Tải bản đầy đủ - 0 (trang)
VI. Contribution of Legume Nitrogen to Rice

VI. Contribution of Legume Nitrogen to Rice

Tải bản đầy đủ - 0trang

Table W



Nitrogen Fertilizer Substitution by Incorporated Green Manures (GM) on a Following Lowland Rice Crop"



Species and

country



G M duration



Interval from

GM incorporation

to transplanting



(d)



(4



NA

50-55

50-55



NA

1



50-55



60

2s

3s

45



1

I

NA

NA

NA



60

30

4s

60

61

50



GM incorporated



Rice yield

(t/ha)



N fertilizer

equivalence

(kg/ha)b



Crop

years



Dry weight

(tlha)



N

(kg/ha)



Without



With



GM



GM



3

1

1

1

2

1

1

1



4.6

3.4

4.0

4.8

5.4

1.5

2.7

3.8



78

97

120

149

110

49

71

89



3.2

2.7

2.7

2.7

3.2

2.8

2.8

2.8



4.0

3.3

3.5

3.1

5.6

3.5

3.8

3.8



38

28

37

52

132

41

62

62



4

4



1



2



3.8



87



3.2



5.3



116



3



15



1



IS



1

1

1

2

1

3

1

1

1

1

1

1



0.85

2.0

6.3

NA

NA

NA

4.8

0.23

2.0

5.6

2.3

3.4

3.7



24

70

170

NA

NA

NA

120

8

58

132

57

87

98



2.3

2.3

2.3

2.6

1.5

1.1

3.0

2.4

2.4

2.4

2.5

2.5

2.5



3.0

4.0

4.5

5.6

2.7

2.2

5.2

4.0



Reference"



Crotalaria juncea



India



Philippines



Cyamopsis tetragonoloba

India

Sesbania aculeata

Bangladesh



India



so



60



30

45

60

50-55

50-55

50-55



1



15

18

15



1

1

1

1



1

1



1

1



5.1



6.2

3.2

3.3

3.6



48d



>w

>w

95

53d

58d

84



42d



>fad

> 120d

24

27

38



1

2

2

2

3

4



5

5

5



6



1

7

8

9

9

9

2

2

2



Sesbania cannabina

India

Sesbania rostrata

Philippines

Sesbania sp. PL Se-17

India



Trifolium alexandrinum

Egypt

Vicia benghalensis

U.S.A.

Cowpea

India

Philippines



7.6

5.5

6.2

5.3

4.8



127

126

128

93

75



10

3



104

104



5.0

3.2

2.7

2.7

2.7



2.9



57



3.2



3.9



33



1



I



I .4



52



2.8



3.5



41



4



15



I

1

1



81

108



103



3.3

3.3

3.3

3.3



5.2

5.1

5.0

5.4



80

74

70

90



12

12



I



-



98



10



51



1



1



60

60

60

60



I



4.1

5.0

4.6

4.6

4.6



125

I08



1

7

14



2

4

4

4



NA



NA



3



25



NA



45

45

45

45



5

0



104



11



11

I1



12

12



NA

NA



NA

NA'



1

1



-



-



6.6

6.6



7.7

7.5



27

40



13

13



NA



NA"



5



NA



NA



3.0



4.9



51



14



60

60



1



45

45



45

15



3

2

1

I



2.8

6.9

2.5

2.3



73

113

70

62



3.0

3.2

3.4

3.3



5.3

5.7

4.2

4.3



89

137

34d

54d



8

3

15

I5



I



NA designates that information was not available in the reference.

Nitrogen fertilizer equivalence was the amount of inorganic N fertilizer computed from rice response models to produce rice grain equivalent to that

obtained with only green manure N applied to rice. Quadratic response models were used unless indicated to the contrary.

I , Bhardwaj et al. (1981);2, Sharma and Mittra (1988); 3, Ben et al. (1989a);4, IRRI (1988, p. 484); 5, Bhuiyan ef al. (1989);6. Dargan eta/. (1975);

7,Tiwarietal. (1980);8,Khindetal.(1982);9,Khindetal.(1983);10,Singhetal.(1988a); ll,BerietaI. (1989b); 12,Ghaietal.(1988); 13,Hamissaand

Mahrous (1989); 14, Williams et al. (1972); 15, John et al. (1989~).

Linear response models were used to calculate N fertilizer equivalence.

Rice was direct seeded rather than transplanted.



Table W

Nitrogen Fertilizer Substitution by lncorporated Legume Biomass, Remaining after Grain Harvest, on a Following Lowland Rice Crop



Species and

country



Interval from

incorporation to

transplanting

(d)



Biomass incorporated

Crop

years



Dry weight

(t/ha)



Lentil

India

Mung bean

India



15

15



3



60

60



I



1



Without

biomass



With

biomass



N fertilizer

equivalence

(kglha)"



Referenceb



~



~



~



Cowpea

Philippines



N

(kg/ha)



Rice yield (t/ha)



2.7

3.1

4.0



53

54

53



2.6

3.4

3.3



3.5

4.5

4.2



31

44'

50'



2

2



1



2.7

5.7



-



4.1

2.8



4.3

3.3



56'

31"



3

3



3



4.6



101



3.2



6.5



1

1



100'



1



4



Nitrogen fertilizer equivalence was the amount of inorganic N fertilizer computed from rice response models to produce rice grain equivalent to that

obtained with only legume biomass N applied to rice. Quadratic response models were used unless indicated to the contrary.

1, Kulkarni and Pandey (1988); 2, John ef al. (1989~);3, John el al. (3989d); 4, Rekhi and Meelu (1983).

Linear response models were used to calculate N fertilizer equivalence.



NITROGEN IN RICE-LEGUME CROPPING SYSTEMS



31



Grain yield (t/ha)



/'8



MCowpea, green manure



Y =4.25t 0.030N - 0.000092 N' ( I?'

4Cowpea ,residue

Y=4.30+0.U22N -0.000035N2 (R'

I Cowpea, residue removed

Y=3.27t0.031N-0.000068N2 (I?'

b.-4 Pre-rice fallow

Y=3.29t0.023N-0.000009N2 (I?'



c-



-



I

29



I

I

58

87

Urea N applied (kg/ha)



= 0.991

=0.98)

~0.99)

~ 0 . 9 17

I



116



FIG. 5. Effect of prerice treatment on response of lowland rice to applied urea in the

Philippines. (Average of 2 years; adapted from John et a / . , 1989c.)



following wet-season rice was greater after green manure (3.2 t/ha) than

after residue incorporation (2.3 t/ha) (IRRI, 1984, pp. 419-420).

At least part of the increase in rice yield from leguminous green manures

and residue can be attributed to increased soil-available N following legume incorporation (Nagarajah, 1988).The trends in rice grain yield for the

four treatments in Fig. 5 are directly related to the accumulation of soilextractable ammonium following rice transplanting (Fig. 6).



R.J. BURESH AND S. K. DE DATTA



32



Soil ammonium N (kg/ha)

45 I



o Cowpea, residue



5 -



A Cowpea, residue removed

A Pre-rice fallow

0



FIG.6. Effect of prerice treatment on ammonium N in the top 30-cm soil layer during a

lowland rice crop in the Philippines. Values for each sampling day followed by a common

letter are not significantly different, based on LSD (.05). (Adapted from John, 1987.)



A. MINERALIZATION

OF LEGUME

N



The soil water regime during decomposition of green manures and residue varies among rice-based cropping systems. In irrigated and rainfed

lowland environments with rice establishment by transplanting or sowing

of germinated seeds, the soil is typically flooded or saturated. In irrigated,

rainfed lowland, and upland environments where rice is established by

sowing in dry or moist soil, the soil is typically aerobic until saturated by

rainfall or irrigation. In rainfed lowlands with erratic rainfall, the soil may

undergo alternate drying and flooding.

Models of organic matter decomposition in upland (van Faassen and

Smilde, 1985)and flooded soils (Bouldin, 1988)have characterized organic

amendments as containing two distinct components: one decomposing

rapidly within a few months and the other decomposing slowly over several years. Singh et a1.(1988b), for example, reported that the initially fast

and subsequently slow release of mineral N during decomposition of



NITROGEN IN RICE-LEGUME CROPPING SYSTEMS



33



7-week-old Sesbaniu aculeata green manure could be described by two

simultaneous first-order reactions. Nitrogen content, C/N ratio, and lignin

content of legumes (Frankenberger and Abdelmagid, 1985) and soil temperature influence N mineralization rate in flooded and upland soils. Temperature must be considered when comparing research results between

tropical and temperate areas.

1 . Flooded Soil



Flooded soils are characterized by a thin, oxidized, surface soil layer

overlying an anaerobic soil layer (Ponnamperuma, 1972). After soil flooding, ammoniacal N accumulates. Extractable soil ammonium normally

peaks by 20 days after rice transplanting and then decreases, presumably

because of plant uptake (Nagarajah, 1988). Formation of nitrate is restricted to oxidized soil zones, but this nitrate can diffuse to anaerobic soil

zones where it is prone to gaseous loss by denitrification (Reddy and

Patrick, 1986). Oxidation of soil during periods of drying may lead to

formation of nitrate, which can be lost by denitrification after soil wetting

(Buresh and De Datta, 1990).

The beneficial effect of green manures on yield of a following rice crop is

dependent upon incorporation of the green manure in both nonpuddled

soils with water-seeded rice (Williams and Finfrock, 1962) and puddled

soils with transplanted rice (IRRI, 1986, p. 406). Green manures and plant

residues incorporated immediately before or after initial soil flooding typically decompose under transitional aerobic to anaerobic soil conditions

and then under anaerobic soil conditions. Although the rate of decomposition of plant material is slower under anaerobic than under aerobic

conditions, the net release of ammonium may be greater under anaerobic

conditions because of the lower N requirement for anaerobic metabolism

(Patrick, 1982).

Under high-temperature tropical conditions, the mineralization of organic amendments can be as great in flooded rice fields as in upland soils

(Neue and Scharpenseel, 1987; Neue and Bloom, 1989). Moreover, net

release of N from added plant material occurs at C/N ratios that are

relatively higher under anaerobic than under aerobic soil conditions.

Leguminous green manures decompose rapidly following incorporation

in tropical flooded soils. Following incorporation of S . cannubina (syn: S .

aculeata) (Khind et al., 1985; Nagarajah, 1988; Beri ef al., 1989a), S.

rostrata (Nagarajah, 1988; Becker e f al., 1990a;Diekmann, 19901, Aeschvnornene afraspera (Diekmann, 19901, Crotalaria juncea (Nagarajah, 1988:

Beri et al., 1989a). clusterbean (Beri et al., 1989a), and cowpea green



34



R. J. BURESH AND S. K . DE DATTA



manure (Beri et al., 1989a; Fig. 6) in tropical lowland rice fields, the

accumulation of soil ammonium peaked at 7-20 days after rice transplanting and then gradually declined. After reaching a peak, soil ammonium N

levels declined more rapidly in the Philippines with broadcast-seeded rice

on wet puddled soil than with transplanted rice because of more rapid early

growth and N uptake for broadcast-seeded rice (Diekmann, 1990).

On a sandy loam soil in Punjab, India, about 80% of the total N in S .

aculeata green manure was mineralized by 10 days after incorporation

(Ben et al., 1989b). About 40% of the green manure C was evolved as

carbon dioxide in 20 days, and about 50% was evolved as carbon dioxide in

40 days (Beri et al., 1989b). Grain yield of irrigated lowland rice in a field

study in the Philippines with three Sesbania species of two growth durations was a direct function of extractable soil ammonium at 10 days after

green manure incorporation (7 days after transplanting) (Furoc and

Morris, 1989).

Nagarajah et aL(1989) reported faster and larger accumulation of soil

ammonium from incorporated S. rostrata than from Azolla microphylla

even though they had similar N contents (41.2 and 42.5 g/kg) and C/N

ratios (1 1 and 9). They attributed the differences to the lower lignin content

of S . rostrata (94 g/kg) than of A. microphylla (205 g/kg) (Frankenberger

and Abdelmagid, 1985). The higher N content, lower C/N ratio, and

slightly lower lignin content of 7-week-old Aeschynomene afraspera than

of S. rostrata grown in the Philippines raised speculation that A. afraspera

green manure may release ammonium faster than does S. rostrata green

manure (Becker et al., 1990b).

Residues of grain legumes, which frequently have a lower N content

than that of green manures, also rapidly release ammonium in tropical

flooded soils. Nagarajah (1987) determined the net N release for two

legume green manures (3.rostrata and Crotafariajuncea) and five legume

residues (cowpea, mung bean, groundnut, pigeonpea, and soybean) after a

50-day incubation in flooded soils without rice. Plant N contents ranged

from 11 g/kg for soybean to 27 g/kg for S . rostrata; C/N ratio ranged from

17 for S. rostrata,cowpea, and groundnut to 38 for soybean; and lignin

content ranged from 57 g/kg for mung bean to 134 glkg for pigeonpea. Net

recovery of plant N as ammonium (AR) at 50 days ranged from 16% for

soybean to 43% for S . rostrata. It correlated directly with plant N (N) and

inversely with C/N ratio (R):

AR

AR



=

=



3.50

46.1



+ 0.59 N

-



0.79 R



r = .94

r = -.91



Similar patterns of accumulation and decline in extractable soil ammonium have been observed for incorporated cowpea green manure (N =



NITROGEN IN RICE-LEGUME CROPPING SYSTEMS



35



27 g/kg, C/N = 15) and residue (N = 13 g/kg, C/N = 30) under field

conditions in the Philippines (Fig.6). The net recovery of incorporated

aboveground cowpea N as extractable soil ammonium N at 15 days after

transplanting (30 days after incorporation) was 37% for green manure and

28% for residue.

Leaves, stems, and roots of legumes can differ in rate of decomposition

and N release because of differing N contents and C/N ratios (Palm et

al., 1988). Ventura et al. (1987) reported much higher N content of leaves

(51 g/kg) than of stems (1 1 g/kg) and root stubble (9 g/kg) for 47-day-old S.

rostrata. Palm et al. (1988) similarly observed higher N content in leaves

(38 glkg) than in stems (4.1 gfkg) and roots (7.3 g/kg), and lower C/N ratio

in leaves (1 1) than in stems (107) and roots (55) for 84-day-old S. sesban.

More than 80% of the leaf N was mineralized in 14 days, but 80% of the

stems and 75% of the roots remained undecomposed after 56 days.

Addition of green manure and plant residue to flooded rice soil can

enhance formation of methane (Tsutsuki and Ponnamperuma, 1987).

Methane is a greenhouse gas linked to global warming (Blake and Rowland, 1988; Bouwman, 1990).Therefore, enhanced emission of methane to

the atmosphere could be an undesirable consequence from use of green

manure and plant residues in flooded rice soils.



2 . Aerobic Soil

In rainfed and irrigated environments where rice is dry sown in nonflooded, aerated soil, the incorporation and initial decomposition of leguminous green manures and residues are normally under aerobic soil conditions. Ammonium formed by mineralization of legume N can oxidize to

nitrate (Singh et al., 1988b; Beri et ul., 1989b). Rates of N mineralization

and nitrate accumulation depend on tillage and depth of residue placement

(Wilson and Hargrove, 1986) and on soil water content, which can fluctuate as a result of intermittent rain or irrigation. In lowland environments,

the soil is eventually flooded by either rain or irrigation, resulting in

depletion of soil oxygen. Under these anaerobic conditions, nitrate can be

lost by denitrification (Williams and Finfrock, 1962).

In Louisiana, the yield of dry-sown rice averaged for four rates of urea

application was about 10% greater following subterranean clover (Trifolium subterraneum L.) than it was following bare fallow (Dabney et al.,

1989). Desiccating the clover with a herbicide and leaving the residue on

the soil surface resulted in rice yields similar to or greater than those

obtained after incorporating the clover with tillage to a 10-cm depth.

Leaving legume residue on the soil surface rather than incorporating the



36



R. J. BURESH AND S. K. DE DATTA



residue decreases the mineralization rate of legume N in aerobic soil

(Wilson and Hargrove, 1986). A reduction in mineralization of legume N

during the period from rice sowing to a permanent flood could decrease the

amount of soil nitrate susceptible to leaching and denitrification losses

before and after permanent soil flooding (Dabney et al., 1987).

In general, only small increases in grain yield and savings in inorganic N

fertilizer have'been reported for tropical dry-sown rice following legumes

grown for grain and green manure production. In Australia, rice sown 39 to

75 days before permanent flooding yielded slightly more following incorporation of either soybean green manure, soybean residue, mung bean

residue, or S. cannabina residue than following either fallow or sorghum

with incorporation of residue (Chapman and Myers, 1987). In the Philippines, yield of dry-sown lowland rice was about 0.2 t/ha higher following

soybean than following sorghum (Furoc et al., 1979). However, the soybean and sorghum treatments in this study were not randomized.

Little is known about the mineralization and N contribution to rice of

haulm (stems and tops), roots, and nodules of postrice legumes grown in

rainfed lowlands (Fig. 4). Haulm of postrice legumes typically remains

unincorporated for several months until after the onset of rains in the next

wet season, when the soil can undergo alternate wetting and drying cycles

before permanent flooding. Soil nitrate formed from mineralization of

legume N before soil saturation will likely be lost by denitrification or

leaching rather than used by rice (Buresh et al.. 1989).

Sisworo et al. (1990) examined the N contribution of plant residues to

the following crop in upland rice-soybean-cowpea and rice-maizecowpea rotations on an Orthoxic Palehumult in the humid tropics of

Indonesia. They concluded that the N contribution of a residue to a

subsequent crop on upland soils was controlled by the N concentration of

the residue and modified by the soil water status. From their results they

developed an equation to predict the percentage of added residue N taken

up by the subsequent crop(s):



S = -1.05



+ 4.37 N + 0.0067 N R



R2 = 0.69



where N is the percent N concentration in the residue and R is the expected rainfall in mm for the subsequent crop.



B. BELOWGROUND

LEGUME

N

Reports vary greatly on the N content, N mineralization, and contribution to rice of legume roots. The conflicting reports may partly result from

large differences in N content and rate of N mineralization for roots among



NITROGEN IN RICE-LEGUME CROPPING SYSTEMS



37



different species and cultivars of the same species (Nnadi and Balasubramanian, 1978). Chapman and Myers (1987) found 23 to 50 kg N/ha in the

nodulated roots and stem bases of mature soybean, mung bean, and S .

cannabina. Roots and stem bases accounted for about 11% of the total

plant N for soybean and about 25% of the total N for mung bean and S.

cannabina. Considerably lower levels of N in nodulated soybean roots

were reported by Bergersen et al. (1989) (Table VIII).

Root nodules have a higher N content and lower CIN ratio than do

legume roots. Consequently, nodules can rapidly decompose and release

mineral N (Chulan and Waid, 1981).Because the fraction of total legume N

contained in root nodules is small (Bergersen et al., 1989),the contribution

of mineralized nodule N to rice yield can be undetectable.

Several studies indicate that roots of green manure legumes contain 10%

or less of the total plant N (Table VIII). The low N content and slow

mineralization of Sesbania roots suggest that they are not important contributors of N to a following rice crop (Crozat and Sangchyosawat, 1985;

Palm et al., 1988). Field studies have confirmed that rice yields are comparable following incorporation of equivalent quantities of green manure

N (Sesbania or cowpea) either grown in situ before the rice crop or

transported from a nearby field (green leaf manuring) (Singh et a / . ,

1988a; Morris et al., 1989). Moreover, when aboveground biomass of S .

aculeata grown before rice was removed, yields of the following rice

were not greater than those obtained following a fallow (Singh et a / . ,

1988a).

There are conflicting reports on the benefits of food legumes to a subsequent rice crop when aboveground legume biomass is removed. De et al.

(1983), in a study with removal of aboveground biomass of crops grown

before rice, reported increased yield of lowland rice following cowpea,

mung bean, and black gram (Vigna mungo) as compared with yields

following a maize fodder crop. Other studies, which compared grain legumes with fallow before rice, reported no increase in rice yield (John er

a / . , 1989~;Fig. 5) or only a slight, nonsignificant increase in rice yield

(John et al., 1989d) when aboveground legume residue was removed.

Incorporation rather than removal of haulm from prerice mung bean

can increase yield of

(Maskina et al., 1990)and cowpea ( John et al., 1989~)

a following transplanted rice crop.

Conflicting reports on the benefits of legumes to rice, when aboveground

legume biomass is removed, can be attributed at least partly to differences

in the treatment used for comparison with the legume. For example, De et

al. (1983) compared legumes with maize fodder, whereas John et al.

(1989~)compared legumes with a traditional weedy fallow. Removal of

maize, unlike incorporation of weeds following a fallow, results in net loss



Table Vm

Nitrogen Accumulation in Aboveground Biomass and Roots of Legumes

Aboveground

Species and

country

Aeschynomene indica

Thailand

Sesbania aculeata

India

Philippines

Sesbania cannabina

Australia

Sesbania rostrata

Philippines

Thailand

Sesbania sesban

Sri Lanka

Mung bean

Australia

Soybean

Australia



Roots



Plant N

abovegroundb



Duration"

(d)



Dry weight

(t/ha)



N

(kdha)



Dry weight

(t/ha)



55



1.9



37



0.3



4



91



1



60

62



4.6

6. I



104

99



0.9

0.8



14

I



88

93



2

3



50



75



4



MAT



152



N

(kdha)



(%I



Reference'



47

55



3.4

7.0



73

128



0.6

0.7



5

6



94

96



5

1



84



4.4



83



I .2



9



90



6



MAT



132



40



77



4



MAT

MATd



292

105-342



37

3-10



89

96-98



4

7



MAT designates crop maturity.

Percentage of the total plant N located in aboveground plant parts.

I , Crozat and Sangchyosawat (1985);2, Ben et al. (1989b);3, IRRI (1986, p. 416); 4, Chapman and Myers (1987);5, Ventura et al. (1987); 6, Palm e f

a / . (1988); 7, Bergersen et al. (1989).

Roots plus nodules sampled at 105 days; aboveground N determined at maturity (146 days).



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

VI. Contribution of Legume Nitrogen to Rice

Tải bản đầy đủ ngay(0 tr)

×