Two experiments were carried out to evaluate the influence of restricted grazing of the legume tree Gliricidia sepium on live weight (LW) gain of growing cattle in Star grass pastures. Three treatments were evaluated in Experiment 1 using 18 animals of 176 kg initial LW with a completely randomized design: (0) grazing pasture alone, (C) pasture plus 1 kg/day of concentrate (16.6 % N x 6.25) and (G) pasture plus restricted grazing for 2 hours/day of Gliricidia (18.9-23.5 % N x 6.25).

Sixteen animals of 124 kg initial LW were used in Experiment 2 and only treatments C and G were evaluated. The pasture used was star grass (6.0-7.1 % N x 6.25) with rotational grazing.

The Gliricidia trees used were two years old and the main stem was cut at 50 cm above soil level. LW gains were 0.42, 0.52 and 0.62 kg/day in treatments 0, G and C of Experiment 1 and 0.21 and 0.32 kg/day in treatments 0 and G of Experiment 2. The LW gain increments of 100-110 g/day obtained with restricted gliricidia grazing are similar to the responses obtained with 0.5 kg/day of a concentrate supplement.

Gliricidia sepium foliage is a valuable feed resource for improving live weight (LW) gain of cattle on basal diets of grasses. LW gain increments of 100 to 200 g/day have been obtained offering the harvested foliage to growing cattle in confinement (Preston and Leng 1987; Seijas et al 1994; Zamora et al 1994), but the "cut and carry" management involved is a limitation to its use at farm level. An alternative to solve this problem could be direct harvesting by the animal managed by restricted grazing of "protein" banks of gliricidia.

Two feeding trials were carried out to evaluate the influence of restricted grazing of Gliricidia on the LW gain of growing cattle on star grass (Cynodon nlemfuensis) pastures.

Both trials were carried out at the Faculty of Agronomy in Maracay, which has a range in monthly average temperatures from 23.4 to 26.5 ºC and a mean rainfall of 989 mm. The first experiment started in July 1994, during the peak of the rainy season, and the second trial was conducted on the same paddocks beginning in October of the same year, at the end of the rainy season.

A completely randomized design was used to compare three treatments, (O) a control group grazing star grass, (G) sames as the control but with restricted grazing Gliricidia during 2hours/day and (C) same as control but with 1 kg/day of concentrate.

Eighteen Holstein x Brahman weaned cattle of 176 kg initial LW were balanced between treatments according to LW and the duration of the experiment was 12 weeks. All animals grazed together in five 3.68 ha paddocks of star grass and received ad libitum a mineral mixture in the field. The paddocks had occupation and rest periods of 10 and 40 days. At 6:30 hours the animals receiving concentrate were taken out of the field for two hours and put in individual stalls where the concentrate was offered. This supplement contained 48.75% rice polishing, 48.75% soya bean husks and 2.50 % mineral mixture.

The animals on the gliricidia treatment were introduced from 6:30 to 8:30 h into a paddock which had a pure stand of this tree. Five 0.2 ha paddocks were used with occupation and rest periods of 10 and 40 days. The trees had been were established at 1 x 2 m distances, were approximately 2 years old when grazing started and after each grazing were cut at 50 cm above soil level.

Sixteen weaned Holstein x Brahman cattle, eight of each sex, were used in a completely randomized design with a 2 x 2 factorial arrangement to compare the restricted grazing of Gliricidia (G) with a control group grazing only star grass (0). The same paddocks and rotations of star grass and Gliricidia of Experiment 1 were used, but the rotations of the grass and the legume were simultaneous and the animals on Treatment G remained continuously on the Gliricidia during the first two days and from 6:30 to 8:30 h the other eight days. The star grass paddocks, but not the Gliricidia banks, were irrigated when it was required. The animals in this trial had an average initial LW of 124 kg and were balanced between treatments according to weight. The duration of the experiment was 12 weeks. All animals received a mineral mixture ad libitum in the field.

The animals were weighed weekly at 06:00 hours and LW gain was calculated by the regression of LW on time. The amount of star grass on offer was estimated by throwing 1.0 x 0.5 m quadrants six times just before and after grazing each paddock and cutting the herbage at ground level. The samples were weighed and mixed to obtain two sub samples, one to determine dry matter and the other to determine the proportion of leaves, stems and dead material. Twelve Gliricidia plants of each paddock were sampled before and after grazing by harvesting the branches from which the leaves and woody stems were separated and weighed. The samples of herbage and concentrate were dried at 65 ºC for 48 hours.

The samples of star grass, concentrate and Gliricidia were grouped in four week periods and ground through a 1 mm screen. They were analysed for crude protein (N x 6.25) and ash (AOAC 1984), neutral detergent fiber (NDF) (Goering and Van Soest 1970), calcium (Fick et al 1979) and phosphorus (Harris and Popat 1954).

Average N x 6.25 content (DM basis) of whole grass samples taken at ground level were 6.0 and 7.1% in the two experiments and increased to over 10 % in leaves (Table 1). The NDF levels were above 75% in all grass analyses with little difference between leaves and total biomass. The offer of standing grass before grazing was between 4,842 and 6,412 kg DM/ha in both trials and after grazing had fallen to 2956 kg/ha (Table 2). The proportion of stems was about 50 % before grazing and leaves constituted less than 30% of total DM available (Table 3). Leaf percentages appreciably decreased at the end of the grazing periods.

Table 1: Chemical composition of the feeds
	N x 6.25	NDF	Ash	Ca	P
	--------------- % of DM--------------
Experiment 1
Star grass
Whole sample	6.0	80.5	8.8	0.32	0.41
Leaves	11.6	75.2	9.2	0.38	0.49
Gliricidia	23.5	43.3	11.8	1.28	0.38
Concentrate	16.6	58.9	9.8	1.19	0.30
Experiment 2
Star grass
Whole sample	7.1	79.0	9.3	0.31	0.40
Leaves	11.4	76.4	10.4	0.50	0.46
Gliricidia	18.9	55.1	9.9	1.77	0.31

Table 2: Weight of standing forage before and after grazing
	Grass	Gliricidia
	--kg DM/ha--
Experiment 1
Before grazing	4842	1037
After grazing	2956	677
Experiment 2
Before grazing	6412	617
After grazing	5204	371

Table 3: Composition of forage samples
	Grass	Gliricidia
	leaf:stem:other*	leaf:woody stems
Experiment 1
Before grazing	18:53:29	73:27
After grazing	7:59:34	39:61
Experiment 2
Before grazing	28:48:24	56:44
After grazing	14:57:29	16:84

The means for N x 6.25 content (DM basis) of Gliricidia foliage were 23.5 and 18.9% in Experiments 1 and 2. NDF levels were lower and calcium percentages higher than those observed in star grass (Table 1). The offer level of Gliricidia branches per unit area was higher in the first trial and in both experiments had a high proportion of leaves (Tables 2 and 3). The amount of DM per ha and the proportion of leaves were markedly reduced after grazing.

LW gains were higher in the first trial. The animals on star grass alone (Treatment 0) gained 0.42 kg/day and LW gain increased to 0.52 and 0.62 kg/day in treatments G and C (Table 4). LW gains also increased in Experiment 2 from 0.21 kg/day for animals in treatment 0, to 0.32 kg/day on restricted grazing of Gliricidia.

Table 4: Mean values for initial live weight (LW) and LW gain of animals
	Init Wt	LW gain
	(kg)	(kg/day)
Experiment 1
O: grazing star grass	175	0.42
G: O+2 h/day gliric	177	0.52
C: O+1 kg/day conc	177	0.62
SE		±0.030
Prob		0.05
Experiment 2
O: grazing star grass	124	0.21
G: O+2 h/day gliric	124	0.32
SE		±0.034
Prob		0.05

The pastures used in both trials were high in fibre and low in N x 6.25; there were large differences in N x 6.25 content between leaves and the whole standing biomass. The proportion of leaves in the herbage was low (Table 3), but the animals selected this fraction and it decreased from 18 to 7 % in Experiment 1 and from 28 to 14 % in Experiment 2. The high offer level of standing biomass before and after grazing in these trials (Table 2) and the spatial distribution of this stoloniferous grass allowed a high degree of selection by the animals. The offer level of the standing biomass was over 2000 kg DM/ha and no restrictions in herbage intake by this factor were to be expected (Minson 1990).

The LW gains recorded when animals grazed star grass alone (0.42 and 0.21 kg/day in Experiments 1 and 2) were within the range of values obtained with cultivated pastures in the tropics (Stobbs 1976). The higher values in Experiment 1 are not explained by the chemical composition of the pastures, which was very similar in both trials (Table 1), or by the herbage on offer which was slightly higher in Experiment 2 (Table 2). The lower LW gains in the latter trial could be related to the stress caused by weaning the animals just before the beginning of the experimental period.

The opportunity to browse the Gliricidia trees resulted in average increases of 100-110 g/day in LW gains. These responses are in the lower range of those found with stall fed animals supplemented with harvested gliricidia foliage (Preston and Leng 1987; Seijas et al 1994; Zamora et al 1994), but labour requirements are larger with the latter management. Neither the amount of legume consumed nor its proportion in the whole diet could be estimated in these trials. But the relatively low offer level and utilisation (Table 2) indicate that it was only a small component of the diet.

Supplementation with 1 kg/day of concentrate in Experiment 1 increased LW gain by 200 g/day, which is similar to the response observed in other trials carried out at this Institute (Combellas 1993). The responses observed with restricted grazing of Gliricidia, of about 100 g/day LW gain, are comparable to those likely to be obtained with a supplementation of 0.5 kg/day of concentrate.

The restricted daily grazing of Gliricidia sepium is a simple and economical management practice, and the results obtained in these trials have shown that moderate LW gain increases could be obtained with animals grazing cultivated tropical grasses.

This research was partially sponsored by the International Atomic Energy Agency, Project VEN/5/017.

Received 20 September 1998