The hypothesis to be tested in the present study was that the biomass yield of the multi-purpose tree Trichanthera gigantea would be enhanced by growing it under partial shade in association with either bananas or the legume tree Leucaena leucocephala. Two experiments of split-plot design were carried out. In the first experiment, the comparison was beween partial shade from bananas (at 5 m spacing) or no shade. In the second experiment, partial shade from Leucaena leucocephala (5 m spacings) was included as a third treatment.

A higher proportion of the Trichanthera cuttings germinated in the shade than in the "sun". The biomass yield of Trichanthera was higher when it was associated with bananas than when planted alone or in association with Leucaena. The absence of effect of the Leucaena may have been because the canopy of this tree is more "open" than in the case of the bananas hence the shade effect was less. There were no significant differences for biomass yield of Trichanthera between plant densities of 20 and 25 thousand plants/ha. On the basis of harvests every 3 months, then the observed yield (10-12 tonnes fresh biomass/ha/harvest) in the associations with bananas is in the range of 40-50 tonnes fresh biomass/ha/year.

Increasing attention is being given to new protein sources, such as the leaves of multi-purpose trees and water plants, in the search for alternatives to conventional protein supplements for both ruminant and monogastric livestock (Gomez et al 1996). Development of these plants as animal feed is also compatible with the strategy for promoting sustainable use of natural renewable resources requiring minimum external inputs (Preston and Murgueitio 1994). The multi-purpose tree Trichanthera gigantea, native to the coffee-growing region of Colombia , is of particular interest as the leaves appear to contain few anti-nutritional compounds (Galindo et al 190) and are readily consumed by pigs (Sarria 1994).

Cuttings of Trichanthera gigantea were introduced into Vietnam in 1991. The tree has been readily adopted by farmers throughout the country because of its tolerance to a wide range of ecological conditions and apparent resistance to pests and diseases (Nguyen Ngoc Ha and Phan Thi Phan 1993; Nguyen Thi Hong Nhan, personal observations).

After five years of observations at the Cantho University, it was apparent that Trichanthera gigantea could grow well in the condition of the Mekong Delta, providing biomass in both the dry and wet seasons (Nguyen Thi Hong Nhan, unpublished data). A feature of this tree is its apparent preference for growing under conditions of partial shade. This has been observed in both Colombia and Vietnam (Preston T R and Binh Van Dinh, personal observations).

The hypothesis to be tested in the present study was that the biomass yield of Trichanthera would be enhanced by growing it under partial shade in association with either bananas or the legume tree Leucaena leucocephala.

Two experiments were conducted during the period from January to October 1995 in Cantho city where the yearly rainfall is 1600-1700 mm, the mean temperature 26-28 ?C and the relative humidity 78-88 %. The soil in the experimental area is "sulphate acid" with pH range of 4.06-3.8. The contents of organic matter, N , P and K are low.

There were two sets of treatments arranged in a 2 x 2 split - plot design. The main plots were with and without shade; the sub-plots were planting distances of 100 x 40 cm vs 120 x 40 cm. The shade was provided by banana trees planted as suckers t 5 x 5 m spacing at the same time as the Trichanthera.

The treatments in a 3 x 2 split-plot arrangement were: shade from banana, leucaena or none (main plots); planting distance of 100 x 40 cm or 120 x 40 cm (split-plots). The bananas (as suckers) and leucaena (seedlings from the nursery) were planted 2 months after the Trichanthera. The Trichanthera was established by planting brown stem cuttings in 10 cm deep holes. The total experimental area, including protection rows, was 2,500 m² .

Table 1. Effect of shade and spacing on germination and growth of Trichanthera gigantea
	Exp*	Sun	Shade Leucaena	Shade Banana	SE/Prob	Spacing, 100*40	Spacing, 120*40	SE/Prob
Germination, %	1 2	74.6 85.7	- 84.72	81.3 78.9	1.2/0.005 2.4/0.13	76.9 82.4	79.0 84.0	1.2/0.54 1.9/0.63
Plant height, cm -1st harvest -2nd harvest -3rd harvest	1 1 2	88.8 115 88.3	- - 87.8	121 139 95.1	3.2/0.001 6.9/0.04 0.2/0.01	109.9 125.7 88.9	110.3 128.0 88.9	3.2/0.94 6.9/0.82 0.2/0.99
No. Branches -60 days -180 days	2 2	6.7 12.8	6.7 13.1	6.5 14.3	0.1/0.4 0.3/0.007	6.5 13.2	6.7 13.6	0.1/0.06 0.3/0.34

* Experiement No

Table 2. Effect of shade and planting distance on biomass yield of Trichathera in experiment 1
	Sun	Shade	SE/Prob	100*40	120*40	SE/Prob
Fresh biomass, tonnes /ha -1st harvest -2nd harvest	9.6 9.1	13.01 12.4	1.1/0.05 1.1/0.06	11.5 11.97	11.1 10.6	1.1/0.8 1.1/0.8
Biomass DM, tonnes/ha -1st harvest -2nd harvest	1.6 1.3	2.1 1.8	0.2/0.07 0.2/0.11	1.8 1.6	1.9 1.5	1.2/0.9 1.2/0.8

The Trichanthera was harvested by cutting all the biomass at 70 cm above ground level. Yields were recorded for the first two harvests, the first 6 months after planting and the second 3 months later. In experiment 2, the harvested material was separated physically into leaves and stems. Bulked samples from all treatments in experiment 2 were dried at 60 ºC and subjected to standard proximate analysis. Soil samples from the plots (top 10 cm) in experment 2 were taken for standard analysis. There was no irrigation of the plots and no pest control in either experiment.

Mean values for germination rate, plant height at harvest and numbers of branches at 90 and 180 days are shown in Table 1 for the two experiments. A higher proportion of the Trichanthera cuttings germinated in the shade than in the "sun" treatment in Experiment 1 but there were no differences in Experiment 2. This was probably because in Experiment 1 the banana suckers were planted at the same time as the Trichanthera cuttings, thus providing a degree of shade during the critical second month after planting when the shoots of the Trichanthera began to emerge. In contrast, in Experiment 2, the shade trees were planted two months after the Trichanthera. In both experiments, the height of the tree at harvest was greater when Trichanthera was associated with bananas, but not with Leucaena (Experiment 2). In Experiment 2, when the branches of Trichanthera were counted, there was a greater number on the banana treatment but no differences due to Leucaena. Density of planting had no effect on any of the parameters.

The biomass yield of Trichanthera was higher when it was associated with bananas than when planted alone or in association with Leucaena (Tables 2 and 3). The absence of effect of the Leucaena may have been because the canopy of this tree is more "open" than in the case of the bananas hence the shade effect was less. There were no significant differences for biomass yield of Trichanthera between different spacing. If 4 harvests are taken per year then the observed yield in the associations with bananas is in the range of 40-50 tonnes fresh biomass/ha/year, which is similar to the yield data reported by Gomez and Murgueitio (1991).

Table 3. Mean values for biomass yield at the first harvest of Trichanthera planted alone or in association with Leucaena or bananas in experiment 2
	Sun	Shade Leucaena	Shade Banana	SE/Prob	Spacing, 100*40	Spacing 120*40	SE/Prob
Yield, tonnes/ha Fresh - Whole - Leaves - Young stems	8.03 6.62 1.43	8.2 6.6 1.6	10.5 8.7 1.2	0.5/0.03 0.4/0.003 0.1/0.04	9.5 7.8 1.7	8.4 6.9 1.6	0.4/0.09 0.3/0.08 0.1/0.4
Dry matter -Whole -Leaves -Young stems	1.6 1.3 0.26	1.7 1.3 0.3	2.1 1.8 0.4	0.1/0.001 0.09/0.01 0.03/0.02	1.8 1.5 0.32	1.7 1.5 0.3	0.08/0.4 0.08/0.8 0.02/0.3

Table 4. Chemical composition in DM of Trichanthera gigantea
	CP	EE	CF	Ash	Ca	P	Carotene
	% in DM						mg/kg
Leaves	18.2	5.9	13.8	19.9	4.3	0.92	385
Stem	11.9	3.7	30.6	31.3	6.4	2.1	0.00

CP N*6.25; EE Ether extract CF Crude fibre

The data for nutrient analysis in table 4 indicate that the leaves of Trichanthera are rich in protein, minerals and carotene and relatively low in crude fibre. It was noted that the Trichanthera coppiced rapidly and quickly regenerated new shoots. The regrowth was more vigorous than from stem planting presumably because the plant does not expend in growing roots.

Trichanthera cuttings were distributed to farmers in several provinces in the Mekong Delta. It was observed that the tree tolerated flooding for periods of up to one month. In severe drought, the leaves died but were quickly regenerated when the rains started. It showed high resistance to pests and diseases. Pseudococcidae has been found recently, that attacks the young plant, but rarely causes serious damage. Once rapid growth begins, Trichanthera forms a dense canopy of foliage that shades out weeds .

* Trichanthera gigantea is easy to establish in the conditions of the Mekong Delta in Vietnam.

* Growing it in association with bananas leads to increased yields of biomass wiith no apparent effect on the productivity of the bananas.

* There were no apparent differences in yield between plant densities of 25 and 20 thousand per ha.

This research was supported financially by the International Foundation for Science through a grant to the senior author.

Sarria Patricia 1994 Efecto del Nacedero ( Trichanthera gigantea ) como reemplazo parcial de la soya en cerdas en gestacion y lactancia recibiendo una dieta básica de jugo de caña . Livestock Research for Rural Development , Volume 6, Number 1:62-73

(Received 1 January 1996)