Milk yield and lactation length in tropical cattle

Bibliography

SHORT COMMUNICATIONS

O. Syrstad

The address of the author is: Norwegian Centre for International Agricultural Development (NORAGRIC), Agricultural University of Norway, PO Box 2, 1432 As, Norway.

Cattle in the tropics have, on average, lower milk yields and shorter lactations than cattle in temperate countries. The difference is caused by both genetic and non-genetic factors.

Cows of temperate dairy breeds under intensive management usually continue to lactate until milk secretion is suppressed by the advanced stage of gestation. If a cow does not dry off naturally, it is usually forced dry five to six weeks before the expected calving date in order to provide a sufficient rest prior to the ensuing lactation. Length of lactation, therefore, is almost completely determined by length of calving interval.

In tropical cattle, however, milk production often ceases several months before next calving and before the depressing effect of gestation on milk production is noticeable. Length of lactation, therefore, is not so greatly influenced by calving interval. Wilson et al. (1987) reported a correlation coefficient of only 0.08 between lactation length and calving interval in Kenana cattle.

Lactation length in tropical cattle is a measure of persistency and it is a heritable trait. Estimates of heritability are few, and these have large sampling errors because they are based on rather small sets of data. Estimates of repeatability, however, suggest that lactation length is heritable to almost the same extent as milk yield. Some of the presumably more reliable estimates of repeatability of milk yield and lactation length in tropical cattle are shown in Table 1.

Several studies have shown a close association between lactation length and milk yield in tropical cattle. Some estimates of the correlation coefficient between the two traits are seen in Table 1. They are all high, ranging from 0.34 upwards.

1 Repeatability of milk yield and lactation length in tropical cattle and correlation between the two traits - Répétabilité des rendements laitiers et des durées de lactation en régions tropicales et corrélation entre ces deux paramètres - Repetibilidad de la producción lechera y duración de la lactancia en zonas tropicales, y correlación entre ambos parámetros

Breed and country	Repeatability		Correlation coefficient	Source
	Milk yield	Lactation length
Kenana, Sudan	0.43	0.19	0.87	Alim, 1960
Nganda, Uganda	0.73	0.42	-	Mahadevan & Marples, 1961
Hariana, India	0.39	0.28	0.69	Singh & Desai, 1961
Butana, Sudan	0.42	0.42	0.86	Alim, 1962
East African Zebu, Kenya	0.55	0.38	0.68-0.81	Galukande, Mahadevan & Black, 1962
Sahiwal crosses, Kenya	0.65	0.33		Mahadevan, Galukande & Black, 1962
Gaolao, India	0.12-0.44	0.20-0.35	0.34-0.67	Patil & Prasad, 1968, 1970
Gir, India	0.40	0.22	0.58	Shulkla & Prasad, 1970
Northern Sudan Zebu, Sudan	0.38	0.29	-	Osman & El Amin, 1971
Deshi, India	0.42	0.18	-	Moulick et al., 1972
Sahiwal, Pakistan			0.62	Ahmed, 1972
Hariana, India			0.67	Ngere et al 1973
Kenana, Sudan	0.47	0.47	0.77	Wilson et al., 1987
Mpwapwa, Tanzania	0.48	0.46	-	Kasonta, 1988
White Fulani, Nigeria	0.32	0.21	0.62	Mrode, 1988

Death of calves is a common reason for short lactations as tropical cattle often fail to let down milk unless stimulated by suckling - Dans les régions tropicales, la mort du veau est une cause fréquente de lactation courte, car les vaches ne donnent leur lait que stimulées par la tétée - La producción de leche del ganado tropical depende del estímulo dado por el ternero. La muerte de este último es causa de lactaciones breves (A)

Death of calves is a common reason for short lactations as tropical cattle often fail to let down milk unless stimulated by suckling - Dans les régions tropicales, la mort du veau est une cause fréquente de lactation courte, car les vaches ne donnent leur lait que stimulées par la tétée - La producción de leche del ganado tropical depende del estímulo dado por el ternero. La muerte de este último es causa de lactaciones breves (B)

2 Factors for extending milk records of lactations terminated by sale or death of cow - Coefficients pour estimer les lactations complètes de vaches vendues ou mortes - Factores de calculo para estimar la lactancia completa en vacas vendidas o muertas

Weeks of lactation	Four-week yield (kg)	Cumulative yield (kg)	Extension factors
1 -4	217.2	-	7.09
5-8	213.8	431.0	3.57
9-12	190.8	621.8	2.48
13-16	171.8	793.6	2.08
17-20	155.1	948.7	1.62
21 -24	141.4	1090.1	1.41
25-28	128.4	1218.5	1.26
29-32	110.2	1328.7	1.16
33-36	89:8	1418.5	1.09
37-40	71.6	1490.1	1.03
41-44	49.8	1539.9	1.00

Source: Mchau and Syrstad, 1991.

Poor milk let down is a serious problem in many crosses of tropical and temperate cattle and short lactations are common - Bien souvent, les vaches issues de croisements entre races tropicales et tempérées ont peu de lait et les lactations courtes sont fréquentes - La escasa producción de leche es un problema grave en los cruces de ganado de zonas tropicales y templadas

The close association between milk yield and lactation length has sometimes been taken as an indication that yield records should be "corrected" for length of lactation. Methods for handling the problem caused by variations in lactation length include the following:

· Records associated with short lactations are excluded from the data. This has been widely practiced in studies done on factors affecting milk yield in tropical cattle. As no standard minimum lactation length has yet been adopted, the various authors have set their own, apparently arbitrary, limits.

Restrictions on lactation length might influence the estimates, not only for lactation length, but also for milk yield, and most probably lead to a downward bias. A frequent reason for short lactations in tropical cattle is that cows fail to let down milk properly. De Alba and Kennedy (1986) estimated the heritability of lactation length at 0.41 when all records were included, but this dropped to 0.08 when only records with normal milk let down were considered.

· Records are restricted to the first 305 days (or another defined interval) after calving. Practised in many temperate countries, the purpose of this is to reduce the variation caused by length of calving interval, but variation associated with length of lactation is reduced as well. The standard period of 305 days has been chosen because this is approximately the normal lactation length of cows calving at intervals of 12 months. In tropical cattle, restricting the lactation records to 305 days would have less effect, as few cows milk for more than 305 days. A shorter interval would be more effective, but it would penalize cows with high persistency.

· Average milk yield per day of lactation is computed and then (if desirable) converted to 305 days, for example, by multiplication. This method ignores the fact that the level of production normally is highest during the first part of a lactation and then decreases gradually over time. Of cows with a given initial yield, those having the shortest lactations will be favoured by this method.

· A correction procedure can be developed by estimating the regression of milk yield on lactation length in a sample of data and using this to construct the prediction equation:

Y_c=Y-b(L-)

where Y is the recorded yield and Y_c the corrected yield in the lactation considered, L is the length of this lactation, is the average length of lactations in the population (i.e. herd or breed) and b is the coefficient of regression of milk yield on lactation length. Alternatively, the records in the sample can be grouped by lactation length, and the correction factors derived as the ratio between average yield in lactations of standard length - for example, 301 to 310 days - and average yield in each of the other groups:

C=Y_r/Y_i

where C is the correction factor (ratio), Y_r is the average yield-of the reference group and Y_i the average yield of the group of lactations of length i.

Both methods overlook the fact that length of lactation is confounded by the inherent producing ability of the cow, i.e. cows with short lactations are usually poor producers. This will inflate the regression of milk yield on lactation length and can lead to seriously biased results. The difference in producing ability between cows with long lactations and those with short lactations will be masked, as the corrected yield is independent of lactation length. Madalena (1990) found that adjusting for lactation length (by regression) removed more genetic-than non-genetic variation.

· Factors (ratios) for extending incomplete lactations (i.e. lactations still in progress or lactations terminated because the cows died or were sold while still milking) are derived by computing the ratios between completed records (usually restricted to 305 days) and partial records of the same lactation. The same factors can be used to adjust complete lactation records (shorter than 305 days) for length of lactation. The underlying assumption is that most short lactations in dairy cattle are caused by management factors rather than by the poor persistency of the cows. This assumption would not be justified in tropical cattle, as they often have short lactations regardless of management. However, cows with short lactations will be favoured less by this method than by the third and fourth methods described above.

· The optimum way of handling variation in lactation length when evaluating milk yield for the purpose of selection is to include lactation length as well as yield in a selection index. This procedure requires that the heritabilities of both traits are known, as well as the phenotypic and genetic correlations between them. If lactation length per se is considered unimportant (of no economic value), the coefficient (weight) of lactation length in the index will be positive or negative depending on the sign of the expression:

h_L r_g - h_Y · r_P

where h_L and h_Y are the square roots of the heritabilities of lactation length and milk yield, respectively, and r_p and r_g are the phenotypic and genetic correlations between the two traits. While several estimates have been made of the phenotypic correlation between lactation length and milk yield (Table 1), very few, if any, reliable estimates of the genetic correlation have been reported. Bastos (1989) estimated the genetic correlation between the two traits in Guzera cattle in Brazil at 0.99, but the data were too limited to provide a reliable estimate. It seems likely, however, that the genetic correlation is higher than the phenotypic. If the heritability of lactation length is slightly lower than that of milk yield, as suggested by the repeatabilities (Table 1), then the expression h_L r_g - h_Y· r_p might be close to zero, in which case the gain in accuracy achieved by including lactation length in the index is marginal.

At present, the best solution might be to accept records of milk yield at their actual value, without any attempts to "correct" for length of lactation. This applies when the lactation has been terminated by normal drying-off. Records of lactations that were short because the cows died or were sold while still in milk should be extended by using appropriate factors. The same factors can be used for lactations that were terminated as a result of the death of the calf (where the calf was necessary for stimulating milk let down) and for lactations still in progress. A set of factors based on records of Mpwapwa cattle in Tanzania are presented in Table 2. The factors were derived simply as the ratio between the average yield of the complete lactation (restricted to 305 days) and the cumulative yield up to the stage in question. These factors are slightly lower than those reported by Rao and Sundaresan (1980), probably because the latter were based on first lactations, which are characterized by a higher persistency than subsequent lactations (Mchau and Syrstad, 1991). Precision would increase slightly if separate sets of factors were used for first and subsequent lactations.

Bibliography

Ahmed, Z. 1972. Genetic analysis of a purebred herd of Sahiwal cattle. Faisalabad, Pakistan, Agricultural University (Ph.D. thesis)

Alim, K.A. 1960. Reproductive rates and milk yield of Kenana cattle in Sudan. J. Agric. Sci. (Camb.), 55: 183-188.

Alim, K.A. 1962. Environmental and genetic factors affecting milk production of Butana cattle in Sudan. J. Dairy Sci., 45242-247.

Bastos, J.F.P. 1989. Genetic and non-genetic factors affecting milk yield and lactation length in a herd of Guzera cows (in Spanish). Arquivo Brasileiro de Medicina Veterinaria e Zootecnica, 41: 185-187; Anim. Breed. Abstr., 58: 82.

De Alba, J. & Kennedy, B.W. 1986. Milk production in the Latin American Criollo and its crosses with the Jersey. In Annual Report, Centre for Genetic Improvement of Livestock, Department of Animal and Poultry Science, Ontario Agricultural College, University of Guelph, Canada. Anim. Breed. Abstr., 56: 10.

Galukande, E.B., Mahadevan, P. & Black, J.G. 1962. Milk production in East African zebu cattle. Anim. Prod., 4: 329342.

Kasonta, J.S. 1988. Population analysis and model calculations for a breeding scheme in the Mpwapwa cattle of Tanzania. University of Hohenheim, Germany. 218 pp. (Dissertation)

Madalena, F. 1990. Crossbreeding effects in tropical dairy cattle. Proceedings of Fourth World Congress on Genetics Applied to Livestock Production, Edinburgh, Scotland, July 1990, Vol. XIV: 310-319.

Mahadevan, P., Galukande, E.B. & Black, J.G. 1962. A genetic study of the Sahiwal grading-up scheme in Kenya. Anim. Prod., 4: 337-342.

Mahadevan, P. & Marples, H.J.S. 1961. An analysis of the Entebbe herd of Nganda cattle in Uganda. Anim. Prod., 3: 2939.

Mchau, K.W. & Syrstad, O. 1991. Production characteristics of Mpwapwa cattle. 2. Shape of the lactation curve. World Anim. Rev., 66: 49-54.

Moulick, S.K., McDowell, R.E., Van Vleck, L.D. & Guha, H. 1972. Potential of Deshi cattle of India for dairy production. J. Dairy Sci., 55: 1148-1155.

Mrode, R.A. 1988. Lactation performance of the White Fulani cattle in southern Nigeria. Trop. Anim. Health Prod., 20: 149-154.

Ngere, L.O., McDowell, R.E., Bhattacharya, S. & Guha, H. 1973. Factors influencing milk yield of Hariana cattle. J. Anim. Sci., 36: 457-465.

Osman, A.H. & El Amin, F.M. 1971. Some dairy characteristics of northern Sudan Zebu cattle. Inheritance of some reproductive and milk production traits. Trop. Agric. (Trinidad), 48: 201-208.

Patil, V.K. & Prasad, R.B. 1968. A study of economic characters of "Gaolao-breed". I. Lactation yield. Nappur Vet. Coll. Mag., 4: 127-130; Anim. Breed. Abstr., 37: 31.

Patil, V.K. & Prasad, R.B. 1970. A study of economic characters in Gaolao breed. Lactation length. Indian Vet. J., 47: 544-547: Anim. Breed. Abstr., 39: 462.

Rao, M.K. & Sundaresan, D. 1980. Comparison of methods for extending part lactation records for sire evaluation. J. Agric. Sci. (Camb.). 94: 57-62.

Shukla, R.K. & Prasad, R.B. 1970. Genetic and phenotypic studies of lactation yield and lactation length in Gir cattle. Indian Vet. J., 47: 140-145; Anim. Breed. Abstr., 38: 581 -582.

Singh, S.B. & Desai, R.N. 1961. Inheritance of some economic characters in Hariana cattle. Indian J. Dairy Sci., 14: 81-88, 89-94, 141-146, 147-153; 15: 1-8, 9-14; Anim. Breed. Abstr., 31:192.

Wilson, R.T., Ward, P.M., Saeed, A.M. & Light, D. 1987. Milk production characteristics of the Kenana breed of Bos indicus in Sudan. J. Dairy Sci., 70: 2673-2679.