Introduction
Globally, soybean [Glycine max (L.) Mer.] is economically and nutritionally the most important oil seed crop of the world. In Pakistan, it is non-conventionally grown on 5137 hectares producing 2885 tones grains during Anonymous, 1994). This national yield per unit area is well below the potential yield of the recently recommended genotypes under modern soybean production technology (Aslam et al., 1995). Reason for the low yield is attributed to many factors. Out of these poor weed management is the most rroortant one. The extent of losses due to weeds, however, vary with the intensity and kind of the weed. According to Sajjapongse and Roan (1981) yield losses were 63 percent. Chandler et al. (1984) estimated that osses in soybean yield due to the weed infestation ranged from 90 percent in Canada to 17 percent in United States. In our country, traditionally weeds are controlled by hand .weeding and very few farmers control weeds by mechanical or chemical means. Sadiq and Habib (1988) revealed that chemical weed control was more economical and efficient as compared to mechanical and hand weeding in rainfed area due to frequent and extensive rains during early growth stages. Balyan (1992) revealed that soybean yield ependimethalin and weed free plots were at par. No doubt the chemical method of weed control can be adopted but it requires technical awareness and more initial investment. For the farmers having low capital investment ability and low technological skill, the use of herbicides becomes a limiting factor. Keeping such points in view this study was conducted to find out an efficient and economically cheaper herbicide application method for effective weed control in soybean.
Materials and Methods
The experiment was conducted at National Agricultural Research Centre (NARC) Islamabad during spring, 1992 and 1993 (Anonymous, 1994). Soybean variety NARC-ll was planted in rows 30 cm apart with single hand drill on a plot size of 2.4 x 5 m. Recommended doses of fertilizer (25:50 NP kg ha1) were applied at the time of sowing. Seed was inoculated at the time of sowing. Experiment was laid out in randomized split plot design with six treatments replicated three times.
Four pre-emergence herbicides pendimethalin at 1.48 kg ai/ha, oxadiazon at 0.45 kg ai/ha, trifluralin at 0.75 kg ai/ha and isoproturon at 1.0 kg ai/ha were applied with the following three different methods.
Spray: A hand operated nap-sack sprayer with flat fan nozzle was used.
Sand mix broadcast: Sand mix broadcast was carried out by mixing the herbicides with 12.5 kg/ha dry sand after sowing of the crop.
Soil incorporated application: The herbicides were applied before sowing of trial and were mixed in the soil.
The weed density and weed biomass was recorded using a quadrate of 25 x 25 cm after 45 days of sowing. Plant height (cm) and number of pods per plant, yield (kg ha1) and 100 seed weight were recorded at harvest. Data over years were consistent, therefore the combined analysis of variance and least significant difference at 5 % probability level were used to test the significant means using MSTATC computer software.
Results and Discussion
Weeds flora present in soybean during spring season experiment (Table 1) consisted of 12 weed species belonging to 10 families, these weeds were grassy, broad leaved and sedges. All the four herbicides had no negative effect a the germination of the crop. Though these herbicides controlled most of the annual grasses and broad leaved weeds but after 60 days of sowing, weeds such as Sorghum halepense, Cynodon dactylon and Cyperus rotundus appeared.
Table 1: | Weeds flora recorded in soybean crop in spring |
 |
Table 2: | Weed density/weed biomass, plant height and number of pods per plant as affected by herbicides in soybean |
 |
Means in a column followed by same letter are not significantly different at 5% level of significance |
Table 3: | Application methods and its effect on weed density/biomass and different characters in soybean |
 |
Means in a column followed by same letter are not significantly different at 5 % level of significance |
The effect of weed control treatments on weed density and their dry matter accumulation as presented in the Table 2 showed that average values were significantly different from each other. Among the herbicide treatments, lowest weed density and bio-mass were recorded in pendimethalin which reveals the effectiveness of this herbicide followed by oxadiazon. The highest plant height was recorded in weed free treatment followed by pendimethalin and the lowest in weedy check (Table 2). Efficacy of herbicidal treatments on the development of pods per plants was statistically similar but significantly different from weed free and weedy check.
Effectiveness of weed control on the grain yield was evaluated through the ranked means of the treatments, Grain yield recorded in weed free and pendimethalin plots was statistically similar which means that weed control efficiency of pendimethalin was at par with weed free treatment. Our results are in accordance with the results of Balyan (1992). Grain yield of weed free treatment gave highest yield while the lowest yield was produced by weedy check. Losses in yield were more than 35 percent. Our results are in accordance with the result of Sajjapongse and Roan (1981) and Chandler et al. (1984).
All the herbicide application method gave better results. Data presented in Table 3 reveals that weed density and biomass as affected by application method were apparently different but statistically similar. Significantly higher plant. height was noted in spray method than other two methods. However the higher number of pods/plant were recorded in-soil incorporation method. This might be due to higher mortality rate of weeds and thus crop utilized more nutrients while the spray and sand mix broad cast method remained at par. The highest yield was obtained in soil incorporation method showing the higher efficacy of this application method for weed control.