Tải bản đầy đủ - 0trang
VII. Soil Moisture–Herbicide Relationships in Upland Rice
WEEDS A N D WEED MANAGEMENT
depth of butachlor movement and the susceptibility of emerging crop
seedlings. In the Philippines, Sankaran and De Datta (1984) found that a soil
moisture content of 35% or more increased the herbicidal activity of
butachlor, oxyfluorfen, and oxadiazon in upland rice. Nako (1977) observed
that increased soil moisture content after applying thiobencarb decreased
establishment and inhibited rice growth at the seedling stage. In very dry soils
in Bolivia, Tollervey et al. (1980) observed that pendimethalin did not
effectively control R.exaltata.
Rao and Dubey (1977) reported that high soil moisture caused severe
toxicity with dinitramine, piperophos plus dimethametryn, and butachlor;
moderate toxicity with nitrofen; and low toxicity with thiobencarb. Jikihara
and Kimura (1979) reported that thiobencarb 50% plus prometryn 5 % at 8
liter/ha gave excellent weed control in wet soil but not in dry soil.
Sankaran and De Datta (1984) reported that soil moisture status determined the success of chemical weed control in upland rice. During the 1983
dry season, they used a line-source sprinkler system to regulate soil moisture
while evaluating the effectiveness of pendimethalin and oxadiazon. With both
herbicides, IR36 yields were similar to those of the hand weeded check at
high moisture levels (81 1 and 691 mm). When moisture levels dropped below
cumulative pan evaporation (688 mm), yields with chemical and hand
weeding were similar to those in the unweeded check (Fig. 7). Below 525 mm
there was no grain yield, although use of herbicides and hand weeding
effectively controlled weeds.
FIG.7. Effect of herbicides on grain yield of upland rice IR36 at different moisture regimes.
(From Sankaran and De Datta, 1984.)
S. SANKARAN AND S . K. DE DATTA
c. LEAFWATER POTENTIAL OF UPLAND RICE AND WEEDS
Limited soil moisture in upland rice areas with inadequate rainfall lowers
the leaf water potential of rice compared with associated weed species (IRRI,
1979). Sankaran and De Datta (1984) found that mid-day leaf water potential
of upland rice was much lower than that of R.exaltata and C. rotundus (Fig.
8). The leaf water potential of the associated weeds did not fall below - 7 bar,
even under severe stress. The leaf water potential of rice plants went down to
-30 bar. Iwata and Takayanagi (1974a,b) reported that upland weeds are
more adaptable to low soil moisture than rice. The leaf water potential of rice
did not vary significantly at high (811 mm) and intermediate (525 mm) soil
moisture levels, with or without weeding. However, at low soil moisture levels
(214 mm), the leaf water potential of rice in the untreated control was below
that in the pendimethalin-treated and hand weeded check plots.
Water applied (rnrn)
FIG. 8. Changes in the leaf water potential of weeds and upland rice at three moisture
regimes. IRRI, 1983 dry season. (From Sankaran and De Datta, 1984.)
WEEDS A N D WEED MANAGEMENT
VIII. WEED CONTROL METHODS IN UPLAND RICE
De Datta (1980) divided weed control and management into four categories: substitutive, preventive, complementary, and direct. He suggested combining these practices to fit farmer resources and to minimize the possible
buildup of a single noxious weed or group of weeds.
Hand weeding is the most widely used weed control method for upland
rice. Normally, two weedings are done, the first 3 weeks after sowing and the
second 2 weeks later. For upland rice, Indonesian farmers still use only hand
weeding with simple tools (Ronoprawiro, 1975). Schiller and Indhaphun
(1979) wrote that upland rice fields in northern Thailand could produce well
with two manual weedings-one before 20 days and the other at 50 DAS. In
India, Upadhyay and Choudhary (1979) reported that weeding 3 and 6 weeks
after sowing upland rice produced maximum yield. However, hand weeding
within the crop row was still necessary. Hand and hoe weedings after upland
rice sowing currently are the best methods of weed control in Nigeria (IITA,
1972). Several workers in India have confirmed the efficiency of hand weeding
as compared with other weed control methods (Ghosh, 1976; Ghosh et al.,
1977; Dixit and Singh, 1981).
Although hand weeding is effective, it is tedious, time-consuming (Aryeetey, 1970; De Datta and Beachell, 1972), and labor intensive (Muller and
Bilderling, 1953; De Datta and Beachell, 1972). Labor is expensive (Muller
and Bilderling, 1953) and sometimes scarce (Jan, 1973), either because of
short periods of high requirement or because of continual shortages (Jan,
1973). Additionally, by the time weeds can be removed by hand, they have
already competed with the crop and reduced yields (Cates, 1969).
Akhanda (1966) reported that hand weeding 25 DAS required 321 laborhr/ha but resulted in high upland rice yields. Ray (1973) estimated that a
single hand weeding needed 300 to 700 labor-hr/ha. Several hand weedings
are needed for successful upland rice production. Delaying hand weeding
beyond 25 DAS sharply increased labor requirements and significantly
In Nigeria, Curfs (1975) reported that hand and hoe weeding performed
better than other mechanical weed control methods, but that their time
requirements were high. At IITA in Nigeria, in an upland field heavily
infested with weeds, one hand weeding at 14 DAS required 244 labor-hr. A 1week delay increased the labor required for weeding by more than 40 hr
S. SANKARAN AND S. K. DE DATTA
(Moody, 1975). Between 25 and 45 DAS each day of delay in weeding
reduced grain yield by 43 kg/ha and increased weeding by 23 labor-hr/ha (De
Datta, 1979). Three weedings produced the highest grain yield but required
530 to 838 labor-hr/ha. In terms of kilograms of rough rice per labor-hour,
one weeding at 25 DAS was most efficient.
Cultural practices such as land preparation, timely weeding (Moody,
1975), planting appropriate rices, using proper seeding rates, and fertilizer
management may partially substitute for chemical weed control in upland
Plant density and row spacing influence weed incidence. Bhan (1968)
found that narrow (15cm) spacing was superior to wide (30 and 45cm)
spacings in minimizing weed competition and increasing productive tillers
and yield in upland rice. However, Clarete and Mabbayad (1978) found that
15-, 30-, or 45-cm row spacings did not appreciably affect total weed count.
Reduced plant stand and wider row spacing left more area for weed growth
and caused low grain yields. Tosh et al. (1981) showed that row seeding
produced higher plant population, lower weed dry matter accumulation and
nutrient uptake by weeds, and higher grain yield than broadcast sowing
(Table XI). Merlier (1978) observed that maintaining optimum plant density
with a high seeding rate helped suppress vigorous weeds during early rice
Effect of Seeding Methods on Weed Dry Matter Accumulation,
Nutrient Uptake, and Grain Yield'**
Weed dry matter
Nutrient uptake by weeds
60 DAS (kg/ha)
Adapted from Tosh et al. (1981).
In a column, means followed by the same letter are not diifferent at the 5 % level by DMRT.
In Sierra Leone, Jones and Tucker (1978) tested rice straw and rice husk
mulches as weed suppressors and obtained 48 and 32% yield increases. Both
mulches controlled weeds as effectively as hand weeding. Results of IRRI
research showed very low soil moisture tension values (85 cbar and 1.75 bar)
in dry soil mulch plots and highest values (25-30 bar) in weedy fallow plots.
Dry soil mulch conserves moisture for the following upland rice crop. In