Livestock Research for Rural Development 15 (11) 2003

Citation of this paper

Cassava hay and Stylo 184 hay to replace concentrates in diets for lactating dairy cows

Krailas Kiyothong and Metha Wanapat *

Khon Kaen Animal NutritionResearch and DevelopmentCenter, Khon Kaen 40260, Thailand
* Department of Animal Science, Faculty of Agriculture
,KhonKaenUniversity, Khon Kaen 40002, Thailand



Sixteen multiparous Holstein-Friesian crossbred cows in mid-lactation were blocked according to days in milk and previous lactation and randomly assigned according to a randomized complete block design with four replications to receive four dietary treatments consisting of different combinations of cassava hay (CH) and / or stylo 184 hay (SH). The dietary treatments were:  allowance of concentrate to milk yield 1:2 (control);  allowance of concentrate to milk yield 1:2 plus 1 kg//hd/d of cassava hay (1CH); allowance of concentrate to milk yield  1:2 plus 1 kg/hd/d of CH+SH  (1CHSH); and allowance of concentrate to milk yield 1:3 plus 2 kg/hd/d of CH+SH (2CHSH). All animals received Ruzi grass ad libitum in a cut-and-carry system as roughage source. The feeding trial lasted for 9 weeks.

There was no significant difference in forage DM intake between the control and supplemented treatments. CP and NDF digestibility of supplemented treatments were significantly  greater than on the control diet but there were no differences among supplement treatments. Milk yield and 3.5% FCM were not significantly different among treatments. Milk protein percentage was higher for CH or CHSH treatments than for the control, but there were no differences among CH and CHSH treatments.

It is concluded that feeding cassava hay in combination with stylo 184 hay as a supplemental protein source could be a potentially valuable strategy in small-holder dairy farming systems in the tropics, as an alternative to balanced concentrates for milk production.

Key words: Cassava hay, concentrate, milk yield, milk composition, stylo 184 hay, tropics


Concentrate supplementation for lactating dairy cows has been practiced by many small-holder farmers in the tropics by giving 0.5 kg concentrates per 1 kg of milk (ratio of 1:2), as a rule of thumb, without taking into account the nature of the roughage used and the actual nutrient requirements (Wanapat M, personal communication). In some areas of Thailand, concentrate use was found to be even higher than 1:1 concentrate to milk yield, which could possibly result in rumen acidosis especially when effective sources of fiber are not available. High concentrate use eventually results in high production costs, of the order of 70% of the total production cost according to Wanapat (1990).  It is, hence, imperative to find means to reduce feed cost.

Cassava hay was reported to be a good source of high protein roughage and has been used as a supplement to improve milk production and quality (Wanapat et al 1997, 1999; Wanapat 2003). Providing a good source of roughage like cassava hay could possibly increase the ratio of protein to energy, and hence could increase productivity in ruminants (Leng 1997). It has been reported that the most economic way to improve energy intake and performance of animals fed on crop residues is to supplement them with good quality forages, including forage legumes (Topps 1997) such as Stylosanthes guianensis CIAT 184, commonly known as "s" (Horne and Stur 1999). At the present time Stylo 184 is widely used in tropical countries (Mannetje and Jones 1992). Stylo 184 was introduced to Thailand in 1993 to evaluate growth and biomass yield, planted at 50 x 30 cm spacing between rows and plants. It was found that it could grow well and produce 12 to 17 tonnes DM/ha/year with 14 to 18 % crude protein in the DM, and could be preserved as hay with high palatability for ruminants (Satjipanon et al 1995).

The objectives of this experiment were to investigate the effect of supplementation with cassava hay and stylo 184 hay to replace concentrates in the diets of crossbred  dairy cows fed Ruzi grass (Brachiaria ruziziensis) as the basal forage in a cut-and-carry system.

Materials and methods

Location of the experiment

The experiment was conducted in Muang District, Mahasarakham, at eight small-holder dairy farms. Researchers and farmers worked closely in a participatory programme from April to December 2002. The experimental site is located 470 km northeast of Bangkok in Northeast Thailand, about 70 km southeast of Khon Kaen..

Dairy farmer selection and training

Eight dairy farmers were selected to participate in this research. The selection was based on farmer qualifications (eg: farmer's willingness, attitude, experience, location, opportunity and potentiality for expansion of technology), their own support and collaboration to participate in the research. Prior to the commencement of the experiment, a one-day training program was held by the researchers and extension officers at Mahasarakham Provincial Livestock Office covering details such as how to feed supplements,  how data should be collected, demonstrations of feed preparation, and dairy cow nutrition. Regular visits to the farms by researchers and extension officers followed, and in addition regular discussions and demonstrations were held. Participating farmers were also to visit other farmers during the demonstrations, which offered a real practical perspective and farmer-to-farmer interactions. As a result of this participation and demonstration scheme, the farmers could learn more effectively and accepted the technology more readily, especially the practical details of the feed preparation, feed establishment, feeding method and feeding management.

Demonstration plots

Eight individual plots sized 20*20 m were established at eight participating farms in late May 2002 in order to demonstrate the growing and processing of the components of the feeding system.  Researchers provided planting materials such as cassava stem, stylo 184 seed and seedlings, and fertilizer. The farmers were also given recommendation on planting, weeding, fertilization, harvesting, hay making and feed supplementation.

Hay production

Cassava hay and stylo 184 hay were produced from cassava (Manihot esculenta ,Crantz) variety "Rayong 72" and stylo 184 (Stylosanthes guianensis CIAT 184) which were planted in late May 2002 at the Experimental Farm of Khon Kaen Animal Nutrition Research and Development Center. The center is located 449 km northeast of Bangkok in Northeast Thailand (16.2 oN and 102.5 oE) at an altitude of 166 m above sea level.

Cassava planting and hay making followed the methods of Wanapat et al (1997, 2000): The cassava crop was planted using stems with 10030 cm spacing between rows and stems. A basal complete fertilizer (188, 188, 188 kg/ha of N, P and K, respectively) was applied at planting. The entire plot area was kept weed-free with hand hoeing at 20 and 75 days after planting and whenever necessary. The foliage was harvested at 3 months after planting in late July 2002 by breaking the cassava stem at 15 cm above the ground.

Planting of stylo 184 and hay making followed the methods of the Department of Livestock Development (2002). Two weeks prior to planting, dolomite was incorporated into plots at the rate of 625 kg/ha. One-month-old seedlings of Stylosanthes guianensis CIAT 184 were planted at 100 x 30 cm spacing between rows and plants. A basal complete fertilizer (188, 188, 188 kg/ha of N, P and K, respectively) was applied at planting, with an additional application of triple superphosphate (125 kg /ha of P) in early July 2002. The entire plot area was kept weed-free with hand hoeing at 20 and 75 days after planting and whenever necessary. The crop was harvested at 3 months after planting in late July 2002 by cutting 15 cm above the ground.

Prior to drying, both forages were chopped to average length of 5 cm using a tractor mounted "Mizubishi" cutter (PTO rotating at 2000 rev./min) then sun-dried for 2 days to attain a moisture content less than 10%. Representative samples of cassava hay and stylo 184 hay were taken for chemical analysis. The hay was collected, packaged in plastic bags and allocated to participating farmers according to treatments to use as supplement for selected lactating dairy cows during August- October 2002.

Animal management

Prior to the commencement of the experiment, sixteen multi-parous Holstein-Friesian crossbred dairy cows (2 cows/farmer), in mid-lactation (98119 days in milk) and with a mean live weight of 420 kg at the beginning of the experiment were selected. The experiment lasted for nine weeks. The first week was used for diet adaptation and the following eight weeks were used for measurements of milk yield and composition. In the last three days of the experiment, forage DM intakes were measured and whole-tract diet DM digestibility was also measured using acid-insoluble ash (AIA) as an internal indicator.

Experimental design and measurements

The experiment was a randomized complete block design (RCBD) with four dietary treatments and four animals per treatment. At the beginning of the experiment, the cows were blocked according to their days in milk and previous lactation into four blocks of four animals each. Within a block, the animals were each randomly allotted to one of the four dietary treatments. The diets comprised a basal roughage of Ruzi grass (Brachiaria ruziziensis) in a cut-and-carry system fed ad libitum with four dietary treatments as follows (CH= cassava hay, SH= stylo 184 hay; ratio of CH+SH is 1:1):

Feeding of the cows was done by participating farmers. Cows were hand milked twice daily, at 07:00 h in the morning and 16:00 h in the evening. The daily morning and evening milk yields of individual cows were recorded. About 100 ml samples of a thoroughly mixed composite of milk (morning and afternoon) of individual cows were taken weekly. The concentrate allowance was given  in two equal amounts during milking time. The CH and SH hay (treatments 1CH, 1SH and 2SH) were given at the same time. The animals had free access to water and mineral blocks. During the last three days prior to termination of the experiment, the basal diet offered was fed in amounts based on the previous day's intake, allowing for a 20% refusal. Daily forage intakes and refusals were weighed each morning and recorded individually before fresh material was given to the animals; fecal samples were collected daily from the rectum for the last three days from each cow, composited, dried and ground. The health condition of the cows was observed daily. Samples of concentrate were taken during the last week of the feeding period.

Chemical analyses

Samples of the cut grass, hay and faeces were dried in a forced-air oven (60C) for 48 h, and ground in a Wiley mill to pass a 2 mm screen and then analysed for dry matter (DM), organic matter (OM), ash and nitrogen (Kjeldahl-N) by the AOAC (1990) procedures. Neutral detergent fiber (NDF), acid detergent fiber (ADF) and lignin (ADL) were determined by the methods of Van Soest and Robertson (1991). Acid-insoluble ash (AIA), measured by the procedure of Van Keulen and Young (1977), was used as an internal indicator to calculate digestion coefficients of the feed. In addition, samples of cassava hay were analyzed for condensed tannins using the vanillin-HCl method (Burns 1971, as modified by Wanapat and Poungchompu 2001). The milk samples were analyzed for fat, protein, lactose, solids-not-fat (SNF), and total solids by using a 'Milko-Scan' (Model 133 V37 GB).

Milk quality analyses were carried out in the Milk Quality Control Laboratory of the Dairy Farming Promotion Organization of Thailand (DPO), in the Northeast Region. Fecal and feed sample analyses were carried out in the Animal Nutrition Laboratory of the Khon Kaen Animal Nutrition Research and Development Center and the Ruminant Nutrition Laboratory of the Department of Animal Science, Faculty of Agriculture, Khon Kaen University.

Statistical analyses

The various data were subjected to the analyses of variance (ANOVA) procedure for a randomized complete block design experiment using the general linear models (GLM) of the SAS System for Windows (SAS 6.12, TS level 020, SAS Institute). Probabilities less than 0.05 were considered significant. Treatment means were compared using Duncan's New Multiple Range test (Steel and Torrie 1980). The statistical model was:

yij= m + tx i + bl j + eij

where m is the overall mean, tx i the ith treatment effect, bl j the jth block effect, and eij is the experimental error of treatment i in block b.


Chemical composition of the experimental feeds

Table 1: Chemical composition of the feeds (on DM basis except for DM which is on as-fed basis)










Ruzi grass


















Cassava hay









Stylo 184 hay









DM = dry matter, OM = organic matter, CP = crude protein, NDF = neutral detergent fiber, ADF = acid detergent fiber, ADL = Acid detergent lignin, CT = condensed tannins; ND = not determined.
Feed intake

There was no difference in intake of cut grass DM between the control and supplement treatments (Table 2). As concentrate intake was determined by milk yield, which was similar for all treatments, total DM intake was higher for 1CH and 1CHSH, which were given in addition to the concentrates (fed at 1 kg per 2 litres milk), and the same for 2CHSH (in which concentrates were reduced to 1 kg per 3 litres milk), compared with the control (Table 2).

Table 2: Mean values for DM intake (kg/day) of milking cows fed cut grass,  concentrates and cassava hay (CH) or stylo 184 hay (SH)


























Total as %  LW





Nutrient digestibility

Digestibility coefficients of crude protein and NDF were higher (P<0.05) on the supplement treatments than on the control, but did not differ among supplements (Table 3).

Table 3: Mean values for coefficients of apparent digestibility in milking cows fed cut grass and partial replacement of concentrates by cassava hay (CH) and stylo 184 hay (SH)





































a,b,c Means  without common superscripts in the same row are different at p<0.05

Milk yield and milk composition

Milk yield and 3.5% FCM were not different among treatments (Table 4). Milk protein percentage for cows given CH or CH-SH was  higher than for cows on the control treatment, but there were no differences among supplement treatments. There were no differences in fat, lactose, solids-not-fat and total solids percentages among treatments.

Table 4: Mean values for milk yield and composition of milking cows fed cut grass and partial replacement of concentrates by cassava hay (CH) and stylo 184 hay (SH)







Milk production, kg/d 





   Milk yield






   3.5% FCM1/






Milk composition, % 

























   Total solids       






a,b Means  in same row without common superscript are different at p<0.05
1/FCM = fat corrected milk, 3.5 % FCM = 0.4*(kg of milk) +15*(kg of fat).


The crude protein  and condensed tannin contents of the cassava hay were in a similar range to those reported by Wanapat et al (1997) and Wanapat (1999). Likewise crude protein percentage of stylo 184 hay was found to be similar to that reported by Satjipanon et al (1995).

The main finding from this experiment was that when cassava hay in combination with stylo hay replaced 33% of the concentrate (treatment 2CHSH), there was no reduction in milk yield and an apparent improvement in milk protein percentage in crossbred dairy cows. Several reports have indicated that cassava foliage is a good source of bypass protein (Ffoulkes and Preston 1978; Wanapat et al 1997), presumably because the condensed tannins act to protect the protein from fermentation in the rumen, thus increasing the supply of amino acids to the small intestine.  The improvement in milk protein percentage could have resulted from this process.

Supplementation of CH solely or in combination with SH did not increase total DM and forage intake. This result is in contrast with the finding of Duong Nguyen Khang and Wiktorsson (2000), Wanapat et al (2000a) and Nguyen et al (2002). The reasons could possibly be the difference in the basal diet, as in the research cited by these authors, urea-treated rice straw was the  basal diet, thus the effect of the cassava hay was to improve markedly the "bypass" protein status of the diet. In the present study the cassava+stylo hay (treatment 2CHSH) replaced an equivalent amount of balanced dairy concentrate thus there was likely to be only a marginal effect on the bypass protein supply.


When hay made from cassava and stylo 184 foliages (50:50 basis) replaced 33% of the concentrate allowance (concentrate: milk ratio  reduced from 1:2 to 1:3) there was no change in milk yields of crossbred dairy cows, but protein percentage in milk was increased.


The Swedish International Development Agency /Swedish Agency for Research Cooperation with Developing Countries (SIDA/SAREC) is gratefully acknowledged for funding of this research. Special thanks are extended to eight participating farmers who allowed the research to be conducted on their farms. The authors thank the farm crew of Khon Kaen Animal Nutrition Research and Development Center  for their enthusiastic assistance in cultivation, harvesting, hay production and allocation to participating farmers. The authors also express appreciation to staff of the Milk Quality Control Laboratory of the Dairy Farming Promotion Organization of Thailand (DPO), the Northeast Region for milk quality test. Appreciation is given to staff of the Ruminant Nutrition Laboratory of the Department of Animal Science, Faculty of Agriculture, Khon Kaen University for assistance in laboratory procedures. The assistance of the following MSc. and PhD. students at Khon Kaen University is also appreciated: Mr. Anan Petlum, Ms. Siwaporn Wora-anu, Mr. Sitisak Khampa, Mr. Pin Chanjula and Ms. Gina Granum.


AOAC  1990  Official Methods of Analysis 15th Ed. Association of Official Analytical Chemists, Arlington, Virginia, USA.

Burns  R 1971  Method for estimatation of tannin in grain sorghum. Agronomy Journal, 63, 511-512.


Department of Livestock Development 2002  Forage legume information.

Duong Nguyen Khang and Wiktorsson  H  2000  Effect of cassava leaf meal on ruminant environment of yellow cattle fed urea-treated paddy straw. Asian-Australian Journal of Animal  Science 13, 1102-1108.

Ffoulkes D and Preston T R 1978  Cassava or sweet potato forage as combined sources of protein and roughage in molasses based diets: effect of   supplementation with soybean meal. Tropical Animal Production (3) :186-192

Horne  P M and Sturr  W W 1999
 Developing forage technologies with smallholder farmers-how to select the best varieties to offer farmers in Southeast Asia. ACIAR Monograph.


Leng  R A  1997  Tree foliage in ruminant nutrition. FAO Animal Production and Health Paper 139


Mannetje  L 't and Jones  R M  1992  Plant Resources of Southeast Asia.No.4 Forages. pp. 300.


Nguyen  T T H, Wanapat  M, Wachirapakorn  C and Pakdee P  2002  Effects of initial cutting and subsequent cutting on yield and chemical compositions of cassava hay and its supplementation on lactating dairy cows. In: Proc. Agriculture Conference, Narasuan University, Pitsanuloke, Thailand, July 26-30.


SAS  1998 User's Guide: Statistics, Version 6.12th Edition. SAS Institute Inc., Cary, NC.


SatjipanonC, Jinosaeng V and Susaena V 1995 Forage seed production project for Southeast Asia, Annual report 1993-1994. KhonKaen Animal Nutrition Research Center, Department of Livestock Development, Ministry of Agriculture and Cooperative. pp. 124-131.


Steel  R G D and Torrie  J H  1980  Principles and Procedures of Statistics: A Biometrical Approach. McGraw-Hill, New York.


Topps  J H  1997  Forage legumes as protein supplements to poor quality diets in the semi-arid tropics. In: Wallace, R.J and Lahloukassi, A. (Eds.), Rumen Ecology Research Planning. Proceedings of workshop held at ILRI, 13-18 March 1995, Addis Ababa, Ethiopia.


Van Keulen  J and Young  B A  1977 Evaluation of acid-insoluble ash as a natural marker in ruminant digestibility studies. Journal of  Animal  Science 44, 282.


Van Soest  P J and Robertson  J B  1991  Methods of dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. Journal of  Dairy Science 74, 3583-3597.

Wanapat  M  2003
 Manipulation of cassava cultivation and utilization to improve protein to energy biomass for livestock feeding in the tropics. Asian-Australasian Journal of  Animal  Science, 16, 463-472.

Wanapat   M  1999 Feeding of Ruminants in the tropics based on Local Feed Resources. Khon Kaen Publ. Comp. Ltd., Khon Kaen, Thailand. pp  236

Wanapat  M  1990 
Nutritional Aspects of Ruminant Production in Southeast Asia with Special Reference to Thailand. Funny Press, Ltd., Bangkok, Thailand.

Wanapat  M,  Petlum  A and Pimpa  O  2000 
Supplementation of cassava hay to replace concentrate use in lactating Holstein-Friesian crossbreds. Asian-Australasian
Journal of  Animal  Science, 13, 600-604.

Wanapat  M,  Pimpa  O,  Petlum  A and Boontao  U  1997 
Cassava hay: A new strategic feed for ruminants during the dry season. Livestock Research for Rural Development, 9(2):

Wanapat  M
and Poungchompu  O  2001 
Method for estimation of tannin by vanillin-HCl method. Department of Animal Science, Khon Kaen University, Khon Kaen 4002, Thailand.

Received 23 August 2003; Accepted 10 October 2003

Go to top