Dairy production and nutritional status of lactating buffalo and cattle in small-scale farms in Terai, Nepal

Y Hayashi, S Shah*, S K Shah*and H Kumagai

GraduateSchool for International Development and Cooperation, Hiroshima University, Higashi-Hiroshima 739-8529, Japan
yoshiha@hiroshima-u.ac.jp
*Institute of Agriculture and Animal Science, Tribhuvan University, Rampur, Chitwan, Nepal

Abstract

Ten small-scale farms were surveyed to identify the feeding traits, milk productivity and nutritional status of lactating buffalo and cattle in Terai, Nepal. Constituents and dry matter (DM) of feed supplied, body condition score (BCS), heart girth (HG), bodyweight (BW), milk yield (MY) and plasma metabolites were obtained in the pasture-sufficient, pasture-decreasing and fodder-shortage periods which were August, November and March, respectively. Milk yield of 305-day lactation was estimated by the MY of seven consecutive days a month. Rice straw and wheat bran were utilized as main basal diets throughout the survey. The variance of feed constituents among the periods induced different supplies of CP, NDF and TDN.

The average concentrations of CP and NDF in the buffalo feed were significantly higher in the pasture-sufficient period than in the other periods (9.6% vs. 8.0% and 65.3% vs. 62.1%, respectively, p<0.01). The corresponding content of TDN in the pasture-sufficient period was similar to that in the pasture-decreasing period (51.1% vs. 50.9%), and higher than that in the fodder-shortage period (51.1% vs. 49.4%, p<0.01). The average supply of TDN was lower in buffalo than in cattle (50.5% vs. 52.3%, p<0.05). The HG and BW of buffalo in the pasture-sufficient period was similar to those in the pasture-decreasing period, and higher than those in the fodder-shortage period (p<0.05). The MY of buffalo was significantly higher in the pasture-sufficient period than in the other periods (7.7 liters/day vs. 6.5 liters/day, on an average, p<0.01). No significant difference was observed in the 305-day MY of buffalo among the calving periods. However, the buffaloes in their second and above lactation had significantly higher 305-day MY than those in the first lactation (2032 liters vs. 1619 liters, on an average, p<0.05). The average concentration of BUN in buffalo was significantly higher in the pasture-sufficient period than in the other periods (p<0.05). The average BUN concentration was higher in buffalo than in cattle (13.4 mg/dl vs. 6.0 mg/dl, p<0.01), while the average contents of total cholesterol and NEFA were higher in cattle than in buffalo (267 mg/dl vs. 127 mg/dl and 0.25 mEq/l vs. 0.19 mEq/l, respectively, p<0.01). The various supplies of CP, NDF and TDN among the periods might have affected MY and nutritional condition in buffalo. It is likely that the higher supplies of CP for buffalo in the pasture-sufficient period improved the nutritional status for milk production.

Key words: Buffalo, cattle, milk production, Nepal, nutritional status

Introduction

The agriculture sector contributed 40.1% of Nepalese growth domestic product (GDP), equivalent to US$5.8 billion in 2003 (World Bank 2004). Livestock was an important agriculture sub-sector in the country, accounting for about 30% of the agricultural GDP. The country raised about 3.7 million buffaloes and 7.0 million cattle. Nationwide in 2003 small-scale dairy farmers produced an approximately 1.2 million MT of milk, of which 65.5% was from buffalo and 28.3% was from cattle (FAO 2005).

Terai is a low-altitude, southern plain region of Nepal and considered as the country's main granary, even though it constitutes only about 14% of the country's total area. Dairy farming integrated with crop production is preferred by farmers in this region, and it has provided about 35% of the national milk production from improved breeds of buffalo and cattle in 2003. In Chitwan, farmers maintained about 35 thousand lactating buffaloes and 19 thousand lactating cattle with a combined production of 52 thousand MT of milk in 2003, which accounted to 4.3% of the national milk production (Agri-information and Communication Center Kathmandu Nepal). Although Terai region is a major contributor to the domestic milk production, expensive rations and fodder shortages are common constraints attributable to small farm size (0.6 ha/farm) (Karki et al 1993; Sharma et al 1994). Recently, Hayashi et al (2004) reported the effects of pasture environments especially on cattle milk production in this region. Those parturitions  in the fodder-shortage period from March to May produced less milk than in the other periods. Although the effects of pasture and fodder environments on cattle dairy production in small-scale farms have been identified, a comprehensive study on dairy nutrition especially in buffalo remains obscure. Hence, the present study was conducted to identify the feeding traits for lactating buffalo and cattle in small-scale farms, and to determine the relationships among feeding traits, milk productivity and nutritional status of these animals in Terai, Nepal.

Materials and Methods

Location and climate

Ten small-scale farms with an adequate number of buffalo and a small number of cattle in lactation were selected for the survey from August 2003 to July 2004 in a village of Chitwan, Terai, Nepal. The village is situated 180 km southwest of Kathmandu, and milk production accounts for a large part of the local economy.

This region has a subtropical climate, and during the survey the average temperatures ranged from 15.2˚C in January 2004 to 29.7˚C in July 2003. Total rainfall over the same period was 2240 mm, and the average relative humidity was 87.1% (National Maize Research Center, Rampur, Nepal). The monthly lowest and highest precipitations during the survey were 0 mm in November 2003 and 549 mm in August 2003, respectively. There are three periods based on environments of pasture and fodder: the pasture-sufficient period, characterized by increased pasture from June to October; the pasture-decreasing period, characterized by decline of pasture due to rainless and cool climate from November to February; and the fodder-shortage period, characterized by scarcity of fodder with dry climate from March to May.

Data and sample collection

Lactating buffalo and cattle were tethered in cowsheds and hand-milked twice daily. Number, breed, age, parity and last calving date of the animals were recorded in August 2003. Average head and parity of lactating buffalo were 2.8 ranging from 1 to 6 and 4.6 ranging from 1 to 18, respectively. The breed of buffalo was Murrah-cross. Average head and parity of lactating cattle were 0.3 ranging from 0 to 2 and 4.7 ranging from 3 to 8, respectively. The breeds of cattle were Holstein-cross and Jersey-cross. Daily milk yield (MY) of each animal was recorded in seven consecutive days a month from August 2003 to July 2004. Milk yield of 305-day lactation from each animal was estimated using the equation of Wood (1969).

Lactating buffalo and cattle were tethered in cowsheds and hand-milked twice daily. Number, breed, age, parity and last calving date of the animals were recorded in August 2003. Average head and parity of lactating buffalo were 2.2 ranging from 2 to 4 and 4.0 ranging from 1 to 18, respectively. The breed of buffalo was Murrah-cross. Average head and parity of lactating cattle were 0.7 ranging from 0 to 3 and 4.1 ranging from 1 to 8, respectively. The breeds of cattle were Holstein-cross and Jersey-cross. Daily milk yield (MY) of each animal was recorded in seven consecutive days a month from August 2003 to July 2004. Milk yield of 305-day lactation from each animal was estimated using the equation of Wood (1969).

Feed resource samples from representative farms were collected with the records of dry matter (DM) supplied for each animal in August and November 2003, and March 2004. The samples were dried to measure DM content and ground to pass through a 1-mm sieve. Composite representative samples of feed resource were made by mixing of the same amounts of the original samples. Concentrations of DM and crude protein (CP) in the representative samples were analyzed by the method of AOAC (1990). Acid detergent fiber (ADF) and neutral detergent fiber (NDF) contents of the samples were determined by the method of Van Soest (1973) and Van Soest et al (1991), respectively. Total digestible nutrient (TDN) of the samples were estimated using the following equations reported by Martin (1985) and Chandler (1990):

TDN (%) in straw = 96.35 - 1.15 x ADF (%)
TDN (%) in native grass = 105.2 - 0.68 x NDF (%)
TDN (%) in supplemental resources = 81.41 - 0.48 x NDF (%)

Concentrations of CP, NDF and TDN in total feed supplied for each animal were calculated from the amount of DM in each feed resource supplied and composition in the representative samples in each period.

Measurement of body condition score (BCS) assessed by the method of Ferguson et al (1994) and heart girth (HG), and blood sampling were conducted at the same time of feed collection. Body weights (BW) of buffalo and cattle was estimated by their HG using the following equation developed by Kumagai et al (2003):

BW of buffalo (kg) = 603 / (1 + 321e^-0.036HG(cm))
BW of cattle (kg) = 605 / (1 + 243e^-0.035HG(cm))

Blood plasma was collected after a centrifugation and stored at -20 ˚C for analyses of the metabolites. Total protein concentrations were determined by a refract meter (SPR-Ne, Atago Co., Ltd., Japan). Concentrations of albumin, blood urea nitrogen (BUN), glucose, total cholesterol and non-esterified fatty acid (NEFA) were analyzed using diagnostic kits (Albumin-HRII, L type Wako UN, Glucose-HRII Wako, L type Wako CHO·H and NEFA-HR, Wako Pure Chemical Industries, Ltd., Japan). Globulin concentrations were calculated by subtracting the concentrations of albumin from total protein.

Statistical analyses

Statistical analyses were conducted using Excel Statistics^TM (Esumi Co., Ltd., Japan). The differences in mean values of feeding traits, nutritional status and productivity according to the survey time, and the effects of calving period and parity on the 305-day MY in buffalo and cattle were analyzed by Student t-test and Duncan's multiple range test (1955).

Results

Dry matter and rates of roughage to supplements supplied for the buffalo and cattle are presented in Table 1.

All of the farmers fed rice straw as a main basal diet for the animals throughout the survey, and its amount for buffalo was the lowest in November (p<0.05). Native grass was also utilized as a roughage resource, and the buffalo was fed the highest amount in August (p<0.01). However, the grass was not supplied for both the buffalo and cattle in November. Although bamboo leaf, corn stover and mustard straw were fed as alternative roughage, the amount was large neither in November nor in March. Wheat bran was fed as a main supplemental feed for both the animal species. Brewery waste, commercial feed, corn flour and rice polish were also fed as supplemental feeds for the animals. The amounts of wheat bran and commercial feed for buffalo were higher in November than in August and March (p<0.05).

Concentrations of CP, NDF and estimated TDN of the feed resources are presented in Table 2.

Bamboo leaf had the highest content of CP among the roughage resources. The lowest NDF and the highest TDN among the roughage were 50.0% and 71.4% in native grass, respectively, on an average. Commercial feed contained the highest concentrations of CP and TDN among the supplements showing 23.8% and 68.7%, respectively, on an average. The average concentrations of CP in wheat bran and brewery waste were 16.3% and 20.9%, respectively, which were higher than those in corn flour and rice polish showing 11.2% and 14.4%, respectively. Commercial feed contained higher CP in August than in November and March.

Estimated concentrations of CP, NDF and TDN in total feed supplied for the buffalo and cattle are presented in Table 3.

The average concentrations of CP and NDF in the total feed given to buffalo were higher in August than in November and March (p<0.01). The corresponding content of TDN in August was similar to that in November, and higher than that in March (p<0.01). No significant difference was observed in the supplied concentrations of CP and NDF between buffalo and cattle showing 8.5% and 7.9% in CP, and 63.1% and 65.6% in NDF, respectively, on an average. However, the corresponding content of TDN was significantly lower in buffalo showing 50.5% than in cattle showing 52.3% on an average (p<0.05).

Parity, BCS, HG, BW, lactation days and MY of the buffalo and cattle are presented in Table 4.

The average parities of buffalo and cattle were 4.2 and 4.1, respectively. The HG and BW of buffalo were higher in August and November than in March (p<0.05). In contrast, the lactation days of buffalo were longer in March than in August and November (p<0.01). The buffalo yielded more milk in August than in November and March (p<0.01). The changes of lactation days and MY in cattle during the survey were similar to those in buffalo. The BCS, HG and BW were significantly higher in buffalo than in cattle (3.6 vs. 3.1 [p<0.05], 193 cm vs. 162 cm [p<0.01] and 459 kg vs. 329 kg [p<0.01], respectively, on an average).

Plasma metabolite concentrations of the buffalo and cattle are presented in Table 5.

The concentrations of total protein, globulin and BUN in buffalo were significantly higher in August than in March (p<0.01). In contrast, the content of total cholesterol in buffalo was significantly lower in August than in March (p<0.05). The BUN concentration was higher in buffalo than in cattle (13.4 mg/dl vs. 6.0 mg/dl on an average, p<0.01). The average contents of total cholesterol and NEFA were higher in cattle than in buffalo (267 mg/dl vs. 127 mg/dl and 0.25 mEq/l vs. 0.19 mEq/l, respectively, p<0.01).

Effects of calving period and parity on 305-day MY of the buffalo and cattle are presented in Table 6.

No significant difference was observed in the 305-day MY of buffalo among the calving periods. The buffaloes in their second and above lactation had a significantly higher 305-day MY than those in the first lactation (p<0.05). The positive correlation was observed between the 305-day MY and HG, and BW in buffalo (r = 0.43 and 0.45, respectively, p<0.05).

Discussion

In Terai region, rice was harvested twice a year, in June and November, leading to its utilization as a major roughage resource for buffalo and cattle throughout the survey (Table 1). Native grass was fed as additional roughage in the pasture-sufficient and fodder-shortage periods attributable to higher temperature and more rainfall in these periods than in the pasture-decreasing period. Since the supply of total roughage DM in November was significantly lower than that in the August and March (11.6 kg vs. 13.7 kg [p<0.01] and 12.8 kg [p<0.05], respectively, on an average), the amounts of wheat bran and commercial feed were likely to increase to compensate the reduction of roughage DM in November. Fodder trees and homemade concentrate mixture, locally known as Kundo, which were common in some areas in Nepal were not utilized in the survey area. The feed constituent was different among the pasture-sufficient, pasture-decreasing and fodder-shortage periods. The different constituents probably caused the variance in supplies of CP, NDF and TDN (Table 3). The lowest supply of NDF for buffalo in November was probably due to the lowest roughage-supplements rate in this time. Since the supplies of CP and TDN for buffalo decreased over time from the pasture-sufficient period to the fodder-shortage period, the intake of these nutrients in buffalo might have declined equally.

The requirements of CP and TDN for lactating buffalo were 5.4 g/kgBW^0.75 and 35.3 g/kgBW^0.75 daily for maintenance, respectively; 90.3 g/kg and 406 g/kg for 1 kg milk containing 6 % fat, respectively (Paul et al 2002). If it was supposed that the milk fat content of the buffalo was 6%, the average requirements of CP and TDN for the buffalo were 1232 g and 6632 g in August, 1129 g and 6179 g in November, and 1103 g and 6030 g in March, respectively. The daily requirement of NDF for buffalo was reported in the following equation (Bartocci et al 2002):

NDF (g/day) = 8864.3 - 198.92 x MY

Based on this equation, the average requirements of NDF for the buffalo were 7333 g, 7571 g and 7591 g in August, November and March, respectively. In contrast, the average supplies of CP, TDN and NDF for the buffalo were 1898 g, 9848 g and 12622 g in August, 1274 g, 8154 g and 9801 g in November, and 1396 g, 8610 g and 10967 g in March, respectively. Since these supplies were higher than the requirements in all of the surveys, the buffalo was considered to receive sufficient nutrients throughout the year. Under this feeding condition, the buffalo MY in August was significantly higher than that in the other times (p<0.01). One possible reason for the higher MY might be the early lactation stage since 53.8% of buffalo were in 1-87 days of lactation in August. The other might be the higher supply of CP in August than in the other times. The higher concentrations of BUN in buffalo in August probably supported a better nutritional condition for MY than the other times (Table 5). A reduction of nutrient intake was considered to lower the concentrations of plasma total protein and BUN in buffalo in March.

Although the supply of CP was decreased in November (Table 3), the BCS, HG and BW in buffalo in this time was not affected (Table 4). The buffalo with enough nutrient ingestion probably utilized excess nutrients for body reserves and milk production. However, if the intake of nutrient was reduced in the pasture-decreasing period, the buffalo possibly preceded to assimilate the nutrients for body reserve rather than milk production. Thereafter, a further decline of nutrient supply was occurred in the fodder-shortage period, and the buffalo might have started to catabolize body tissue, resulting to a decrease in BCS, HG and BW.

The 305-day MY of Murrah buffalo was reported that the averages were from 1645 kg to 2130 kg, and the elite of the same breed yielded over 3000 kg in Europe (Alexiev 1992). Patro and Bhat (1979) reported that the mean 300-day MY of Murrah ranged from 1573 kg to 1964 kg in India. The present study showed that the average 305-day MY of buffalo was 1932 liters ranging from 1259 liters to 3037 liters. These findings were considered to coincide with the previous reports. Additionally, the average 305-day MY of buffalo did not significantly differ from that of cattle (2314 liters). Thus, buffalo might be utilized as a major dairy animal in the area concerned. A positive correlation was observed between the 305-day MY and BW in buffalo in the present study. The BW of buffalo that calved in the pasture-decreasing period was higher than that in the pasture-sufficient period (472 kg vs. 447 kg, p<0.05). Therefore, the 305-MY of buffalo that calved in the pasture-decreasing period tended to be higher than that in the pasture-sufficient period. Besides, the higher parity resulted a higher 305-day MY in buffalo attributable to a positive correlation between parity and BW in the present study (r = 0.22). The effect of parity on the 305-day MY agreed with earlier reports that the 305-day MY was higher in the second parity than in the first (Cady et al 1983; Tekerli et al 2001).

The variance of feed constituents among the three periods caused the difference in nutrient supplies for cattle (Table 3). The requirements of CP and TDN in diet on a DM basis for lactating cattle with BW less than 400 kg and daily MY less than 8 liters were reported to be 13% and 63%, respectively (NRC 1989). However, the supplies of these nutrients for the cattle were lower than the requirements during the entire survey. Thus, the cattle was considered to have insufficient nutrient supplies throughout the year. The lower concentration of BUN than the normal range in cattle (8-23 mg/dl in BUN, Whitaker et al 1995) were possibly affected by the lower intakes of CP and TDN (Table 5).

Even though there was no difference of CP supply between buffalo and cattle (Table 3), the higher BUN content in buffalo than in cattle was observed (p<0.01) (Table 5). Norton et al (1979) reported that plasma urea concentration was higher in buffalo than in cattle, and suggested that there was a higher renal resorption of urea in buffalo. Kawashima et al (2000a) reported ammonia nitrogen in rumen fluid was higher in buffalo. Thus, buffalo possibly have a better ability to maintain the recycling of urea and mobilize energy from body tissue protein than cattle even during scarcity of nutrients. The content of NEFA in the present study was higher in cattle than in buffalo (p<0.01). Kawashima et al (2000b) reported that NEFA concentration was higher in cattle than in buffalo under the same feeding management. Cattle was likely to mobilize more energy from fat than buffalo. The concentration of plasma total cholesterol was higher in cattle than in buffalo (p<0.01), though a similar feed constituents were supplied to both the animal species. The same finding was reported by Hayashi et al (2004). However, the exact reason is still poorly understood.

In conclusion, the periods divided by the environments of pasture and fodder caused the variance in feed constituents. Although the buffalo was considered to have sufficient nutrient supplies throughout the year, the variance induced the difference in supplies of CP, NDF and TDN. As a consequence, milk production and nutritional status of buffalo were also affected. The buffalo decreased MY from the pasture-sufficient period to the pasture-decreasing period but maintained BW even in the pasture-decreasing period. This might show the buffalo's preference for assimilation of nutrients to body reserve rather than milk production. Additionally, the difference of protein and energy mobilization in the body between buffalo and cattle was suggested in the survey. Research and development of economical and locally available feed resources for the nutrient improvement are necessary. Further studies on how management skills affect the traits of dairy animals are also needed.

Acknowledgements

The study was supported by a Grant-in Aid for scientific research (No. 12505025) from the Japanese Society for the Promotion of Science (JSPS).

References

Alexiev A I 1992 B reeding and management of river buffaloes in Europe, Egypt and Iraq. In: Buffalo Production (Editors: N M Tulloh and J H G.Holmes). Elsevier Science Publishers, Amsterdam, the Netherlands. pp 59-76.

AOAC 1990 Official Methods of Analysis (15th Edition). Association of Official Analytical Chemist, Arlington, Virginia.

Bartocci S, Tripaldi C and Terramoccia S 2002  Characteristics of foodstuffs and diets, and the quanti-qualitative milk parameters of Mediterranean buffaloes bred in Italy using the intensive system. An estimate of the nutritional requirements of buffalo herds lactating or dry. Livestock Production Science. 77:45-58.

Cady R A, Shah S K, Schermerhorn E C and McDowell R E 1983 Factors affecting performance of Nili-Ravi Buffaloes in Pakistan. Journal of Dairy Science. 66:578-586.

ChandlerP 1990 Energy prediction of feeds by forage testing explored. Feedstuffs. 62, 36, pp 12.

Duncan D B 1955 Multiple range and multiple F test. Biometrics. 11:1-42.

FAO 2004 Food and Agriculture Organization of the United Nations, FAO Statistical Database. http://faostat.fao.org/

FergusonJ D, Galligan D T and Thomsen N 1994 Principal descriptors of body condition score in Holstein cows. Journal of Dairy Science. 77:2695-2703.

Hayashi Y, Maharjan K L and Kumagai H 2004 Milk production and nutritional status of dairy cattle and buffaloes of small-scale farms in Terai region, Nepal. In: Proceedings of the 11th Animal Science Congress (Editors: H. K. Wong, J. B. Liang, Z. A. Jelan, Y. W. Ho, Y. M. Goh, J. M. Panandam and W. Z. Mohamad), VolumeII. The Asian-Australasian Association of Animal Production Societies, Kuala Lumpur, Malaysia. pp 30-33.

Karki U, Kolakchhapati M R and Paudyal N 1993  Impact of crossbred cows on the economy of small farmers and their performance in Gitanagar village development committee, Chitwan, Nepal. IAAS Research Report (1985-1991). Institute of Agriculture and Animal Science, Rampur, Nepal. pp 600-608.

Kawashima T, Kurihara M, Sumamal W, Pholsen P, Chaithiang R and Boonpakdee W 2000a Comparative study on rumen physiology between Brahman cattle and swamp buffalo fed with Ruzi grass hay with or without filter cake and rice bran mixture. In: Improvement of cattle production with locally available feed resources in Northeast Thailand (Editor: T Kawashima). Japan International Research Center for Agricultural Science, Tsukuba, Japan. pp 67-73.

Kawashima T, Sumamal W, Pholsen P, Chaithiang R, Boonpakdee W and Kurihara M 2000b Comparative study on energy and nitrogen metabolisms between Brahman cattle and swamp buffalo fed with low quality diet. In: Improvement of cattle production with locally available feed resources in Northeast Thailand (Editor: T Kawashima). Japan International Research Center for Agricultural Science, Tsukuba, Japan. pp 116-122.

Kumagai H, Maharjan K L and Nakao T 2003 Surveys of age, live weight and heart girth of cattle and buffalo in Chitwan, Nepal. In: Proceedings of the 102nd Japanese Society of Animal Science Congress (Japanese). Japanese Society of Animal Science, Tokyo, Japan. pp 45.

Martin N 1985 Agricultural extension service. no. 2637. University of Minnesota, Minneapolis, Minnesota.

Norton B W, Moran J B and Nolan J V 1979 Nitrogen metabolism in Brahman cross, buffalo, Banteng and Shorthorn steers fed on low quality roughage. Australian Journal of Agricultural Research. 30:341-351.

NRC 1989 Nutrient Requirements of Dairy Cattle (sixth edition). Nutrient requirements of domestic animals, no. 3. National Research Council, National Academic Press, Washington, DC. pp 138-147.

Patro B N and Bhat P N 1979 Inheritance of production traits in buffaloes. Indian Journal of Animal Science. 49:10-14.

Paul S S, Asit B M and Nitya N P 2002 Feeding standards for lactating riverrine buffaloes in tropical conditions. Journal of Dairy Research. 69:173-180.

Sharma M, Kharel M, Kolachhapati M R, Dhakal I P, Joshi N P and Gajurel K P 1994 Comparative performance of local and Murrah buffaloes with their crossbreds under farmers' management in Mangalpur, Chitwan. IAAS Research Report. (1992-1993). Institute of Agriculture and Animal Science, Rampur, Nepal. pp 130-136.

Tekerli M, Kucukkebabci M, Akakin N H and Kocak S 2001 Effects of environmental factors on some milk production traits, persistency and calving interval of Anatolian buffaloes. Livestock Production Science. 68:275-281.

Van Soest P J 1973 Collaborative study of acid detergent fiber and lignin. Journal of the Association of Official Analytical Chemists. 56:781-784.

Van Soest P J, Robertson J B and Lewis B A 1991 Methods for dietary fiber. Neutral detergent fiber and non-starch polysaccharide in relation to animal nutrition. Journal of Dairy Science. 74:3584-3597.

Whitaker D A, Kelly J M and Eayres H F 1995 Use and interpretation of metabolic profiles in dairy cows. Department of Veterinary Clinic Studies, University of Edinburgh, Midlothian, UK. pp 13.

Wood P D P 1969 Factors affecting the shape of the lactation curve in cattle. Animal Production. 11:307-316.

World Bank 2004 The World Bank Group, World Development Indicators Database.   http://www.worldbank.org/data/wdi2002/