%, respectively. Compared to the weedy check, percent of Cichorium intybus seeds contamination in alfalfa seeds in the first year of the experiment was reduced 82% by the application of bromoxynil, in the second year it was also reduced by the application of metribuzin at 0.5 and 0.75 kg/ha and bentazon by 67, 95 and 71%, respectively, and in the third year it was reduced by 85, 75 and 86%, respectively.