| Livestock Research for Rural Development 9 (4) 1997 | Citation of this paper |
Agricultural College of Ho Chi Minh University, Ho Chi Minh city,
Vietnam
Rowett Research Institute, Bucksburn, Aberdeen, Scotand (e-mail; blk@rri.sari.ac.uk)
Leaves from twelve different trees and three agricultural crops, commonly available in the tropics were selected, dried and subjected to analysis by the in sacco nylon bag degradability test, the in vitro gas production test and the Kjeldahl test for nitrogen.
The protein content of the leaves was negatively related with the proportion of leaves in the harvested fresh foliage (r = -0.63). The relative ranking of the feeds was similar when evaluated by the in vitro and in sacco methods after 24 or 48h incubation, and was in line with practical observations on the relative value of these feeds for goats. The exception was the sugar cane bagasse (pressed cane stalk after extraction of 2/3 the juice in an artisan crusher) which had a higher gas production than the best of the tree leaves (Cassava, gliricidia and I.teysamii) but a lower in sacco degradability. A similar situation prevailed in the case of the sugar cane leaves which were rated relatively higher by the in vitro gas test tnan the in sacco degradability method. The ranking of the feeds using the sum of the coefficients "a" + "b" (the total potential degradability) derived from the equation p=a+b(1-e-ct) in the in sacco rumen degradability test was broadly similar (r = 0.56) to the ranking by the 24h dry matter loss. The "c" coefficient was not related (r = 0.16) to the 24h dry matter loss nor to the practical assessment of the nutritive value of the feeds.
It is concluded that the in vitro gas production and in sacco rumen degradability techniques are equally useful for predicting the potential nutritive value of tropical tree leaves but that the in vitro gas method gives erroneous results for crop residues containing moderate levels of soluble sugars.
The in sacco rumen degradability (rskov et al 1980) and the in vitro gas production (Menke and Steingass 1988) tests are useful for the rapid screening of feeds to assess their potential as energy sources for herbivorous animals (Preston 1995). This paper documents the results of using these methods to evaluate the leaves of trees and shrubs commonly found in Vietnam. Included in the leaves selected for the study are those from well known species such as Gliricidia. sepium, Trichantera. gigantea and Leucaena. leucocephala and some from species which have been studied less (such as Artocarpus heterophyllus [jackfruit]) or not at all (Anacardium occidentale [cashew]). Rice straw, bagasse and leaves of sugar cane and banana were included for comparative purposes.
Acacia auriculiformis and Acacia mangium are two of many species of Acacia which are widely used for reforestation in Asia. They are leguminous, fast growing trees which produce abundant foliage. Acacia auriculiformis has been planted on thousands of hectares in Vietnam as part of a reforestation programme.
Cashew (Anacardium occidentale) is a tree well known for it's fruit and seeds which are used for human consumption. It is a non-leguminuos, ever-green tree which will grow up to 12m in height.
Artocarpus heterophyllus (jackfruit) is a common, non-leguminous tree growing up to 15m high which bears a large fruit popular for human consumption. Jackfruit leaves are widely used as a feed source despite their apparently low protein digestibility thought to be due to the high tannin content.
Bambusa arundinacea (bamboo) is a very common multi-purpose tree which is used for construction, firewood, and human and animal foods.
Gliricidia sepium is a leguminous tree receiving much attention for use as a multipurpose tree species. It is a fast growing tree (up to 15m in height) widely used in Lain America, the Caribbean and SE Asia. It is used in coffee and cocoa plantations to shade the plants.
Indigofera teysami is another fast growing leguminous tree with good potential as a multipurpose resource. It is very similar in appearance to Gliricidia and has long been used for reforestation purposes in Vietnam.
Leucaena leucocephala is native to Mexico but very common and widely utilised in Asia as a fodder crop. It is drought resistant which makes it an invaluable dry season feed. However, L. leucocephala suffered badly from attack by the psyllid pest Heteropsylla cubana in the late 1980's. Leaves and seeds contain mimosine, a toxic amino acid which can produce ill effects in ruminants which do not have the necessary micro-organisms in their rumen microflora to detoxify it.
Manihot esculenta (cassava) is a shrub or small tree growing up to 4m high with edible roots containing high levels of starch which are widely used for both human and animal consumption in the tropics and sub-tropics. The leaves can be used as animal feed either fresh, dried or ensiled for storage and feeding during the dry season.
Trichantera gigantea is a non-leguminous tree native to Colombia and Venezuela. It is known to be low in phenols and steroids with a high degradability and is therefore receiving much attention for use as an animal feed. However, the leaves have a slightly hairy surface which may reduce their palatibility. T. gigantea grows well in a mixed cropping system with banana plants which provide shade.
Musa spp (banana) is a plant widely grown for it's fruit for human consumption. The banana plant is between 3 and 5m high. Banana leaves are a crop residue used as emergency feed for ruminants but the high tannin content appears to cause problems with digestibility.
Leaves from ten different trees and four commonly available crop residues in the tropics were selected, dried and subjected to analysis by the in sacco nylon bag degradability test, the in vitro gas production test and the Kjeldahl test for nitrogen.
Samples of the tree leaves and feeds were collected as they would be harvested to feed goats. The leaves were removed from the stems and the percentage by weight of leaf and stem was calculated from the fresh samples. The leaf samples were dried for a minimum of 36 hours at 60C and their dry matter content was estimated. All samples were then ground in a laboratory mill through a 1mm screen.
The nitrogen content of each of the samples was estimated using the Kjeldahl technique and Kjeltech apparatus as described by Davidson et al (1970); the crude protein content was calculated by multiplying the nitrogen content by 6.25.
The procedure followed was that described by Menke and Steingass (1988). Samples of rumen fluid were taken from a rumen fistulated Bos indicus heifer (diet was 50% urea-treated rice straw and 50% ensiled cassava peel residue with 500g cottonseed meal daily and molasses-urea block ad libitum) using a plastic tube fittted with a suction valve. The rumen fluid was strained through gauze and mixed with buffer in a ratio of 1:2. Carbon dioxide was provided in a steady stream over the buffer/rumen fluid mixture throughout the mixing and dispensing procedure.
200mg of the previously dried and milled samples were weighed into glass syringes which were incubated in a waterbath at 38C for 30 minutes before addition of the buffer/rumen fluid mixture. 30ml of buffer/rumen fluid mixture was dispensed into each syringe by means of a semi-automatic pump dispenser; all gas was expelled from the syringe and the volume of the contents recorded. The syringes were suspended in a waterbath at 38C and the volume of the contents measured after 3, 6, 12, 24, 48, 72 and 96 hours. Samples were done in duplicate with G. sepium included in every run as a standard and blanks containing only buffer/rumen fluid mixture as a control.
The procedure used to determine the dry matter degradability of the feeds was that described by Ørskov et al (1980). 5g of the previously dried and ground samples were weighed into nylon bags (pore size 50 microns) which were then attached to plastic tubes. Bags were incubated in the rumen of rumen fistulated Bos indicus heifers for 0, 3, 6, 12, 24, 48 and 72 hours. After incubation, bags were washed by hand until the water ran clear and dried to a constant weight at 60C.
Replicates of each run were placed in the rumens of 3 different animals all of which were given a diet of 50% urea-treated rice straw and 50% ensiled cassava peel residue with 500g cottonseed cake daily and molasses urea block ad libitum.
Due to the very small particle size of the milled samples, it was necessary to check
the loss of particles through the pores in the nylon bag. This was done by washing 1g of
each sample in filter paper with 1 litre of water, or until the filtrate was colourless,
then drying at 110C.
In presenting the results, the feed samples were divided arbitrarily into
"leaves" (from trees) and "crop residues" (from sugar cane, rice and
banana). The exception is in Table 1 where banana leaves are included with the other
leaves. The proportions of leaves obtained from the harvested biomass are shown in Table
1. For banana leaves, the harvested matter was both leaf blade and leaf mid-rib. Only the
leaf blade was considered in the calculation of percentage leaf.
Table 2 shows the dry matter and crude protein contents of each of the feed samples.
The protein content of the leaves appeared to be negatively related with the proportion of
leaves in the harvested fresh foliage ((r = -0.63; Figure 1). Highest protein contents
(24-29% in DM) were in leaves from cassava and from the three legume trees, I.
teysamii, L. leucocephala and G. sepium; the lowest (10%) was in cashew.
The observation that the proportion of leaf as percentage of fresh foliage harvested was positively related to the crude protein content of the dry matter of the leaves has no obvious explanation. One possible reason is that in plants where the greater proportion of the biomass is lignified supporting material (ie: more stems and branches and lower proportion of leaf), the leaf is less lignified and contains more protein, but in plants where there is less supporting material (fewer stems and branches), the distribution of protein is more uniform.
The results for in sacco degradability and in vitro gas production of the range of leaves are ranked in Figure 2 according to the values at 48 hr for the in sacco method..
Data for the crop residues are presented in a similar way in Figure 3. The data for all incubation times for in vitro gas production and in sacco rumen degradability, are in Appendices 1-3. The mean values for the coefficients "a", "b" and "c", derived by fitting the equation p=a+b(1-e-ct) (McDonald 1981) are shown in Appendix 3.
The relative ranking of the feeds (Figures 2-3) was similar for both methods of evaluation and in line with practical observations concerning the relative value of these feeds for goats and sheep (Dinh Van Binh and T R Preston, unpublished observations) The exception was the sugar cane bagasse which had a higher gas production but a lower in sacco degradability than the best of the tree leaves (Cassava, gliricidia and I.teysamii) (appendices 1-3). A similar situation prevailed in the case of the sugar cane leaves which were rated relatively higher by the in vitro gas test than the in sacco degradability method (appendices 1-3).
The sugar cane bagasse was the pressed stalk remaining after squeezing out the juice in a 3-roll crusher. This artisan method of juice extraction leaves a fibrous residue with about 25% sugar in the dry matter. The sugar cane leaves can contain up to 10% of sugars. It would seem that the presence of high concentrations of soluble sugars results in gas production data which are not related to the digestibility of the feed in the animal.
L. leucocephala produced an unexpectedly small amount of gas compared to the high rumen degradability observed. This could be an effect of the mimosine content of the leaves. This compound may have inhibited the fermentation in the small volume of the syringe but in the larger volume of the rumen, the dilution is greater thus reducing the effect. Jackfruit leaves are known to have a high content of tannins, but the similar results for gas production and in sacco rumen degradability indicate that these compounds have less effect on the rumen fermentation as compared to the mimosine in L. leucocephala.
The ranking of the feeds for the coefficients "a" + "b" (the total
potential degradability) in the in sacco rumen degradability test was broadly
similar (r = 0.75) to the ranking by the 48hr dry matter loss. The "c"
coefficient was not related (r = 0.08) to the 48hr dry matter loss nor to the practical
assessment of the nutritive value of the feeds (Dinh Van Bien and T R Preston, unpublished
observations). For example the cashew leaves had the highest "c" value but the
lowest in sacco dry matter loss and the lowest rate of gas production.
This study has shown that there are potentially many sources of fodder for ruminants in
the form of tree leaves which are not, as yet, being utilised to the maximum possible
extent. The in sacco rumen degradability measurement and the in vitro
gas production appear to be equally suitable for assessing the potential nutritive value
of tree leaves. However, for feeds with moderately high concentrations of soluble sugars
(10-20%), the gas production technique gives inflated results.
Davidson J, Mathieson J and Boyne A W 1970 The use of automation in determining nitrogen by the Kjeldahl method, with final calculations by computer. Analyst 95: 181-193.
McDonald I 1981 A revised model for the estimation of protein degradability in the rumen. Journal of Agricultural Science (Cambridge) 96: 251-252
Menke K and Steingass H 1988 Estimation of the energetic feed value obtained from chemical analysis and in vitro gas production using rumen fluid. Animal Research and Development 28.
Ørskov E R, Hovell F and Mould F 1980 The use of the nylon bag technique for the evaluation of feedstuffs. Tropical Animal Production 5: 195-213.
Preston T R 1995 Tropical
animal feeding: A manual for research workers. FAO, Rome
Received 20 June 1997
