OBSERVATIONS ON THE FORAGE POTENTIAL OF VELVET BEAN

Absnaa: A field trial was conducted with velvet bean (Mucuna spp.) planted during the Maha season of 1983184. Whole crop samples ot tne above ground dry matter were taken at four different stages of growth (60, 90, 120 an& 150 days) t o assess the interaction of dry matter yield and forage quality during maturity. Dry matter yield increased with maturity, but quality parameters steadily declined. The results suggest that the velvet bean forage should be harvested around 90 days. Harvested at this maturity, a dry matter yield of 3.1 tonnes per hectare, with a crude protein content of 20.6% and 55.4% in mtro digestibility, could be obtained.


Introduction
The usefulness of legumes as components in forage mixtures is well recognised.The beneficial effects attributed to the legume derive partly from its ability to fix atmopheric nitrogen and partly from the high protein contents of its dry matter.14y20In Sri Lanka, at least on account of the latter reason, establishment of legumes is justified.
Velvet bean (Mucuna spp; Syn Stizolobium ~p p ) , ~O which was origi-\ nally ihtroduced as a cover crop in coconut p1antations;has the potential as a forage crop.The major characteristics of this forage legume has been recently reviewed1 and, in fact, its use is recommended foi the hill country regions1 * The study reported herein is a preliminary attempt to assess the productiviby and nutritive value of velvet bean forage at four different growth stages.Such assessment has not been reported previously for velvet bean grown in Sri Lanka.

Description i
Several selections of velvet bean are known t o exist in Sri ~a n k a , ' but the exact classification is difficult as the taxonomy of the species is confusing.6

C"
They are all annual-perennial legumes with vigorously growing vines which may extend up to 6 m in length.The leaves are trifoliate with large, ovate ledets.The white to dark purple flowers appear in long pendant clusters.Pods, 10 to 14 in a cluster, are borne singly and have a greyish-white pubescence of short, silky hairs.Three seedcoat colour types, namely black, white and mottled, are common. 1 5 , 2 0 , 2 3riginating in India, it is now naturalized in all tropical countries.Their distribution has been extended into the temperate countries by breeding.' The legume is tolerant to drought due to its deep roqtsn6 .' The velvet bean variety used in the present evaluation was black-seeded.The seeds were obtained from the Coconut Research Institute, Lunuwila.

Growing conditions
Velvet beans were planted (three seeds per hill) at a spacing of 60 x 60 cm during the 1983184 Maha season in an experimental plot (6.0 x 12.0 m) at the Department of Animal Science, University of Peradeniya.The soil at the site was reddish brown latasolic soil with a pH of 6.1.A well distributed rainfall was received during the experimental period.
The plot received a basal application of one tonne of poultry litterlha (equivalent to approximately q2 kg N, 36 kg P205 and 35 kg K 2 0 per hectare) at planting.The stand was thinned t o two plants per hill after one month and bamboo trellises were provided for support.The supports were given mainly t o prevent matting of plants from adjacent hills and t o facilitate the yield estimations.,

Sampling procedures
The flowering in velvet beans commences around 90 days and continues for another 50 -70 days.The sampling of the forage was done at 60 days (pre-bloom), 90 days (start of bloom), 120 days (mid-bloom) and 150 days (post-bloom).At each age of harvesting, twenty plants (from ten hills) were randomly selected and cut 1 0 cm above ground.Leaves and stems (and pods at 120 and 150 d a y s of age) from each plant were hand-separated, and weights were recorded.Sub-samples (ca 500 g) of the different anatomical parts were removed t o the laboratory for dry matter determination.Further sub-samples (ca 2 kg) were taken and dried at 60°C in an unitherm oven for 36 -48 hours.The dried samples were then ground and stored in air tight plastic jars for subsequent laboratory evaluation.vklvet bean pods and seeds are also reported to have been used for livestock feeding.611For this reason, samples of green pods (immature and s mature), seeds and pod husks were also collected for analysis.The dry matter, crude protein, ether extract and ash contents were determined according to standard procedures2 The detergent fibre composition was determined using the method of Goering and Van ~oest.' Znuitro organic matter digestibility (IVOMD) was determined according t o the method of Tilley and ~e r r y ~~ as modified by ~a r n e s .~ Itumen liquor was a collected from two fistulated cattle which have been previously fed a mixture of straw and legumes.Five standard samples of known in vim digestibility were used in each in vim run t o correct for run-to-run variation.
The standard samples consisted of legume and straw samples.

Results and Discussion
The forage dry matter yields of velvet bean increased with maturity (  The relative dry weight distribution of different anatomical parts (Table 1) show that at 60 days the leaves constituted 55% whereas only 45% was stems.This changed during growth t o 30% leaves and 61% stems by 150 days.The reproductive parts, mainly pods, constituted the balance 9%.
The general tendency of a decreasing 1eaf:stem ratio during growth is in agreement with previous studies reported for other tropical and temperate legumes.As shown in Table 2, the velvet bean forage contained 20.6 -24.8%.crude protein during the pre-bloom stage.The crude protein content declined rapidly with flowering, due to a redistribution of nitrogen to the seeds.5Despite this decline, the crude protein content of the legume herbage (even at 150 days) remained at a higher level than of common grasses avaiIable in Sri b k a .O CeIl walls are a major part of the plant.The amount and composition .of.the'cell walls is probably the most important factor influencing the nhtritiv'e'value of forage^.^But studies of the composition of the cell walls in .' tropical .foragesare scanty.As such, comparison of the present data with other, work is not 'possible.The data (Table 2) shows that all .cellwall components steadily increased with maturity.
. The in vitro digestibility of the velvet bean forage decreased from 66.4% to 42.2% as the plant matured.This is to be expected, since the general trend of decline in the digestibility of forage crops with plant maturity is well documented.8119This decrease can be attributed t o a .combination of factors interalia decreasing leaf to stem ratio, decreasing amount of cfude protein, increasing amount of cell walls and increasing lignification.Although not reported separately in the present data, the fact that the chemical composition of leaves and stems during crop maturity followed distinctly different patterns needs a special mention.The contents of crude protein, cell walls and lignin and in vitro digestibility of the leaf, fraction changed only a small extent during maturity, while in the stems all these parameters decreased rapidly.
?'able 3. Chemical composition and IVOMD of some reproductive parts of velvet bean.Surprisingly the mature green pods had a higher crude protein content than tender green pods (Table 3); but the IVOMD of mature pods was lower.The crude protein and IVOMD values of mature velvet bean seeds were 26.4% and 70.7%, respectively.These values compare closely to those reported for the common food legumes grown in the Asian ~o n t i n e n t .~ Recent reports suggest that the velvet bean seeds have potential as a protein source both in human and animal nutrition, "118 The data show pod husks to be a poor quality roughage.The crude protein content and IVOMD value of pod husks are similar to those generally quoted for paddy straw.lA0ii On the strength of these preliminary results, the forage potential of velvet bean appears good.The data indicate that the velvet bean forage 2 should be harvested around 90 days.Harvested at this age, velvet bean would provide a yield of 3100 kg/ha of moderate quality (20.6% crude protein and 55.3% IVOMD) forage dry matter.The productivity of velvet bean forage is sufficiently encouraging t o warrant further field trials.The persistency of the legume under repeated harvesting and its compatibility with companion grasses are aspects that need be evaluated in future studies. .

c
Extrapolated yie d based on a plant density of 27,225ha.

Table 1 )
. The observed yields are in the same range as those reported by King e t al.3in Australia and by Takahashi and ~i p ~e r t o n ~ in Hawaii.

Table 1 .
Some productive parameters of velvet bean forage.

Table 2 .
Chemical composition and IVOMD of velvet bean forage at four different stages of growth (percentage DM basis).