4 FIELB RESEAREHi
Centre for Soil Researc;i Ji. Ir H. Juanda 98: Bogor: Indonesia (0251) 23012
Contact: TROPSOILS; Box 02; Sitiung 1A; Sumatera Barat
DATE: JANUARY 1987 NUMBER: 36
TITLE: Source and Management of Green Manure 1985/86 Season
EXPERIMENT NO.: 3503
RESEARCHERS: Carl Evensen and Russell Yost
1. To compare three legume green manure crop-s for soil
ameliorating effects on subsequent crops.
2. To compare three methods of production and application of
green manure organic material.
3. To determine the extent by which green manures.reduce lime
4. To determine the response curve of upland rice to
inorganic N and to compare this response to that from the
application of a legume green manure.
SOIL: Typic Paleudult; clayey, kaolinitic, isohyperthermic.
FCC: Ceak. Cleared by manual tree felling and then
bulldozing during the 82/83 wet season but never cropped.
The experiment is being conducted in a Randomized Complete
Block design with the following treatments:
Species: 1) Calo0oonium mucunoides
2) Crotalaria usaramoensis
3) Centrosema pubescens
Methods of Green Manure Application:
1) In situ incorporation
2) Cut and carry
3) Root residue and and stubble from plots used to
produce th.e cut and carry material
Levels of Lime: 1) LO = no lime
2) L1 = 375 kg lime/ha
3) L2 = liming to 40% Acid (Al+H) saturation
4) L3 = liming to 20% Acid saturation
LeveT of Nitrogen: 1) NO = no N
2) N1 = 30 kg N/ha (split application)
3) N2 = 60 kg N/ha
4) N3 = 120 kg N/ha "
Green Manure s.. Aplication
(GM species and application m
species 1 method 1
species 2 method 1
species 1 method 2
species 2 method 2
species 1 method 3
species 2 method 3
check (no GM)
species 3 method 1
(Lime response with GM also a
species 2 method 1
species 2 method 1
species 2 method 1
species 2 method 1
(Lime response with no GM app
check (no GM) -
check (no GM)
check (no GM)
check (no GM)
(Inorganic nitrogen response)
check (no GM)
check (no GM)
check (no GM)
check (no GM)
Method (Nitro en,Limej
MATERIALS AND METHODS:
This report covers the crops grown during the 1985/86 cropping
season. These include the green manure (GM) crops grown during the
1985 dry season, and then upland rice and peanuts grown during the
following rainy season. See FRB 16 for a history of the uniformity
trial and plot selection preceding these plantings. Following the
uniformity trial (in which all plot had received 40 kg P/ha as
TSP), the liming treatments were applied according to the treatment
plan above on June 13, 1985. However, lime was applied at
different rates in each replication to create the desired acid
saturations (using the Cochran equation) due to differing mean acid
saturations and extractable acidities (Table 1).
Table 1. Soil analysis means and lime applied by replication.
Rep. A1+H Acid Sat. --- Lime Applied (kg7ha) ---
(cmol/kg) (%) 40% Acid Sat. 20% Acid Sat.
I 1.99 70.1 1000 1600
II 2.53 77.1 1400 2080
III 2.23 74.1 1200 1900
The GM crops (see Experimental Factors above) were planted
on June 14, 1985 at a spacing of 40 cm between rows and densely
within the rows (for Crotolaria about 50 seeds/m and for
Cal-opogonium and Centrosema about 25 seeds/m). The plot size was
6m x 4m. The months of June, July, and August are generally
quite dry, averaging 100 mm of rainfall/month, as opposed to 250
to 400 mm during the rainy season. During these dry months crop
production is quite tentative and farmers often leave their
fields fallow. These GM crops were planted during this normally
fallow period to make this technology more attractive to farmers
since food crops are not replaced). The GM growth was very slow
at first due to a drought and gaps in the stand were replanted on
July 18th. Several days of rain followed and the germination and
subsequent growth of the GM crops was quite vigorous.
The GM crops were harvested on Sept. 19, 1985 with a harvest
area of 4m x 2.4m and incorporated with hoes according to the
treatment plan as given above. Litter on the soil surface
and root residue produced by the green manures were also
estimated since this is what is-left in the green manure residue
treatments. Two strips, each 40 cm wide and 4 m long, were
harvested in each of the green manure residue plots (treatment
numbers 5 and 6). Surface litter and stubble was collected and
roots dug to 30 cm depth and shaken free of soil. These samples
were dried and then sieved to remove remaining soil. This simple
procedure did not result in complete recovery of fine or deep
roots and is only a rough estimate. TSP (10 kg P/ha) and KCL (25
kg K/ha) were also incorporated at the same time as the green
manures on all plots.
A local variety of upland rice (disease tolerant and
preferred by local farmers) was planted at a spacing of 25cm x
25cm on Sept. 24, 1985. The seed was planted in dibble holes
along with 3 % carbofuran granules (625g a.i./ha). The
germination was fairly good but there were a few gaps in the
stand which were replanted after the first weeding on October
8th. Half of the required nitrogen was applied in the N
treatments on October 23rd (ie. 15, 30,.and 60 kg N/ha) and the
other half on November 27th. Following the second appTication,
the field was weeded again. Insect control was achieved with
regular sprayings of diazinon on Oct. 12, Dec. 6 and 30, Jan. 2,
12, 15, and 17. The diazinon was mixed with Azodrin
(monokrotophos) on Dec. 30 and Jan. 2 and with Sevin (carbaryl)
on Dec. 12 and 17.
Blast and Helminthosporium Brown Spot caused severe leaf
damage in November, with many leaves completely dieing, but
amazingly by December the rice had produced new leaves and
recovered well. This tolerance to diseases is the reason a local
variety was grown. Grain set was very heavy and uniform, with
treatment effects not evident. However, very heavy winds and
rain in mid-January caused serious lodging throughout the
experiment. Due to this lodging, the harvest areas were reduced
to portions of the plots that were still intact for the first
harvest of mature panicles on January 21 as well as the final
harvest on February 1, 1986. Harvest areas2ranged from 3.3 m2 to
10.5 m with most (45 plot) measuring 8 m The rice straw was
all returned to the plots and buried by hoe along with TSP at 10
kg P/ha. Treatment 15 was adapted by discarding the rice straw to
determine the effect of straw on yield of a subsequent crop. In
regression analyses of the lime response with no GM (Group 3),
treatments 7 and 17 will be substituted for treatment 15.
A local variety of red peanut (Valencia type) was planted at
25cm x 25cm on February 16, 1986. The crop was sprayed once with
Sevin (carbaryl) on March 24th. The peanut growth was quite good
but rat damage began to occur in mid-April. This was relatively
minor (less than about 5 %) and was corrected by slightly
reducing harvest areas to exclude damaged hills. The peanuts
were harvested on May 13, 1986 with most harvest areas being 8 m
Excellent nodulation was observed on most plots during harvest.
Of the treatments, only the liming rates had a clearly observable
influence on peanut yields, with higher rates producing larger,
greener tops and better filling of the nuts. After peanut
harvest, the green manure crop treatments were again planted.
The yields of the green manure crops are shown in Table 2.
Above ground dry matter yields of crotolaria increased with
increasing liming rates and was higher than the yields of
calopogonium and centrosema at the-equivalent lime rate (lime 2).
However, all green manure crops grew fairly well and produced
fairly large amounts of dry matter, considering the dryness of
the growing season. It would seem that all three of these green
manure crops are suitable for growth during the dry season.
Table 2. Green manure crop yields.
_C__r --e-en__M-a-n-u-r-e ------- __-ove-C_--u-jB-- _So-1Su-~~~ce- -- _00 _
Green Manure Above Ground Dry Soil Surface Root
Species Matter Yield Litter Residue
-------------------- kg7ha ------------------
Crotalaria Lime 0 1923
Lime 1 2806
Lime 2 3408 75 690
Lime 3 3710
Calopogonium (Lime 2) 2198 400 640
Centrosema (Lime 2) 2303
LSD (0.05) 949 -
The growth of the upland rice was very uniform and yields
quite high, averaging 3150 kg/ha (with a range of 1880 to 4330
kg/ha). There were no significant treatment effects. Table 3
shows that there were no significant differences between
treatments comparing methods of green manure application. The
coefficient of variability was a low 10.1 %, indicating that this
lack of treatment effect is quite believable and not due to
unexplained variability. Also, liming and nitrogen rates did not
significantly effect yield. It seems that at the moderately high
liming rates (to at least 40 % acid saturation for most of the
treatments), the upland rice did not respond to either green
manure or nitrogen application. A crop more responsive to
liming, such as maize, should be used as the test crop next year.
Table 3. Analysis of variance for effects of methods of green
manure application on rice and peanut yields.
Source df Upland Rice Peanut
------- PROB > F -------
GM TREATMENTS (6) 0.627 0.135
GM vs. no GM 1 0.379 0.230
Crotalaria vs. Calopogonium 1 0.620 0.031 *
Stubble vs. others 1 0.640 0.948
Cut and Carry vs. In situ 1 0.494 0.111
REPLICATION 2 0.030 0.003 *
C.V. (%) 10.1 13.6
Yields of the subsequent crop of peanuts were influenced by
treatment effects and ranged from 100 to 2000 kg nuts/ha (with an
overall mean of 980 kg nuts/ha). However, the residual effects
of green manures applied in .September did not have a significant
influence on yields, as shown in Table 3. Crotalaria GM did
cause significantly higher yields than Calopogonium GM, but since
green manuring did not differ significantly from no green
manuring, this may mean little. The coefficient of variability
for these GM comparisons was again an acceptably low 13.6 %. It
should be noted that the overall C.V.s (all 20 treatments) was
14.6 % for the upland rice and 25.1 % for the peanuts. Coupled
with the significant replication effects for both crops, this
indicates that the preliminary uniformity trial and special
efforts of stratifying plot variability to replications (see FRB
No. 16) helped to reduce experimental error due to soil
variability. This may prove to be a useful technique in field
experimentation on soils with high micro-variability, such as on
newly cleared land.
Liming significantly increased peanut yields, but linear
regressions with or without green manure did not differ
significantly, as determined by comparing confidence intervals of
slopes and Y-intercepts (Table 4). Since lime rates were
determined by percent acid saturation, this was used in the
regression. Figure 1 shows the linear regression between peanut
yield and soil acid saturation percentage (A1+H/ECEC) for liming
treatments both with and without green manure. This provides a
better expression of the response relationship than would liming
rate since soil microvariability is high.
Table 4. Comparison of linear regression equations for peanut
response to soil acid saturation, with and without GM.
Soil 5 % Confidence 5 % Confidence
Ammendment Interval for Slope Interval for Y-intercept
Lime with GM -20.8 + 10.3 2024 + 537
Lime without GTl -18.8 + 7.1 1741 + 357
To better identify the soil chemical factors associated with
peanut yield, regressions were calculated between grain yield and
a number of soil test variables (Table 5). The quadratic
regressions with exchangable Ca (cmol/kg soil) have the strongest
relationships to peanut yield. Also the weaker relationships with
extractable acidity (Al+H) suggest that the Ca nutrition rather
than the reduction of Al toxicity is more important in liming for
peanut production. The major point of interest, however, is that
green manuring using the rates and methods described, did not
reduce liming requirements as had been hypothesized. Table 6
shows soil analyses for samples taken on April 14, 1985 during the
growth of the peanut crop. It is interesting to note the apparent
increase of exchangable Mg with increasing liming rates and that
the desired acid saturations were achieved fairly well.
The legume species Crotalaria usaramoensis, Calopo.onium
mucunoides, and Centrosema pubescens can be successfully grown as
green manures during the dry season in Sitiung. However, during
the 1985/86 growing season, neither upland rice nor peanuts
responded to green manure application. The well distributed
rainfall may in part explain the lack of response, since rainfall
was eratic in previous trials in which strong responses to green
manures were observed. Also, a crop more responsive to N and less
tolerant to Al, such as maize, may respond to green manures
better. It is possible, however, that herbaceous green manures
produced in situ on infertile sites, and not transported in from
other more fertile sites, may not increase yields of a rotation of
upland rice and peanuts.
/ 1 1 I1
20 30 40 50
60 70 80 90
% ACID SATURATION
Figure 1. Peanut yield response to percent acid saturation.
A= No GM
Y = 1862 19,6X
r = 0.809 -
Table 5. Regression equations for peanuts with and without GM.
LIMING TREATMENTS WITH GM
X-Variable Equation R-s ua r e
% Acid Sat. Y = 2024 20.8x 2 0.6709
% Acid Sat. Y = 2115 24.9x + 0.04 x 0.6713
Exch. Ca Y = 201 + 807x 0.7717
Exch. Ca Y = 528 156x + 501 x 0.8153
Extr. A1+H Y = 2007 723x 0.5955
Exch. Mg Y = -33.7 + 4889x 0.4585
LIMING TREATMENTS WITHOUT GM
% Acid Sat. Y = 1741 18.8x 0.7167
% Acid Sat. Y = 427 + 41.6x 0.61x2 0.8434
Exch. Ca Y = 288 + 536x 0.6522
Exch. Ca Y = -375 + 2265x 803 x 0.9137
Extr. A1+H Y = 1941 871x 0.5937
Exch. Mg Y = 296 + 2681x 0.3409
LIMING TREATMENTS WITH AND WITHOUT GM
% Acid Sat. Y = 1862 19.6x 0.6551
% Acid Sat. Y = 1319 + 5.18x 0.246x2 0.6734
Exch. Ca Y = 268 + 636x 0.6528
Exch. Ca Y = -78.1 + 1573x 454 x 0.7133
Extr. A1+H Y = 1879 732x 0.5184
Exch. pg Y = 195 + 3463x 0.3689
Table 6. Analyses of soil samples (0-15cm) taken during the growth
of the peanut crop on April 14, 1986.
SLimeC a Mg K ECEC Pb Z Acid
Rae A a ---a cmol kg ---------- ppm Sat.
0 1.84 0.31 0.12 0.23 2.51 6 73
375 kg/ha 1.57 0.63 0.19 0.28 2.69 4 58
40% Acid Sat. 1.12 1.22 0.21 0.24 2.82 5 40
20% Acid Sat. 0.86 1.70 0.30 0.27 3.13 4 27
Extracted with IN KC1.
b Extracted with Mehlich I extractant.
Field Status Report September to December 1986
Center for Soils Research
1002. Lime Reaction. The maize was nearly destroyed by wild
pigs and harvesting will be difficult. Liming treatment responses
were quite strong throughout the growth of the crop, indicating
that the earlier effect of low lime requirement on this recently
cleared soil was diminishing The next crop will be soybeans.
1003. L x P Model. The maize was in late dough stage at the
end of December and is expected to be harvested in the third week
of January. Plant biomass samples at three growth stages were
collected and will be sent to Bogor for analysis.
1103. Lime Reaction. The maize crop responded well to lime and
will be harvested in early January.
1104. Ca & Mo on Legumes. The experiment was harvested on
September 25 T5th harvest) and leaf samples were taken. Most of
the Desmodium intortum plot have now died and there still seems to
be no response to the molybdenum treatments. This means that over
half of the plots are no longer providing useful information.
Because of this and since we have a year and a half of good data,
it is recommended that this experiment be discontinued. The site
and the treatments may be useful for other experimental purposes,
such as perhaps Istiqlal Amien's work with Al tolerance of peanuts.
1105. Mg. The maize and rice crops are growing well and were
both in flowering in December. Early visual symptoms of the
maize seems to indicate that it is responding to Mg, but like
last year, this may not translate into yield responses. In any
case, low Mg levels get rid of deficiency symptoms. The upland
rice shows no response to Mg.
1106. .Flooded Rice Fe Toxicity. Rice was harvested in late
December and the data is being processed. We are waiting for
clarification from Bogor as to whether or not this trial will be
.1107. Rock Phosphate. Upland rice, 'Sentani' variety was in
milk stage in late December. A moderate infestation with
Helminthosporium Brown Spot was noted.
1108. Nitrogen Sources. Maize was in early dough stage at the
end of December. Pigs had attacked several plots even though the
experiment was fenced and bordered. A night watchman was hired
to guard from further pig problems.
1109. Peanut Variety x Limin_. This new experiment, which will
provide preliminary information for Istiqlal Amien's dissertation
research on Al tolerance of peanuts, was planted on Dec. 22, 1986.
It consists of a split plot design with three lime levels as main
plots (0, 0.5, and 2 T/ha) and nine peanut variety as subplots
('Anoa', 'Banteng', 'Gajah', 'Pelanduk', 'Rusa', 'Tapir', 'Tupai',
'Blanco Tarapoto', and a local variety--of these only the 'Blanco
Tarapoto' did not germinate well). Amien will use this
information to select Al tolerant, intolerant, and intermediate
1202. Variability. The crop of upland rice was in milk stage
in late December and showed some symptoms of Helminthosporium
Brown Spot. Growth variation due to soil variability is very
obvious. In some spots, there isn't even vegetative growth while
in other spots, growth is lush.
1203. Lime Rate. This experiment is planted in a split plot
design with a local and an improved variety being the subplots.
The local variety was in early milk stage and the improved
variety in dough stage at the end of December. The improved
variety showed some symptoms of Brown Spot while the local
variety did not.
1404. Lime Reaction. The maize crop shows a strong response to
lime and will produce good yields. It is expected to be
harvested in mid-January.
1405. L x P Model. The upland rice crop shows a strong
response to the lime and P treatments. It was still in flowering
stage at the end of December. There is a slight disease
infestation, probably Helminthosporium Brown Spot.
1406. Nitrogen Sources. Upland rice is growing well on some
plots but poorly on other and no pattern of treatment effects is
evident. This variability is most likely due to soil
2001. Lime-Residue management. The soybeans were in pod filling
stage in late December. Weeds were very troublesome for plots
with plant residues incorporated, due to the large number of weed
2002. Tillage-Residue. The soyb-ean crop was in pod filling
stage in late December. Growth is fair.
2003. Crop Production Land Manaaement. The experiment is
underway. The cover crop was planted in December and the plots
are not yet full. Final treatments and planting are expected to
be implemented in March.
2004. Root Growth Limitations. This experiment is underway.,
Plastic cores have been cut. A large supply of Al saturated soil
has been brought from Sitiung ID. It has been dried, crushed and
sieved. Preliminary runs will begin soon.
2401. Erosion. Maize was in late dough stage in late December.
Repairs were underway on the soil collectors. All run off and
soil loss measurements are continuing.
3002. Leguminous species collection. Monitoring continues.
3003. Tree Legumes. Tree height and diameter measurements were
taken at 9 months and 1 year from transplanting seedlings (in
September and December). The most vigorous species is still
Albizia falcataria which averaged about 7.5 m tall at 1 year.
Other vigorous species are; Acacia mangium, Calliandra
calothyrsus, and Cassia siamea. On Dec. 23, 1986 a new increment
was added to the trial with the transplanting of six Leucaena
leucocephala varieties which were obtained from the Nitrogen
Fixing Tree Association in Hawaii. Several of the new varieties
are purportedly acid tolerant.
3101. CIAT Trial B. This experiment was harvested after a
regular 2 month interval on October 29, 1986. This was then the
start of another set of intensive (every 3 weeks)_harvests which
will be the last harvests for the experiment as it currently
stands. The most productive species continue to be the grasses
Brachiaria dictyoneura and Andropoon2 2ayanus and the legumes
Centrosema macrocarpum, Desmodium ovalifolium, and Stylosanthes
suianensis. Since two years have elapsed since planting (5
regular harvests and 4 intensive harvest cycles), sufficient data
has been collected to characterize the growth of the species in
this experiment. The site will now be handed over to Dr. Ron
Guyton who will probably restructure the experiment.
3502. Alley CroLpin". The tree prunings were applied on Sept.
4 and a local variety of upland rice planted on Sept. 13, 1986.
The trees were again pruned on Nov. 1 and the prunings applied as
mulch. The growth of the rice was excellent at first, but by late
November had developed a general leaf yellowing, lower leaf
necrosis and leaf spot. This is a general condition of upland
rice throughout Sitiung V this year and may be the disease
Helminthosporium brown spot, possibly exacerbated, by K-'deficiency.
The yields will still be fairly good in'this experiment and may
show a very strong replacement effect of green manure for lime.
Rice harvest will begin in mid-January 1987.
3503. Cover Crop Management. The cover crops which had grown
through the dry season, were applied on Sept. 11 and maize
planted on October 6, 1986. The growth of the maize test crop
was quite good and strong treatment effects were shown to liming
.while nitrogen and green manure effects may also be evident.
Treatment differences will be much greater than for last year's
crop of upland rice. Pig damaged a few plants but subsequent
guarding of the crop precluded further damage. Harvest will take
place in early January 1987.
3505. Tree Legumes. No new data was collected during this
period, however, visually the trees in this trial are very
similar in growth to the trees in trial no. 3003.
3701. Green Manure Management. This experiment in Kuamang
Kuning has had some serious problems in that 2 of the green
manure plantings (out of 4 total) did not grow. The reasons
reported for these failures were shading during the 3rd growth
period (green manure planted into a standing soybean crop) and
drought during the 4th growth period. Nevertheless, the existing
green manures were incorporated and maize was planted in October,
1986, to try to get some useful data out of trial. Continuation
of this trial will depend on the outcome of this planting.
4504. Farmer Practice and Production Study. Data on last
cropping season (approx. July-November) has been collected. This
represents primarily peanut and soybean production along with
some rice yields. In order to more accurately document home
garden production, an intensive twice-weekly survey was begun in
October. Field Research Brief # 32 Characterization of Home
Gardens in Aur Jaya was completed and distributed.
4505. Economics Survey. Survey has been completed and a
preliminary report written. Final report will be written upon
completion of more sophisticated analyses at Univ. of Florida.
4511. Record Keepien. Stephenie Kan has completed data
collection and summarized her findings in a preliminary report.
She returned to the Univ. of Florida in December and will
complete her data analysis and final report writing there.
4512. Consumption Survey. In order to determine type, amounts
and source of daily food consumed by Sitiung V farmers and its
specific nutrient contribution a survey of consumption patterns
was begun in October. One half of the targeted farmers have been
surveyed; the rest will be completed when the researcher gets a
medical clearance to continue work in Sitiung V.
4513. Farmer Practice Survey. In November a 20% random sample of
farmers in Sitiung I Blok D was surveyed as to their methods of
soil preparation, fertilizer and pesticide use, use of crop
residue, constraints to production, etc. This intensive survey
will be duplicated in Sitiung V Blok C in order to obtain more
accurate information regarding these practices and to determine
--if any differences exist between longer term (Sit ID) and shorter
term (Sit VC) transmigrant farmers.
4602. Time Allocation. Data collection has been completed.