Tải bản đầy đủ - 0 (trang)
V. Management of Tree Legumes

V. Management of Tree Legumes

Tải bản đầy đủ - 0trang



experiments investigating the effect of cutting height and/or cutting frequency on Leucaena yield are summarized in Table IV.

A major problem with many research reports is the use of incorrect or

ill-defined yield terminology (e.g., Mendoza and Javier, 1980). Many reports are of limited utility because of a failure to accurately define “herbage” without making it clear whether this refers to edible material only or

to total biomass yield. Most of the yield figures in Table IV conform with

the assertion of Blom (1980) that annual edible dry matter production from

Leucaena will generally range form 6 to 18 t DM ha-’ year-’. Comparison

of yields and the effects of cutting height and frequency on Leucaena is

made difficult because of the different cultivars used by the various researchers. The findings of cutting management experiments will be dependent on whether arboreal (“Salvador”), shrubby (“Hawaiian”), or lowbranching (“Peru”) cultivars have been used. Although the Salvador types

with high leaf: stem ratios have often outyielded the Hawaiian and Peru

types, the results have been confounded by the method of harvesting

Table IV

Cutting Height and Cutting Frequency Effects on Yield of LeucaendPb

Cutting height












( 120)

( 150) (60)







30/60/90 nsp







30-90 ns



42-84 ns






(t ha-’ year-‘)


50.6 FW total

28.7 FW total

39.4 FW total

28.0 FW total

113.0 FW total

13.0 DW leaf

9.0 DW leaf

8.2 DW leaf

5.4 DW leaf

14.6 DW total

12.5 DW total

I 1 .O DW leaf

14.2 DW leaf

15.2 DW leaf

4.0 DW leaf

Takahashi and Ripperton (1949)r

Castillo el al. (1979)

Krishnamurthy and Munegowda (1982a)

Krishnamurthy and Munegowda (l982b)

Siregar (1983)

Guevarra et al. (1978)

Ferraris (1979)

Savory and Breen (1979)

Pathak et a / . (1980)

Perez and Melendez (1980)

Osman (1981a.b)

Semali et al. (1983)

Topark-Ngarm (1983)

Evensen (1984)

Isarasanee et al. (1985)

“ After Horne et al. (1986).

Figures not in parentheses are the best height/frequency used. Figures in parentheses are

the only heightlfrequency used.

Cited by Evensen (1984).

Maximum height or longest frequency used.

No significant differences.



(Evensen, 1984). Results from experiments investigating effects of cutting

height and cutting frequency are often conflicting and will be considered

separately here to clarify general trends emerging from the literature.



Little information exists on the effect of age of tree legumes at first

cutting on their productivity and persistence. Bhumibhaman er al. (1984)

assessed the coppicing of 1-year-old Leucaena leucocephala trees and

concluded that the amount of regrowth was positively related to stump

size. Similar results were reported by Pathak and Patil(l983). Ella (1987)

reported an increase in both leaf and wood production in L . leucocephala,

C. calothyrsus, S . grandiflora, and G . sepium as age to first cutting increased from 13 to 21 months (Table V).

Subsequent yields recorded over a 12-month period revealed that the

yields of S . grandiflora and C . calofhyrsus were not related to age of first

cutting, whereas in L . leucocephala and G . sepium, the highest yields were

recorded from the trees that were first cut at 21 months.

These results have prompted research to investigate the role of fastgrowing, weakly perennial species in supplying forage to animals (Bray er

al., 1989). These studies have recorded high leaf yields from Cajanus

cajan, Codariocalyx gyroides, and Sesbania sesban at sites in Indonesia

and Australia. Planting of these species would allow farmers to leave the

more perennial tree species uncut for a longer period and hence increase

overall productivity.

Table V

Leaf and Wood Yield (mg DM Tree-’) of Tree

Legumes“ Cut to 100 cm at Various Ages

from Sowing

Age (months)








0.35 a*”

0.40 a

0.71 b

1.35 c

2.11 d

0.47 a

0.60 a

1.53 a

3.59 b

6.79 c

a Mean of L. leucocephala, C. calorhyrsus, S .

grandgora, and G. sepium.

Numbers followed by the same letter within a

column are not significantly different.




Recommendations from cutting height experiments in the literature deal

mostly with Leucaena and show wide variability, ranging from extremes

as low as 5 cm up to 3 m. Takahashi and Ripperton (1949, cited by

Evensen, 1984), in an early investigation of the effect of cutting height on

Leucaena, compared total production when trees were cut at 5,38, or 76

cm and found that a cutting height of 5 cm gave maximum yields. By

contrast, other workers have often found that higher cutting heights give

better yields. Mendoza et al. (1975) found that yields of Leucaena cut at

300 cm were more thar double that of Leucaena cut at a height of 15 cm.

Unfortunately, they c id not include an intermediate range of cutting

heights in their experiment. Herrera (1967, cited by Evensen, 1984),

Pathak et al. (1980), and Perez and Melendez (1980) all found that the

highest yields were obtained from the highest cutting height under investigation, but as these were 75 cm or below, it might be that their experiments

did not cover a wide enough range to properly investigate this effect.

Horne and Blair (1991) found no significant difference in leaf or wood

yields when L . leucocephala cv. Cunningham was cut at 30 or 100 cm.

Krishnamurthy and Munegowda (1982a) found that 150 cm gave higher

yields than 15-cm and 75-cm cutting height treatments, and in an additional

comparison (Krishnamurthy and Munegowda, 1982b), the highest cutting

height treatment (150 cm) again resulted in the highest yields compared

with cutting heights of 60,90, or 120cm. Most other workers have reported

optimum cutting heights in the range of 90-120 cm (Osman, 1981c; Siregar,

1983; Isarasanee et al., 1985).

Lazier (1981) imposed three cutting frequency treatments (14-, 40- and

56-day intervals) with three heights of cutting ( 5 , 25, and 50 cm) on Codariocalyx gyroides and found that the 56-day cutting interval and the two

higher cutting heights gave the best yields (2.7 t leaf DM ha-' year-').

Siregar (1983) imposed a range of cutting height treatments (up to 150 cm)

on Calliandru calothyrsus and Flemingia congesta and reported that cutting height at 1 m gave the best yield.


Cutting frequency treatments imposed by researchers range from harvest intervals of 4 weeks (Perez and Melendez, 1980; Osman, 1981a) up to

16-17 weeks (Takahashi and Ripperton, 1949, cited by Evensen, 1984;

Ferraris, 1979). Most experiments have set fixed cutting intervals, but it

has been agreed that in areas with large variability in climatic conditions



(e.g., with distinct wet and dry seasons), optimum production is more

likely if trees are cut on the basis of stage of regrowth (Evensen, 1984).

Guevarra et al. (1978), Herrera (1967, cited by Evensen, 1984),Isarasanee

et al. (1985), and Ferraris (1979) reported experiments in which cutting

frequency treatments were imposed on the basis of attainment of some

predetermined height of regrowth. In a more recent experiment,

Catchpoole and Blair (1990a) found that leaf production in L . leucocephala, G . sepium, C calothyrsus, and S . grandifIora was unaffected by

cutting to a height of 100 cm when the trees reached 150,200,or 250 cm. By

contrast, stem production averaged over species increased from 10.16 to

21.45 t ha-' as attainment height was increased from 150 to 250 cm. Data

from cutting experiments that use predetermined stages of regrowth to

indicate the time of harvesting are more meaningful and more readily

applicable to other locations than if only fixed cutting intervals are used

(Hegde, 1983).

Results of cutting frequency experiments with Leucaena reported in the

literature vary considerably, and the variation in cutting heights makes it

difficult for any general conclusions to be drawn. For example, whereas

Takahashi and Ripperton (1949, cited by Evensen, 1984) and Ferraris

(1979) obtained the best yields from infrequently cut trees at low cutting

heights, Pathak et al. (1980) and Perez and Melendez (1980) found that

yields were maximized at high cutting heights with shorter cutting intervals.

The situation becomes a little clearer when the results from investigations reporting the effect of cutting frequency on total yield are considered

separately from studies on the effects of cutting interval on yields of leaf or

edible forage from Leucaena. Most workers investigating the effects of

cutting frequency on total Leucaena yields have generally found that

infrequent harvesting gives the highest total dry matter yields. Osman

(1981a) compared cutring intervals of 30, 60, 90,and 120 days, with the

90-day treatment resulting in the highest total yield. The higher total DM

production under infrequent cutting is generally attributed to the higher

proportion of the total yield made up of stem material (Osman, 1981b;

Horne et a / . , 1986).

Ferraris (1979) found that leaf made up 54% of the total Leucaena yield

after 60 days of regrowth, but only 31% after 120days. Similarly, Guevarra

el al. (1978) reported a decline in the edible fraction from 81% after 70 days

to 60% after 110 days of regrowth, and Topark-Ngarm (1983) from 69% at

40 days to 60% after 60 days. Thus it appears that the higher total yields of

Leucaena obtained with infrequent cutting are due to a relative increase in

the stem component (a lower leaf stem ratio) with lengthened harvest


However, it is difficult to draw general conclusions from experiments



investigating the effects of cutting frequency on leaf yields, as the results

are much less consistent. Pathak et al. (1980) and Das and Dalvi (1981)

found that the most frequent cutting treatments studied (40-day and 60-day

intervals, respectively) gave the best yields of Leucaena leaf. In contrast,

Semali et al. (1983) found that infrequent cutting (1 10 days) resulted in the

highest leaf yields. Topark-Ngarm (1983) obtained higher leaf yields with a

60-day harvest interval than from a more frequently cut 40-day regrowth

treatment. Evensen (1984) reported no significant differences in leaf yield

across cutting heights of 30,60, and 90 cm, with harvest intervals of 40,60,

and 120 days, despite the lower leaf fraction in the total yield at 120 days

(44%) compared to at 40 days (69%). These results of Evensen (1984) have

particular relevance to management recommendations for Leucaena production, as his findings imply that infrequently cut Leucaena can be used

as a source of fuelwood without adversely affecting forage yield.

However, such a practice is dependent on the nature of the farmer's tree

legume resource and forage requirements-a smallholder with a limited

number of trees will have to cut each tree more often to obtain a continuous supply of forage (Home et al., 1986).

Although the available literature on the effects of cutting frequency on

leaf dry matter yields is confusing, it is well known that the forage quality

of Leucaena leaf declines with age, and therefore material from infrequently cut trees with a higher proportion of older leaf has a lower nutritive

value than material harvested from frequently cut trees (Takahashi and

Ripperton, 1949, cited by Evensen, 1984;Tangendjaja et al., 1986). Semali

et al. (1983) showed that calcium, phosphorus, and nitrogen levels in leaf

declined as cutting interval was lengthened beyond 60 days. An exception

is that of Evensen (1984), who found no significant effect of cutting frequency on crude protein content, although this could be a reflection of the

highly fertile soil on which his experiments were conducted.

There is very little information concerning the effects of cutting height

and frequency on yields of tree legume species other than Leucaena. Gore

and Joshi (1976, cited by Evans and Rotar, 1987a) obtained the highest

yields of N,P,K-fertilized Sesbania sesban with infrequent cutting (every

9-10 weeks) than with frequent cutting (5-6 weeks), over a 31-week

growing period (14.2 t leaf DM ha-').


The variation in tree spacing and planting configuration between published investigations into Leucaena cutting management strategy is a further confounding factor in the interpretation of the effects of cutting height

and frequency on yield. The density employed in cutting height and fre-



quency experiments is usually determined with a view to the intended

practical application. Rarely has equidistant spacing been used, but rather

Leucaena has been planted in rows to simulate hedgerow or alley cropping

management systems and, as such, density expressed as number of trees

per unit area is less meaningful than specifications of inter- and intrarow

tree spacing. Generally most researchers have found that Leucaena

planted at higher densities gives better yields than at lower densities.

Pathak et al. (1980) reported higher leaf dry matter yields (5.4 t ha-'

year-') from trees at a density of 40,000 trees ha-' than at a density of

15,000 trees ha-'. (Both of these density treatments were cut at a height of

30 cm every 40 days.) Castillo et al. (1979) compared four densities (3000,

5000, 60oO, and 10,000 trees ha-') and obtained significantly higher yields

from the two highest densities. Three densities (lO,OOO, 30,000, and 60,000

trees ha-') were compared by Savory and Breen (1979), who reported that

60,000 trees ha-' gave the highest yields. Van den Beldt (1983) reported

that wood yields were higher at wider spacings.

Ella (1987) studied tree legumes in mixtures with grasses in a distinctly

monsoonal climate in South Sulawesi, Indonesia, where soil moisture

deficits limit plant growth in the dry season and incident light is potentially

limiting in the wet season. Four tree legume species (Sesbania grandijlora,

Calliandra calothyrsus, Leucaena leucocephala, and Gliricidia sepium)

were grown at four densities (5000, 10,OOO, 20,000, and 40,000 trees ha-')

with an underplanted grass (Panicum maximum). Cutting intervals of 6

and 12 weeks were applied to the trees (cut at 1 m) and the grass (cut at

12 cm). Although there was some variability in results between species,

there was a general trend toward increased grass leaf dry matter yields and

reduced tree leaf dry matter yields with reduced tree density in the wet

season. In the dry season, however, grass yields were positively related to

tree density. Ella (1987) attributed this seasonal reversal of the tree density

effect on grass yields to competition for light, which inhibited grass growth

at the higher densities and longer cutting intervals, in the wet season and to

reduced direct sunlight and transpiration benefiting grass growth in the

same treatments in the dry season.

A further confounding factor influencing the effect of tree density on

grass yields is nitrogen fixation. Kitamura (1986) investigated the longterm effects of varying tree densities and regrowth intervals on the productivity of low-cut Leucaena (30 cm) interplanted with tall tropical grasses

(Pennisetum purpureum and Panicum maximum). Total dry matter yields

of Leucaena were reduced to only 2 1% of monoculture yields by Penniseturn, compared with 5% in the experiment of Horne and Blair (1990). This

is probably because of the very high densities employed (10-cm intrarow

spacings and 15-,30-, 60-, and 100-cm interrow spacings). Kitamura (1986)

found that edible dry matter yields of both Leucaena and Penniserum were



maximized at the highest densities and longest growth intervals (attainment of 200-cm height by the Leucaena canopy). The explanation

given for this was substantial transfer of nitrogen from the Leucaena to the

grass in these treatments.

The effects of interrow and intrarow spacing on yield have been investigated by Guevarra et al. (1978), Ferraris (1979), and Savory and Breen

(1979). Ferraris (1979) found no differences between interrow spacings of

30 or 90 cm. Savory and Breen (1979) also found no significant differences

between Leucaena leaf yields of the two interrow treatments (60-cm and

100-cm spacing between rows), but there was a significant effect of intrarow spacing, with significantly higher yields (7 t leaf DW ha-’ year-’)

from trees 30 cm apart within a row compared to the 7.5- and 15-cm

intrarow treatments. This is in contrast to the findings of Guevarra et al.

(1978), who reported that wider intrarow spacing (45 cm) gave higher

yields than closer spacings (15 or 30 cm). Thus the general observation that

high densities result in higher yields must, however, take into account the

planting configuration of trees, as the evidence from Guevarra et al. (1978)

indicates that the closer spacing within rows can result in lower yields.




One possible factor giving rise to the often conflicting findings of experiments investigating the relationship between height and frequency on leaf

yields might be that of density. For example, the variation in cutting height

recommendations could be due to the effect of density on rooting volumes.

At high densities, which limit the soil volume that can be exploited by a

single tree, a low cutting height will more severely impair regrowth than at

lower densities. This shows the importance of the size of stumps remaining, from which reserves can be mobilized for regrowth, and explains the

better yields obtained by Pathak et al. (1980), Perez and Melendez (1980),

and Mendoza et al. (1975) at higher cutting heights.

The literature in not consistent concerning cutting frequency effects on

leaf yield and might be due partly to the different densities adopted by the

various experimenters. There is a strong interaction between cutting frequency and density (Horne et al., 1986). The ideal time to harvest the

woody species is after canopy closure and attainment of maximum Leaf

Area Index, but before any shedding of leaves due to shading has begun.

Harvesting after this stage will result in lower yields due to leaf abscission,

whereas harvesting before this optimum stage will increase the relative

proportion of the regrowth period taken up by the initial “lag phase”

before regrowth commences (Evensen, 1984). Thus, at low densities,

infrequent cutting will give the best yields, whereas at high densities, the



leaf canopy will close earlier and therefore shorter cutting intervals will

result in higher yields. Planting arrangement will also affect the optimum

time to harvest.

Trees growing in widely spaced rows will not be subject to the same

degree of light competition as trees equidistantly spaced, because of lateral

irradiation. Consequently, one might expect longer cutting intervals to

give the highest yields with trees planted in rows, whereas more frequent

cutting is required with equidistantly spaced trees at the same density. At

this stage, however, there is not yet a clear indication of the effect of

density and/or plant arrangement on leaf yields or the leaf : stem ratio.

Another possible reason for the conflicting findings concerning optimum

cutting frequency is a potential interaction with height of cutting. Low

cutting heights could limit the number of buds from which regrowth can

arise. Evidence for this comes from the work of Perez and Melendez

(1980), who showed that Leucaena cut at 30 cm formed fewer buds (89)

after a harvest than trees cut at 50 cm ( 1 12 buds). This might explain why

Guevarra et al. (1978) and Ferraris (1979), with very low cutting heights of

5- 10 cm, found that longer harvest intervals resulted in higher leaf yields,

whereas those workers who reported that frequent cutting gave maximum

leaf yields had higher cutting heights. Evensen (1984), however, did not

find any interaction between cutting heights (30,60, and 90 cm) and cutting

interval (40, 60, and 120 days).

There is little information available on the long-term effect of cutting

management on growth habit and form of each tree. The work of Savory

and Breen (1979) suggested that frequent cutting ultimately results in a

greater number of shoots regrowing from the stump than occurs with less

frequent cutting. Consequently, it is possible that the optimum harvest

interval might change with time. However, to date no cutting frequency

experiment has been conducted long enough for this to become apparent.

There are many other cutting management factors besides height and

frequency that might be important determinants of yield and persistence

but that have not yet been addressed by researchers. For example, it is

likely that regrowth will be more rapid from shoots remaining on the tree

after cutting than if regrowth arises solely from new buds, but no information is yet available to compare the effect of total versus partial removal of

leaf on long-term yields. Another factor that has received little attention is

the effect of age or size of the tree at the time of the first harvest on

subsequent yields. Ella (1987) has found that an additional consideration

that has not received any attention is whether initial cuts should be made

at a lower height than subsequent harvests in order to promote a multibranched stem with theoretically more potential bud sites from which

regrowth can arise.





Yields of nitrogen in material cut from leguminous trees are presented in

Table VI. Information about possible nitrogen yields is scanty, even for

Leucaena, because many of the experimenters who have reported dry

matter yields did not take samples of leaf or stem material for analysis of

nitrogen content. Generally, most researchers have found a mean nitrogen

concentration in leaf tissue within the range of 3.5% to 4.5% N , although

Endrinal and Mendoza (1979, cited by Mendoza, 1984) showed the N

content of young leaves to be as high as 6% N , whereas for mature leaves

the nitrogen concentration was found to drop to as low as 2.6% N.

Table VI

Reported Nitrogen Yields in Material Harvested from Tree Legumes




Sesbania sesban

Sesbania grandgora

Sesbania cannabina

Calliandra calothyrsus

Gliricidia sepium

Nitrogen yield

(kg N ha-' year - I )


614 (leaf)

Evensen (1984)

414 (leaf)

581 (leaf)

195 (leaf)

485 (leaf, 9 months only)

240 (total)

751 (total)

613 (total)

715 (total)

630 (total)

450 (total)

731 (leaf)

848 (total)


(leaf, 221 days only)

120 (leaf)

160 (total)


(total, 70 days only)

474 (leaf)

581 (total)

586 (leaf)

700 (total)

Ferraris (1979)

Guevarra et al. (1978)

Hutton and Beattie (1976)

Hutton and Bonner (1960)

Anslow (1957)"

Eriksen and Whitney (1982)

Ferraris (1979)

Guevarra et al. (1978)

Kitamura (1986)

Takahashi and Ripperton (1949)"

Catchpoole and Blair (1990a)

" Cited by Hutton and Bonner (1960).

Cited by Evans and Rotar (1987a).

Gore and Joshi (1976)b

Catchpoole and Blair (1990a)

Yost et al. (1985)

Catchpoole and Blair (1990a)

Catchpoole and Blair (1990a)



The nitrogen yield of 715 kg N ha-' year-', obtained by Guevarra et al.

(1978), has often been quoted by other researchers in the past decade as

evidence of the capacity of Leucaena to fix large quantities of atmospheric

nitrogen. However, as Halliday and Somasegaran (1983) pointed out, this

figure cannot be attributed solely to biological nitrogen fixation because of

the contribution from soil nitrogen. Eriksen and Whitney (1982), in estimating nitrogen fixation by Leucaena, attempted to allow for nitrogen that

had been taken up from the soil. They subtracted the average N uptake of

three grasses grown adjacent to the tree legume plots from the total nitrogen accumulation in the harvested Leucaena material and obtained an

estimate of nitrogen fixation of 656 kg N ha-' year-'. The limitation of this

approach is that soil N uptake by the relatively shallow-rooted grass

species might not provide a good indication of soil N uptake by deeprooted perennial Leucaena. This and other difficulties associated with

techniques to estimate N fixation by tree legumes are discussed later.

High-protein forage can, in theory, be continually cut from three legumes fertilized with all nutrients except nitrogen without adversely affecting nitrogen yields, because of the capacity for symbiotic nitrogen

fixation. However, it is most unlikely that tree legumes used as sources of

cut-and-carry forage throughout developing countries ever receive any

fertilizer, and consequently yields will decline with nutrient depletion.

Humphreys (1978) noted that where herbaceous pastures form part of a

cut-and-carry animal feeding system, the reduction in pasture soil fertility

is very rapid if the dung, urine, and reject forage are not carefully collected

and returned to the pasture.


Although cut-and-carry animal feeding systems are the norm in densely

populated areas of the tropics, in other areas where population pressure is

not as high, or land is unsuitable for cropping, ruminant animals are

permitted to graze freely. There are a number of advantages with grazing

compared to cut-and-carry systems. The labor input is generally less, and

nutrient rundown of the pasture is likely to be greatly reduced as a result of

nutrient recycling through the animal. Disadvantages include trampling

and the need for fencing and provision of water. In both systems there is

potential for large amounts of nitrogen to be lost through ammonia volatilization.

There are few reports of the direct contribution of tree legumes to an

associated nonlegume. In a field experiment conducted by Catchpoole and

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

V. Management of Tree Legumes

Tải bản đầy đủ ngay(0 tr)