Tải bản đầy đủ - 0trang
III. Biomass and Nitrogen Accumulation in Green Manures
YADVINDER SINGH E T A L .
waterlogged conditions and produced significantly less biomass than Sesbania and sunn hemp (Morris et al., 1986a). In California, Williams et al.
(1957) reported 0.9-1.4 t/ha dry matter yield of purple vetch. Dry matter
yield of subterranean clover ranged from 3.4-6.0 tlha (Dabney et al., 1989;
Hoyt and Hargrove, 1986).
Rainfall and temperature influence the biomass production of green leaf
manures. Gliricidia sepium shrub raised on the field boundaries produces
about 1-3 t dry leaf material per month from 10,000trees/ha. In India, G.
maculata was reported to produce about 14 kg green leaf per shrub in the
month of July, which could be used for green manuring the main season
rice crop. Sesbania speciosa planted during the month of August on the
boundaries of a hectare field (1 150 m) produced, within four months, an
average of about 42 t/ha of green matter for incorporation in the rice crop
transplanted in January. The seedlings planted in February on the boundaries during the second season gave about 4.0 t/ha green matter for the
main season rice crop transplanted in July (Vachhani and Murty, 1964).
In Sri Lanka, Weerakoon and Gunasekera (1985) showed that about
2.5 tlha leaf dry matter of Leucaena leucocephala could be obtained every
cropping season. S. sesban used as GLM gave total N yields of 32,46, and
47 kg Nlha in the first, second, and third year, respectively, of growth on
sodic soils in northern India, (Rao et al., 1989). In Thailand, S. sesban
produced leaf yields of 542-683 g/plant in wet season and 167-377 g/plant
in dry season, and was more productive than S.formosa and S . grandijiora
(Arunin et al., 1988).
Mukherjee and Agarwal (1950) and Ghai et al. (1985) reported that N
content of different green manures (8 weeks old) ranged from 1.5 to 4.85%.
Roger and Watanabe (1986)reported that N content in legumes varies from
0.2 to 0.6% (fresh weight basis). Ghai et al. (1985) and Hernandez et al.
(1957) observed that N content in Sesbania, sunn hemp, and T. candida
tops was maximum at 45 days of growth and decreased thereafter. In milk
vetch, N content before flowering was 4.5%, decreasing gradually to 3.2%
at the full-broom stage (Ishikawa, 1988). Morris et al. (1986a) observed
that different green manures showed a linear relationship between N
accumulation and dry weight irrespective of green manure species, and
were affected by the age of green manure crop. It was found that green
manures had maximum N content of 2.54% at 45 days and decreased to
1.88% in 60-day-old green manures.
N accumulation in the tops of several leguminous green manure crops is
shown in Table I. The values in excess of 100 kg N/ha are common for
45-50-day-old legumes. Hernandez et al. (1957) found that N yield of
45-day-old green manures ranged from 56-226 kg N/ha. Milk vetch and S.
cannabina in China fixed about 100-350 kg N/ha at full blooming. In the
GREEN MANURING IN WETLAND RICE
United States, hairy vetch, crimson clover, subterrarean clover, and common vetch accumulated 56-209 kg N/ha, but in most cases it was between
100-150 kg N/ha (Smith et al., 1987). In Hawaii, Evans and Rotar (1987)
reported that high yielding accessions of annual Sesbunia produced 817 t/ha dry matter containing 150-245 kg N/ha when sown at 125,000
plantslha and harvested at 98 days after sowing. The highest yielding
varieties were related to S . cannuhinu, which are grown as green manures
in the Asian lowland rice system.
Like dry matter production, N accumulation is related to the age of the
green manure crop. Ishikawa (1963) reported that N yield of milk vetch
was 1.8 kgiha at the start of flowering, and increased to 156 kg Nlha during
flowering (14 days later). N accumulation of S. aculeata as related to its
age is shown in Fig. Ib. Palaniappan et ul. (1990) reported that at 45 days,
S. aculeata and S. rostrata in south India accumulated 185 and
219 kg N/ha, respectively. Chapman and Myers (1987) reported that soybean green manure at early flowering stage ( I 10 days old) fixed 124-167 kg
N/ha in different years. Evans et af. (1989) indicated that dry matter yield
varied from 2.0- 14.3 t/ha and
of narrow leaf lupin (Lupinus ~~iigustifalius)
total N in the shoot ranged from 45-267 kg N/ha. Mahler and Auld (1989)
studied the green manuring potential of Austrian winter peas (Pisurn
satiuurn aruense spp.) and reported mean biomass and N yield of 8.3 tlha
and 167 kg N/ha, respectively.
The contribution of roots to total N yield is generally small in most
leguminous green manures. In S. uculeafu,root dry matter averaged about
1.O t/ha, adding about only 10 kg N/ha at 50-60 days of growth (Morris et
ul., 1986a; Beri et al., 1989a; Meelu et ul., 1990). The root-to-top ratio of
milk vetch is about 0.05 and will supply only a negligible quantity for rice
production (Ishikawa, 1988). Westcott and Mikkelson (1988) reported that
sweet and crimson clovers have 24-26% of the entire crop biomass in the
roots, which contain 2.0-2.3% N; the tops contain 2.4-2.9% N. Vetch has
17-19% of the total biomass in the roots.
Using ''N isotope dilution technique, Chapman and Myers (1987) estimated that 60-72% of total plant N was from biological N2 fixation when
the legumes were grown after 12 months of fallow, and 93-95% when
grown immediately following dry season crop. Smith et al. (1987) concluded that N2 fixation values for legume cover crops ranged from
6 7 4 4 % . Using difference method, Meelu et al. (1990) observed that 80%
of the total N in the tops of 52-day-old S . uculeata was from biological N2
fixation. Similar estimates of N2 fixation have been made for S . rostrata
and S. cannabina by Pareek (1989) using different methods of evaluation.
In the past, only limited research has been directed toward enhancing
biological N2 fixation in the leguminous green manures. The important
YADVINDER SINGH ETAL.
approaches in this direction may involve: selection of superior plant and
rhizobium genotypes, including the consideration of plant-microbe interaction; refinement of inoculation technology; increased understanding of
rhizobium ecology; and improved management practices (Smith and
Knight, 1984). Alikhan et a f . (1983) developed a SSI selection of sunn
hemp which gave 37% increase in green matter yield and 70% increase in N
yield at 50 days of growth.
Application of inorganic fertilizers (N and P) and organic matter has
been reported to stimulate nodulation and Nz fixation by legume crops
(Gibson ef a f . , 1982). Phosphorus, which is required for efficient N2 fixation, is often a limiting nutrient in the tropical lowland soils. Results from
several studies have shown that P application increased biomass and N
accumulation of green manures (Table 11). The green manures generally
responded more to P application on soils low in available P and pH.
Venkatachalam et al. (1969) found that uptake of 32P by rice was greater
from P applied to the green manure crop than its direct application made in
four soil types. In China (Liu, 1988) and Japan (Ishikawa, 1988), application of P has been recommended for obtaining high biomass and N yield of
milk vetch. Beri and Meelu (1980) found that on a soil testing low in
available P, application of 13 kg P/ha increased biomass production and N
accumulation of S. aculeata green manure and gave better yield of rice
than P applied directly to rice. Many other research workers (Sanyasi
Raju, 1952; Sen and Rao, 1953; Desai et al., 1957; Singh and Verma, 1969;
Chen, 1988) have reported similar results.
On soils testing high in available P, application of P fertilizer did not
show any beneficial effect on biomass yield and N accumulation of green
manures (Relwani and Ganguly, 1959; Desai et a f . , 1957). Gu and Wen
(1981)reported that if available P content is below 15 mg/kg in the acid and
neutral soils or below 10 mg/kg in the calcareous soils, P application to the
green manures would be markedly efficient. The results of 311 experiments showed that application of 11-167 kg P/ha gave an average response
(1988) reported that
of 347 kg fresh biomass and 1.21 kg N/kg P ~ O SChen
while rice absorbed 66.2% of P applied to the green manure crop, it
absorbed only 14.7%of the P added directly just before transplanting.
The good effects of K fertilizer on the yield of green manure has been
found in some soils, and K can give effects similar to P on such soils (Chen,
1986). In China, Liu (1988) recommended application of K2S04/KCI at
75-105 kg/ha to milk vetch.
Effect of P Application on Accumulation of Biomass and N of Green Manures
S . aculeara
1. Singh el nl. (1968); 2, Beri and Meelu (1981); 3. Sharma and Mittra (1988); 4, Herrera e r a l . (1989); 5, Singh (1990).
YADVINDER SINGH ETAL.
Rational application of N fertilizer can also increase the N2 fixation rate
of a green manure crop. At low level of N, a starter dose of N was found to
promote nodulation and N2 fixation by legume crops (Gibson et al., 1982).
Gu and Wen (1981) reported that 1 kg of fertilizer N could increase
1.7 0.9 kg N in the green manures. Chapman and Myers (1987) found
that application of a starter N dose (25 kg/ha as urea) increased 10 and 30%
N in the tops of Sesbania and soybean at the flowering stage in the first
year of study. Sharma and Mittra (1988) observed that application of 15 kg
N/ha as urea increased N accumulation of sunn hemp and Sesbania by 23
and 30 kg/ha, respectively. Application of 25.5 kg N/haat the stem elongation stage gave a three- to 4-fold increase in fresh biomass of milk vetch
over its application at the seedling stage. The plant recovery of fertilizer N
applied at the seedling stage was 32%, and that applied at the stem elongation stage was about 78% (Gu and Wen, 1981).
Soil organic matter can affect growth and survival of rhizobia in soil.
Application of 7.5 t/ha of farmyard and poultry manure enhanced root
nodulation of soybean and increased N2 fixation by 209% and 149%,
respectively, over unamended treatments (Dev and Tilak, 1986). In
Thailand, Herrera et al. (1989) observed that application of small rates of
farmyard manure (3 t/ha) was slightly more advantageous when applied to
S. rostrata rather than to rice.
Inoculation enhances the onset and number of effective nodules and N2
fixation by legumes. Ishikawa (1988) observed that inoculation of milk
vetch seed with rhizobia before sowing increased green matter production
by more than 3 times than without inoculation during the first year. In
subsequent years the increase was about 48%. Chu (1954) reported that
inoculation of soybean seed used for green manuring can help increase rice
yields by 20.7% over no inoculation.
Ladha et al. (1989) reported that both stem and soil + seed + stem
inoculation methods produced significantly more nodules (stem and roots)
and biomass of stem-nodulating legumes ( S . rostrata) than did the control.
The plants that were not inoculated on the stem did not develop stem
nodules. Alazard and Duhoux (1987) reported that plant dry weight of
9-week-old A . afraspera was 8.0 g/plant when only roots were inoculated
but it increased to 46 g/plant when both stem and roots were inoculated.
The corresponding values for N accumulation were 158 and 366 mg/plant.
Arunin et al. (1988) reported that inoculation of S . rostrata with improved strains of ORS 571 markedly increased (34-50%) its dry matter
GREEN MANURING IN WETLAND RICE
(plant weight + pod weight) under both flooded and upland conditions.
The improved strain was more effective than native strain. In S. cannabina, S. speciosa, and S . uculeatu inoculation helped to improve their
growth (29-242%) under flooded conditions only. In Sri Lanka, Kulasooriya and Samarakoon (1990) reported that decapitation and stem inoculation of S . rostratu increased dry weight/plant by more than two times and
N yield by about three times over the control.
C. EFFECTOF IRRIGATION
Singh and Lamba (197 1) recommended that cowpea should be irrigated
when the available water in the 180-cm profile is depleted by 35%. Gaul et
al. (1976) reported that during summer in northern India, about 600650 mm of irrigation water would be required for raising a 74-day-old green
manure crop of S . aculeata on alkali soil. N . T. Singh et al. (1981) found
that irrigation frequency (irrigation water/pan evaporation = 0.5- 1 .O) exerted a significant influence on the dry matter and N yields of 7-week-old
green manure crops of S. aculeata, cowpea, and clusterbean in semiarid
regions of Punjab, India. In China, Gu and Wen (1981) reported that for
optimum yield of milk vetch, surface soil moisture (0-10 cm) should be
maintained at about 70% of water-holding capacity until winter in order to
speed up the root growth.
IV. TIME AND DEPTH OF INCORPORATION
OF GREEN MANURES
Traditionally, green manures were grown in fallow fields on rainwater
and incorporated 2 to 4 weeks before sowing of the following crop. This
practice is, however, not feasible in the context of intensive agriculture
when there is a fallow period of only 40-60 days before transplanting of
rice. While studying the possibility of green manuring in the present-day
rice-based cropping systems, Bhardwaj (1982), Ghai et al. (1988), and Beri
et al. (1989b) showed, on the basis of yield responses, that a 2-week delay
between incorporation of green manure and transplanting of rice was not
only unnecessary but also disadvantageous (Table 111). In fact, Williams
and Finfrock (1962) and Vachhani and Murty (1964) had already demonstrated that green manure could be incorporated even at the time of
transplanting rice seedlings. The reason for low efficiency of green manure
when incorporated for a longer period before transplanting rice or flooding
YADVINDER SINGH E T A L .
Effect of Interval between Incorporation of Green Manure and
Rice Transplanting on Rice Yield (t/ha)
Beri et al.
Ghai et al.
could be the loss of green manure N released during aerobic decomposition through ammonia volatilization, nitrification-denitrification, and
leaching after flooding of rice fields (Ishikawa, 1988; Chapman and Myers,
1987; Williams and Finfrock, 1962). Ishikawa (1963) observed that with
simultaneous flooding and milk vetch incorporation, loss of N through
apparent denitrification was small. When flooding began 10 days after
application, milk vetch decomposed rapidly under 10-day aerobic conditions and NH4+-N was converted to N03--N, which was lost upon flooding, possibly through denitrification.
A few investigations have shown that it is not always necessary to
incorporate green manure a day or two before transplanting rice (Iso, 1954;
Staker, 1958; Roy et al., 1988; Tiwari et al., 1980; Rana et al., 1988).
Swamp (1987) showed in field experiments that allowing decomposition of
Sesbania green manure for 1 week under flooded conditions in sodic soils
significantly improved rice yields over simultaneous incorporation and
transplanting of rice, possibly through improvement of physicochemical
properties of sodic soils. Wen (1984)and Herrera et al. (1989) reported that
it is better to turn under green manure crop about 15 days before transplanting rice seedlings so that plants do not suffer damage from the decomposition products of the green manure. To avoid losses of green manure N
it was recommended to keep the fields flooded during the decomposition
period before transplanting rice.
Ishikawa (1988) concluded that only in poorly drained fields with low
rates of nitrification was rice yield not influenced by flooding time in
relation to application of milk vetch green manure. Soil moisture conditions during the preflooding period significantly influenced the effectiveness of green manure (Williams and Finfrock, 1962). Under conditions
favorable to nitrification during the preflooding period, the shorter the
period, the greater the effectiveness.