- Email: [email protected]
Effect of differently managed legumes on the early development of oil palms (Elaeis guineensis Jacq.)
Agro-Ecosystems, 6 (1981) 315--323
315
Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands
E F F E C T OF D I F F E R E N T L Y MANAGED LEGUMES ON THE E A R L Y DEVELOPMENT OF OIL PALMS (ELAEIS GUINEENSIS JACQ.)
P. AGAMUTHU 1, Y.K. CHAN 2, R. JESINGER s, K.M. KHOO 4 and W.J. BROUGHTON ''s
'Department of Genetics and Cellular Biology, University of Malaya, Kuala Lumpur (Malaysia) 2Department of Agriculture, Jalan Swettenham, Kuala Lumpur (Malaysia) 3Present address: Rohm and Haas Company, 513 November Drive, Durham, North Carolina 27712 (U.S.A.) 4Sime Darby Agronomic Advisory Unit, Seafield Estate, Batu Tiga, Selangor (Malaysia) sPresent address (and author for correspondence): Max-Planck-lnstitut f~r Z~chtungsforschung, D-5000 Kbln 30 (Federal Republic of Germany) (Accepted 30 July 1980)
ABSTRACT
Agamuthu, P., Chan, Y.K., Jesinger, R., Khoo, K.M. and Broughton, W.J., 1981. Effect of differently managed legumes on the early development of oil palms (Elaeis guineensis Jacq.), Agro-Ecosystems, 6: 315--323. We studied the effects of different covers (none, legume and natural) established in different ways (with legumes only: hand-weeded, hand-weeded plus extra fertilizer, and with pre-emergent application of the herbicide Oxyfluorfen at 0.25 kg active ingredient ha -~ ), on the first 3.5 years' growth and yield of oil palms (Elaeis guineensis Jacq). Legumes were a mixture of Centrosema pubescens and Pueraria phaseoloides. Natural generation consisted mostly of Nephrolepis biserrata and Paspalum conjugatum. Early rates of growth of the palms (as measured by frond area, girth, height, etc.) were greatest in the bare plots and the first yields of fresh fruit bunches were significantly higher than with any other treatment. On the other hand, relative growth rates and net assimilation rates were higher in legume plots (particularly those established with Oxyfluorfen) throughout most of the experimental period. This means that the rate of growth of palms in legume treated plots steadily overtook the rate of growth of those in bare plots. Presumably, these differences resulted from rapid exploitation of soil nutrients by palms in bare plots, and the "banking" of these same nutrients in legume plots.
INTRODUCTION
Establishment of leguminous cover crops under oil palm (Elaeis guineensis Jacq.) and rubber (Hevea brasiliensis Muell. Arg.) is an integral part of plantation practice in Malaysia. Covers like Calopogonium caeruleum, Ca. muconoides, Centrosema pubescens, and Pueraria phaseoloides have been widely tested and their effects on soil fertility and yield of the main crops
0304-3746/81/0000--0000/$02.50 © 1981 Elsevier Scientific Publishing Company
316 well d o c u m e n t e d (Broughton, 1976, 1977; Corley, Hardon and Wood, 1976). There is a certain amount of contention over the best way of establishing legumes, and their effects on the early, vegetative growth of the plantation crop. In an earlier communication, we showed that covers could be cheaply and effectively established by using the pre-emergence herbicide "Oxyfiuorfen" {"Goal") (Agamuthu et al., 1980). The present communication reports on the effect of " O x y f l u o r f e n " established legumes on the early growth and yield of oil palms in relation to those established by the normal plantation practice of hand-weeding. MATERIALS AND METHODS
Experimental design Eight different treatments, each replicated four times were used as follows; " O x y f l u o r f e n " and "Nitrofluorfen" at two different levels were used as pre-emergence herbicides to establish C. pubescens and P. phaseoloides; mixtures of C. pubescens and P. phaseoloides which were manually kept free of weeds (the "Control"); a similar set in which the experimental palms received twice the normal plantation dosage of fertilizer ("Fertilizer" plots); a set in which natural regeneration was allowed (the "Natural" plots, dominated by Nephrolepis biserrata and Paspalum conjugatum) and a set in which the ground surface was kept completely free of legumes and weeds ("Bare"). Each plot was 0.17 ha, and carried 25 palms, of which the central nine were used for measurement. One-year old palms were planted in the experimental area in October 1975. Sowing of the legumes and pre-emergence herbicide application began two months later. Further details are given in Agamuthu {1978), Agamuthu et al. {1980), Agamuthu and Broughton (1981).
Palm analysis Measurements of the palms included the rate of frond and leaflet production, the individual area of fronds (as calculated b y Corley et al., 1976), the total area of fronds on an individual palm, and the girth of the palm (measured at ground level). First measurements were made one year after planting the palms, and thereafter at six-monthly intervals. Relative growth rates and net assimilation rates were calculated on frond no. 9 according to Sestak, Catsky and Jarvis (1971). Yield (in terms of bunches of fresh fruit) was recorded for six weeks, three and a half years after planting (flowering began six months after field planting, but following standard agronomic practices in plantations the inflorescences were removed until the palms were 3 years old).
Chemical analysis Twelve leaflets from the middle of frond no. 9 were collected from each
317
of the nine central palms per plot and used for the determination of N, P, K, Ca, Mg and Mn. The mid-rib was removed from the central 20 cm portion of the leaflets and was discarded. The remaining leaf tissue was cut into ca. 1 cm pieces, oven dried at 70 ° C for 48 h, ground by a laboratory mill and the powder passed through a 1 mm sieve. After mixing thoroughly, this material was used for mineral analysis as described by Agamuthu et al.
(1980). RESULTS
Agamuthu et al. (1980)listed eight different treatments in their first report of these trials. Four of them dealt with legume establishment using different herbicides. As this communication is concerned with the effect of different covers on the early growth of oil palms, only the best herbicide treatment -- " O x y f l u o r f e n " -- applied at 0.25 kg active ingredient ha -1, will be considered. It is obvious from Fig. 1 that the various covers only slightly affected growth of the palms. Still, the difference in fruit yield attributable to legume covers is only about 10% (see Broughton, 1976), and some of the differences observed here are of this order. As an example, the number of fronds per palm was consistently 10 to 20% more in "Bare" plots than in those treated with " O x y f l u o r f e n " (Fig. 1). As a similar trend existed with individual frond area, the total size of the assimilatory apparatus was commensurately greater in "Bare" plots. Components of the increased frond area included both length and width, and resulted in increased frond weight. Considered in these terms, the plots in which the palms received twice the normal estate amount of fertilizer ("Fertilizer" plots), produced the next best growth. Other measures of palm growth give a similar perspective. Initially, the rate of increase in girth of palms, grown in "Bare" and "Fertilizer" plots exceeded those in the others, a difference that remained until the end of the experiment (Table I). Similar observations were also made with respect to the rate of frond production and the increase in height. TABLEI M e a n r a t e o f p a l m f r o n d p r o d u c t i o n , p a l m h e i g h t i n c r e a s e a n d p a l m girth i n c r e a s e o v e r the experimental p e r i o d ; all d a t a a r e s h o w n -+ s t a n d a r d e r r o r s
Treatment
Palm frond production ( m o n t h -1 )
Palm height increase ( c m m o n t h -1 )
P a l m girth increase ( c m m o n t h -~ )
Fertilizer Control (hand-weeded) Natural Bare O x y f l u o r f e n 0.25
2.64 2.55 2.53 2.78 2.57
13.15 10.23 11.43 12.39 11.71
6.80 6.12 6.04 6.82 6.06
+- 0 . 0 6 -+ 0.14 -+ 0.07 -+ 0 . 1 3 ± 0.05
± 1.15 -+ 1.28 -+ 1.12 +- 1.08 -+ 0.82
-+ 0.61 -+ 0.39 -+ 0 . 5 0 + 0.36 -+ 0.59
318
80.
2/-0
E "~
li0.
o
~0
f
2O
e3 . ~
..¢p 200
~
160
120 7oo t
I
I 300
O
_=o
200
~d
I./. w
v ,.c o) c o
1.2
3.
7
~.o U.P
o LL 1
I
I
I
1./,
,'~
1.o
~
o~
I
I
0.8
• Fertiliser =Control A Natural o Bare o OxyfLu0rfen 0.25
0.2 1.5
20
2.5
TIME
(years
1.0
after
i
i
i
1.5
2.0
2.5
planting}
Fig. 1. Effect of various types of covers (and different ways of establishing t h e m ) on the early g r o w t h of oil palms. The data for a specific frond refer to frond No. 9. All data represent measurements m a d e on the central nine palms of each plot, and were averaged over the f o u r replicates.
319
A clearer picture emerges if these data are taken together, and considered in plant growth terms (Table II). Initially, palms from " C o n t r o l " and "Oxyfluorfen" treated plots had higher relative growth rates (R) than those in the other plots. This trend changed with time, but the growth rate of palms from " O x y f l u o r f e n " treated plots was still higher than in any other treatment at the end of the experiment. Similar trends were observed with the net assimilation rate (E), though initially the differences were not so great. Final values of E were lowest in palms from " C o n t r o l " and " O x y f l u o r f e n " plots, something that stemmed directly from their leaf area ratios (F). Whilst there was also an ontogenetic drift of F with time, these especially high values in palms from both treatments suggest a different physiology (thicker leaves) that may well explain yield increases observed in older palms (see below). A direct cause of the changing rates of development seems to be the levels of the various nutrients in the fronds. One year after planting, palms from every treatment had higher nitrogen levels in their fronds than those from " O x y f l u o r f e n " treated plots (Fig. 2). At the end of the experiment, however, this difference was completely reversed so that palms from "Oxyfluorfen" treated plots had significantly higher nitrogen levels (P >1 0.05) than the next best treatment. This was also true of phosphorus (P = 0.05), but not of the other elements measured. Potassium levels in palms of all three legume plots ("Fertilizer", "Control", and " O x y f l u o r f e n " ) were higher than those in palms from the other treatments. In this respect, palms from "Oxyfluorfen" treated plots had the second highest potassium levels. Surprisingly, palms from "Bare" plots had significantly higher (P = 0.01) levels of manganese in their fronds than palms from any other treatment. Increased levels of nitrogen, phosphorus, and potassium in palms from legume plots (particularly those treated with " O x y f l u o r f e n " ) were evident two years after planting. Initial fruit yields (Table III), displayed a pattern similar to that of palm growth. The number of bunches harvested was far higher in "Bare" plots, and lower in "Natural" plots. Yields did n o t vary greatly in the legume plots, especially when the other herbicide treatments were included in the comparison (data not shown). Both components of yield were affected by the method (or type) of cover establishment. Mean weight of the bunches was greatest in "Fertilizer" plots and lowest in "Natural" plots. Palms in the "Bare" plots had the next highest mean bunch weight, and so the extraordinary yields obtained in these plots were a combination of the large number and weight of the bunches. Similarly the poor performance of palms in the "Natural" plots was a combination of a low number of light bunches. Palms in all the legume plots yielded a more consistent number of fresh fruit bunches of more uniform weight.
320
0 0 0 0 0 0
0 0 0 0 0
~o
qqqqq 0 0 0 0 0
.= o ~.E
0 0 0 0 0
~
~.N
"o ~ e 0 0 0 0 0
~.~
~
0
0
" ~ o
o~
~s4 <
~.~
.o
321
TIME
(years
aft¢r
planting)
-i&l
~
,
%N I
0 0.3 0.2 % P 01 0 3 O.
%K
--] T T I - - - 1-]-
2 1
o~
0
~o 3 0.
E
%Ca
o (.o
2 I
r--T-r
0
m
]---1 r - -
~
3 Z %M
2 T 0
- -
--r--r~-r~
m~l--1-- l-
5 4 ppm .in (xl0 2 ) 3 2 I
0
Fcrti[iser
Control
Natural
Bare
Oxyfluorfcn 0.25
TREATMENT
Fig. 2. Mineral contents of frond No. 9 of palms grown in association with the different types of cover (and different ways of establishing them) at various times after establishment. All data represent measurements made on the central nine palms of each plot and were averaged over the four replicates.
322 TABLE III Early yield of palms established with different ground covers. Data were recorded over a six week period beginning three and a half years after planting the palms Treatment
No. fresh fruit bunches (month -1)
Fertilizer Control Natural Bare Oxyfluorfen 0.25
382 436 310 594 373
Weight of fresh fruit bunches
Yield (fresh fruit bunches)
(kg -I)
(ha -1 month =1)
7.6 6.6 5.6 7.1 6.0
2899 2859 1741" 4224* 2242
*Significantly different (P = 0.05) from the mean.
DISCUSSION
Over the last t w e n t y years, commercial organizations operating in peninsular Malaysia have documented the effects of various types of covers on the growth of oil palms (see Broughton, 1976). The benefits of this include: reduced soil erosion (and improved water penetration), increased soil organic matter contents, improved soil structure (and therefore aeration), reduced soil temperatures, and reduced competition between the main crop and weeds. All these combined lead to increased levels of nutrients in the palm fronds, faster rates of growth of the palms, and ultimately a yield increase of about 2 metric tonnes fresh fruit bunches ha-ly -' . In all these experiments, data collection usually began in the year after planting, and was intense only following harvesting of the fresh fruit bunches. One of our principal aims, therefore, was to document the effects of covers on the early phases in the establishment of oil palm growth. Whilst soil t y p e has a large influence on cover effects (Broughton, 1976), perhaps the most significant finding of this communication is that the benefits of well-managed legumes are not immediately available to the oil palms. Indeed, by several different measures, palms in the "Bare" plots had the fastest initial rate of growth. Whilst these differences diminished with time, early yields of fresh fruit bunches were significantly higher in these plots. If one considers relative growth rates and net assimilation rates as an index of later palm performance, there can be little d o u b t that palms from the legume plots would have quickly out-yielded those from "Bare" plots. These differences in early performance must, therefore, reflect varying abilities to exploit available soil nutrients. Broughton's 1977 hypothesis suggests that covers bank some of the soil nutrients, slowly releasing them again as competition with the palms for water and light gradually diminishes their vigour. By this time, the palms have an extensive root system fully capable of absorbing re-released nutrients. Absence of a cover would mean that more
323
nutrients were intially available to the palm, but leaching and run-off losses would quickly reduce this. Quite obviously, palms in the "Bare" plots absorbed some of these nutrients and, therefore, grew faster. As the rate of exploitation exceeded the supply, however, growth in comparison with that in legume plots decreased. Taken as a whole, these data suggest that establishment of palms, even in legume plots, could be facilitated by fertilizer application rates above those normally used (cf. "Fertilizer" plots). ACKNOWLEDGEMENTS
We wish to thank Han Siew King, and the staff of the Sime Darby Agronomic Advisory Unit, T.R.D. Eakin of Seafield Estate, J.I.D.R. Furtado, S.S. Dhaliwal, S. Jodhy and A. Kokilam of the University of Malaya for their help with many aspects of this work. Financial assistance was provided by Sime Darby Plantations Bhd., Kuala Lumpur, Rohm and Haas Asia Inc., Singapore, and the University of Malaya.
REFERENCES Agamuthu, P., 1978. Factors Affecting the Development of Oil Palm (Elaeis guineensis Jacq.) seedlings. M.Sc. Thesis, University of Malaya, Kuala Lumpur, 202 pp. Agamuthu, P. and Broughton, W.J., 1981. Nutrient recycling between covers and oil palms in developing plantations. Agro-Ecosystems, in preparation. Agamuthu, P., Chan, Y.K., Jesinger, R., Khoo, K.M. and Broughton, W.J., 1980. Effect of diphenyl ether pre-emergence herbicides on legume cover establishment under oil palm (Elaeis guineensis Jacq.). Agro-Ecosystems, 6: 193--208. Broughton, W.J., 1976. Effect of various covers on the performance of Elaeis guineensis (Jacq.) on different soils. In: D.A. Earp and W. Newall (Editors), International Oil Palm Developments, Inc. Soc. of Planters, Kuala Lumpur, pp. 501--525. Broughton, W.J., 1977. Effect of various covers on soil fertility under Hevea brasiliensis (Muell. Arg.) and on growth of the tree. Agro-Ecosystems, 3: 147--170. Corley, R.H.V., Hardon, J.J. and Wood, B.J., 1976. Oil Palm Research. Elsevier, Amsterdam, 532 pp. Sestak, Z., Catsky, J. and Jarvis, P.C., 1971. Plant Photosynthetic Production. Manual of Methods. Dr. W. Junk, den Haag, pp. 343--391.
315
Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands
E F F E C T OF D I F F E R E N T L Y MANAGED LEGUMES ON THE E A R L Y DEVELOPMENT OF OIL PALMS (ELAEIS GUINEENSIS JACQ.)
P. AGAMUTHU 1, Y.K. CHAN 2, R. JESINGER s, K.M. KHOO 4 and W.J. BROUGHTON ''s
'Department of Genetics and Cellular Biology, University of Malaya, Kuala Lumpur (Malaysia) 2Department of Agriculture, Jalan Swettenham, Kuala Lumpur (Malaysia) 3Present address: Rohm and Haas Company, 513 November Drive, Durham, North Carolina 27712 (U.S.A.) 4Sime Darby Agronomic Advisory Unit, Seafield Estate, Batu Tiga, Selangor (Malaysia) sPresent address (and author for correspondence): Max-Planck-lnstitut f~r Z~chtungsforschung, D-5000 Kbln 30 (Federal Republic of Germany) (Accepted 30 July 1980)
ABSTRACT
Agamuthu, P., Chan, Y.K., Jesinger, R., Khoo, K.M. and Broughton, W.J., 1981. Effect of differently managed legumes on the early development of oil palms (Elaeis guineensis Jacq.), Agro-Ecosystems, 6: 315--323. We studied the effects of different covers (none, legume and natural) established in different ways (with legumes only: hand-weeded, hand-weeded plus extra fertilizer, and with pre-emergent application of the herbicide Oxyfluorfen at 0.25 kg active ingredient ha -~ ), on the first 3.5 years' growth and yield of oil palms (Elaeis guineensis Jacq). Legumes were a mixture of Centrosema pubescens and Pueraria phaseoloides. Natural generation consisted mostly of Nephrolepis biserrata and Paspalum conjugatum. Early rates of growth of the palms (as measured by frond area, girth, height, etc.) were greatest in the bare plots and the first yields of fresh fruit bunches were significantly higher than with any other treatment. On the other hand, relative growth rates and net assimilation rates were higher in legume plots (particularly those established with Oxyfluorfen) throughout most of the experimental period. This means that the rate of growth of palms in legume treated plots steadily overtook the rate of growth of those in bare plots. Presumably, these differences resulted from rapid exploitation of soil nutrients by palms in bare plots, and the "banking" of these same nutrients in legume plots.
INTRODUCTION
Establishment of leguminous cover crops under oil palm (Elaeis guineensis Jacq.) and rubber (Hevea brasiliensis Muell. Arg.) is an integral part of plantation practice in Malaysia. Covers like Calopogonium caeruleum, Ca. muconoides, Centrosema pubescens, and Pueraria phaseoloides have been widely tested and their effects on soil fertility and yield of the main crops
0304-3746/81/0000--0000/$02.50 © 1981 Elsevier Scientific Publishing Company
316 well d o c u m e n t e d (Broughton, 1976, 1977; Corley, Hardon and Wood, 1976). There is a certain amount of contention over the best way of establishing legumes, and their effects on the early, vegetative growth of the plantation crop. In an earlier communication, we showed that covers could be cheaply and effectively established by using the pre-emergence herbicide "Oxyfiuorfen" {"Goal") (Agamuthu et al., 1980). The present communication reports on the effect of " O x y f l u o r f e n " established legumes on the early growth and yield of oil palms in relation to those established by the normal plantation practice of hand-weeding. MATERIALS AND METHODS
Experimental design Eight different treatments, each replicated four times were used as follows; " O x y f l u o r f e n " and "Nitrofluorfen" at two different levels were used as pre-emergence herbicides to establish C. pubescens and P. phaseoloides; mixtures of C. pubescens and P. phaseoloides which were manually kept free of weeds (the "Control"); a similar set in which the experimental palms received twice the normal plantation dosage of fertilizer ("Fertilizer" plots); a set in which natural regeneration was allowed (the "Natural" plots, dominated by Nephrolepis biserrata and Paspalum conjugatum) and a set in which the ground surface was kept completely free of legumes and weeds ("Bare"). Each plot was 0.17 ha, and carried 25 palms, of which the central nine were used for measurement. One-year old palms were planted in the experimental area in October 1975. Sowing of the legumes and pre-emergence herbicide application began two months later. Further details are given in Agamuthu {1978), Agamuthu et al. {1980), Agamuthu and Broughton (1981).
Palm analysis Measurements of the palms included the rate of frond and leaflet production, the individual area of fronds (as calculated b y Corley et al., 1976), the total area of fronds on an individual palm, and the girth of the palm (measured at ground level). First measurements were made one year after planting the palms, and thereafter at six-monthly intervals. Relative growth rates and net assimilation rates were calculated on frond no. 9 according to Sestak, Catsky and Jarvis (1971). Yield (in terms of bunches of fresh fruit) was recorded for six weeks, three and a half years after planting (flowering began six months after field planting, but following standard agronomic practices in plantations the inflorescences were removed until the palms were 3 years old).
Chemical analysis Twelve leaflets from the middle of frond no. 9 were collected from each
317
of the nine central palms per plot and used for the determination of N, P, K, Ca, Mg and Mn. The mid-rib was removed from the central 20 cm portion of the leaflets and was discarded. The remaining leaf tissue was cut into ca. 1 cm pieces, oven dried at 70 ° C for 48 h, ground by a laboratory mill and the powder passed through a 1 mm sieve. After mixing thoroughly, this material was used for mineral analysis as described by Agamuthu et al.
(1980). RESULTS
Agamuthu et al. (1980)listed eight different treatments in their first report of these trials. Four of them dealt with legume establishment using different herbicides. As this communication is concerned with the effect of different covers on the early growth of oil palms, only the best herbicide treatment -- " O x y f l u o r f e n " -- applied at 0.25 kg active ingredient ha -1, will be considered. It is obvious from Fig. 1 that the various covers only slightly affected growth of the palms. Still, the difference in fruit yield attributable to legume covers is only about 10% (see Broughton, 1976), and some of the differences observed here are of this order. As an example, the number of fronds per palm was consistently 10 to 20% more in "Bare" plots than in those treated with " O x y f l u o r f e n " (Fig. 1). As a similar trend existed with individual frond area, the total size of the assimilatory apparatus was commensurately greater in "Bare" plots. Components of the increased frond area included both length and width, and resulted in increased frond weight. Considered in these terms, the plots in which the palms received twice the normal estate amount of fertilizer ("Fertilizer" plots), produced the next best growth. Other measures of palm growth give a similar perspective. Initially, the rate of increase in girth of palms, grown in "Bare" and "Fertilizer" plots exceeded those in the others, a difference that remained until the end of the experiment (Table I). Similar observations were also made with respect to the rate of frond production and the increase in height. TABLEI M e a n r a t e o f p a l m f r o n d p r o d u c t i o n , p a l m h e i g h t i n c r e a s e a n d p a l m girth i n c r e a s e o v e r the experimental p e r i o d ; all d a t a a r e s h o w n -+ s t a n d a r d e r r o r s
Treatment
Palm frond production ( m o n t h -1 )
Palm height increase ( c m m o n t h -1 )
P a l m girth increase ( c m m o n t h -~ )
Fertilizer Control (hand-weeded) Natural Bare O x y f l u o r f e n 0.25
2.64 2.55 2.53 2.78 2.57
13.15 10.23 11.43 12.39 11.71
6.80 6.12 6.04 6.82 6.06
+- 0 . 0 6 -+ 0.14 -+ 0.07 -+ 0 . 1 3 ± 0.05
± 1.15 -+ 1.28 -+ 1.12 +- 1.08 -+ 0.82
-+ 0.61 -+ 0.39 -+ 0 . 5 0 + 0.36 -+ 0.59
318
80.
2/-0
E "~
li0.
o
~0
f
2O
e3 . ~
..¢p 200
~
160
120 7oo t
I
I 300
O
_=o
200
~d
I./. w
v ,.c o) c o
1.2
3.
7
~.o U.P
o LL 1
I
I
I
1./,
,'~
1.o
~
o~
I
I
0.8
• Fertiliser =Control A Natural o Bare o OxyfLu0rfen 0.25
0.2 1.5
20
2.5
TIME
(years
1.0
after
i
i
i
1.5
2.0
2.5
planting}
Fig. 1. Effect of various types of covers (and different ways of establishing t h e m ) on the early g r o w t h of oil palms. The data for a specific frond refer to frond No. 9. All data represent measurements m a d e on the central nine palms of each plot, and were averaged over the f o u r replicates.
319
A clearer picture emerges if these data are taken together, and considered in plant growth terms (Table II). Initially, palms from " C o n t r o l " and "Oxyfluorfen" treated plots had higher relative growth rates (R) than those in the other plots. This trend changed with time, but the growth rate of palms from " O x y f l u o r f e n " treated plots was still higher than in any other treatment at the end of the experiment. Similar trends were observed with the net assimilation rate (E), though initially the differences were not so great. Final values of E were lowest in palms from " C o n t r o l " and " O x y f l u o r f e n " plots, something that stemmed directly from their leaf area ratios (F). Whilst there was also an ontogenetic drift of F with time, these especially high values in palms from both treatments suggest a different physiology (thicker leaves) that may well explain yield increases observed in older palms (see below). A direct cause of the changing rates of development seems to be the levels of the various nutrients in the fronds. One year after planting, palms from every treatment had higher nitrogen levels in their fronds than those from " O x y f l u o r f e n " treated plots (Fig. 2). At the end of the experiment, however, this difference was completely reversed so that palms from "Oxyfluorfen" treated plots had significantly higher nitrogen levels (P >1 0.05) than the next best treatment. This was also true of phosphorus (P = 0.05), but not of the other elements measured. Potassium levels in palms of all three legume plots ("Fertilizer", "Control", and " O x y f l u o r f e n " ) were higher than those in palms from the other treatments. In this respect, palms from "Oxyfluorfen" treated plots had the second highest potassium levels. Surprisingly, palms from "Bare" plots had significantly higher (P = 0.01) levels of manganese in their fronds than palms from any other treatment. Increased levels of nitrogen, phosphorus, and potassium in palms from legume plots (particularly those treated with " O x y f l u o r f e n " ) were evident two years after planting. Initial fruit yields (Table III), displayed a pattern similar to that of palm growth. The number of bunches harvested was far higher in "Bare" plots, and lower in "Natural" plots. Yields did n o t vary greatly in the legume plots, especially when the other herbicide treatments were included in the comparison (data not shown). Both components of yield were affected by the method (or type) of cover establishment. Mean weight of the bunches was greatest in "Fertilizer" plots and lowest in "Natural" plots. Palms in the "Bare" plots had the next highest mean bunch weight, and so the extraordinary yields obtained in these plots were a combination of the large number and weight of the bunches. Similarly the poor performance of palms in the "Natural" plots was a combination of a low number of light bunches. Palms in all the legume plots yielded a more consistent number of fresh fruit bunches of more uniform weight.
320
0 0 0 0 0 0
0 0 0 0 0
~o
qqqqq 0 0 0 0 0
.= o ~.E
0 0 0 0 0
~
~.N
"o ~ e 0 0 0 0 0
~.~
~
0
0
" ~ o
o~
~s4 <
~.~
.o
321
TIME
(years
aft¢r
planting)
-i&l
~
,
%N I
0 0.3 0.2 % P 01 0 3 O.
%K
--] T T I - - - 1-]-
2 1
o~
0
~o 3 0.
E
%Ca
o (.o
2 I
r--T-r
0
m
]---1 r - -
~
3 Z %M
2 T 0
- -
--r--r~-r~
m~l--1-- l-
5 4 ppm .in (xl0 2 ) 3 2 I
0
Fcrti[iser
Control
Natural
Bare
Oxyfluorfcn 0.25
TREATMENT
Fig. 2. Mineral contents of frond No. 9 of palms grown in association with the different types of cover (and different ways of establishing them) at various times after establishment. All data represent measurements made on the central nine palms of each plot and were averaged over the four replicates.
322 TABLE III Early yield of palms established with different ground covers. Data were recorded over a six week period beginning three and a half years after planting the palms Treatment
No. fresh fruit bunches (month -1)
Fertilizer Control Natural Bare Oxyfluorfen 0.25
382 436 310 594 373
Weight of fresh fruit bunches
Yield (fresh fruit bunches)
(kg -I)
(ha -1 month =1)
7.6 6.6 5.6 7.1 6.0
2899 2859 1741" 4224* 2242
*Significantly different (P = 0.05) from the mean.
DISCUSSION
Over the last t w e n t y years, commercial organizations operating in peninsular Malaysia have documented the effects of various types of covers on the growth of oil palms (see Broughton, 1976). The benefits of this include: reduced soil erosion (and improved water penetration), increased soil organic matter contents, improved soil structure (and therefore aeration), reduced soil temperatures, and reduced competition between the main crop and weeds. All these combined lead to increased levels of nutrients in the palm fronds, faster rates of growth of the palms, and ultimately a yield increase of about 2 metric tonnes fresh fruit bunches ha-ly -' . In all these experiments, data collection usually began in the year after planting, and was intense only following harvesting of the fresh fruit bunches. One of our principal aims, therefore, was to document the effects of covers on the early phases in the establishment of oil palm growth. Whilst soil t y p e has a large influence on cover effects (Broughton, 1976), perhaps the most significant finding of this communication is that the benefits of well-managed legumes are not immediately available to the oil palms. Indeed, by several different measures, palms in the "Bare" plots had the fastest initial rate of growth. Whilst these differences diminished with time, early yields of fresh fruit bunches were significantly higher in these plots. If one considers relative growth rates and net assimilation rates as an index of later palm performance, there can be little d o u b t that palms from the legume plots would have quickly out-yielded those from "Bare" plots. These differences in early performance must, therefore, reflect varying abilities to exploit available soil nutrients. Broughton's 1977 hypothesis suggests that covers bank some of the soil nutrients, slowly releasing them again as competition with the palms for water and light gradually diminishes their vigour. By this time, the palms have an extensive root system fully capable of absorbing re-released nutrients. Absence of a cover would mean that more
323
nutrients were intially available to the palm, but leaching and run-off losses would quickly reduce this. Quite obviously, palms in the "Bare" plots absorbed some of these nutrients and, therefore, grew faster. As the rate of exploitation exceeded the supply, however, growth in comparison with that in legume plots decreased. Taken as a whole, these data suggest that establishment of palms, even in legume plots, could be facilitated by fertilizer application rates above those normally used (cf. "Fertilizer" plots). ACKNOWLEDGEMENTS
We wish to thank Han Siew King, and the staff of the Sime Darby Agronomic Advisory Unit, T.R.D. Eakin of Seafield Estate, J.I.D.R. Furtado, S.S. Dhaliwal, S. Jodhy and A. Kokilam of the University of Malaya for their help with many aspects of this work. Financial assistance was provided by Sime Darby Plantations Bhd., Kuala Lumpur, Rohm and Haas Asia Inc., Singapore, and the University of Malaya.
REFERENCES Agamuthu, P., 1978. Factors Affecting the Development of Oil Palm (Elaeis guineensis Jacq.) seedlings. M.Sc. Thesis, University of Malaya, Kuala Lumpur, 202 pp. Agamuthu, P. and Broughton, W.J., 1981. Nutrient recycling between covers and oil palms in developing plantations. Agro-Ecosystems, in preparation. Agamuthu, P., Chan, Y.K., Jesinger, R., Khoo, K.M. and Broughton, W.J., 1980. Effect of diphenyl ether pre-emergence herbicides on legume cover establishment under oil palm (Elaeis guineensis Jacq.). Agro-Ecosystems, 6: 193--208. Broughton, W.J., 1976. Effect of various covers on the performance of Elaeis guineensis (Jacq.) on different soils. In: D.A. Earp and W. Newall (Editors), International Oil Palm Developments, Inc. Soc. of Planters, Kuala Lumpur, pp. 501--525. Broughton, W.J., 1977. Effect of various covers on soil fertility under Hevea brasiliensis (Muell. Arg.) and on growth of the tree. Agro-Ecosystems, 3: 147--170. Corley, R.H.V., Hardon, J.J. and Wood, B.J., 1976. Oil Palm Research. Elsevier, Amsterdam, 532 pp. Sestak, Z., Catsky, J. and Jarvis, P.C., 1971. Plant Photosynthetic Production. Manual of Methods. Dr. W. Junk, den Haag, pp. 343--391.