CHAPTER 5
REVIEW OF LITERATURE ON DAIRY CATTLE CROSSBREEDING IN THE TROPICS (continued)

5.5. U.S.A.

In the southern states of the U.S. poor performance of dairy cattle has long been a problem, and one major reason has been supposed to be the lack of sufficient heat tolerance in Bos taurus dairy breeds. In 1946 a research project investigating the possibility of introducing heat tolerance through crossbreeding with Bos indicus was initiated. Two males and two females of Red Sindhi were imported from India and formed a nucleus for crossbreeding with Holstein, Brown Swiss and Jersey. The project was carried out jointly by several experiental stations, and the program was to compare animals with various proportions of Bos indicus inheritance. Results as reported by Branton et al. (1966) are presented in Table 5.5.1 (Jersey crosses) and Table 5.5.2 (Holstein and Brown Swiss crosses).

TABLE 5.5.1.
PERFORMANCE OF JERSEY AND JERSEY × RED SINDH1 CROSSBREDS AT THREE EXPERIMENTAL STATIONS IN U.S.A. Source: Branton et.al. (1966)

In age at first calving various Jersey × Red Sindhi crosses were rather similar and similar to the Jersey contemporaries. Milk yield, in contrast, increased with increasing proportion of Jersey inheritance, and this increase could partly be ascribed to the longer lactations in purebred and high grade Jersey compared to the lower grades, line authors found little evidence of heterosis in the data.

TABLE 5.5.2.
PERFORMANCE OF HOLSTEJN, BROWN SWISS AND THEIR CROSSBREDS WITH RED SINDHI AT EXPERIMENTAL STATIONS IN U.S.A. (1) Source: Branton et al. (1966).

(1) The Holstein and Brown Swiss crosses were kept at different stations, and their results are therefore not comparable.

In the Holstein and Brown Swiss crosses only F¹ and backcrosses to the Bos taurus breed were produced. By and large the results were consistent with those obtained for Jersey crosses.

A few years later an experimental program of up-grading Brahman females with Jersey bulls was started at Texas Agricultural Experimental Station (College Station). The same Jersey bulls were used to produce purebred Jerseys and Jersey × Brahman crosses. Results are given in Table 5.5.3.

TABLE 5.5.3.
PFFFORMANCE OF JERSEY × BRAHMAN CROSSBREDS AT TEXAS AGRICULTURAL EXPERIMENTAL STATION, U.S.A. Source: Branton et al. (1966).

The increase in yield with increasing Bos taurus proportion is even more dramatic in this experiment than in that reviewed above, as F¹'s produced only 35 percent and 3/4 J only 50 per cent of the milk yield of pure Jersey. It should be noticed that many of the cows, particularly among F¹ and 3/4 J, failed to yield 4.5 kg milk per day after calving, and were dried off. These cows were recorded with zero yields.

On the basis of the findings in both experiments the authors concluded that “dairy cattle breeders in the Southern United States would profit more from selective breeding within the existing European breeds of cattle for production performance than from any program of introducing Zebu breeding”.

5.6. AUSTRALIA.

Australia has no indigenous cattle, but Bos taurus cattle were introduced by European settlers from late in the 18th century. As dairy animals these cattle did not perform well in the tropical areas of the country, and it was decided to make an attempt to improve their performance by introducing Bos indicus genetic material (Hayman, 1974). Sahiwal and Red Sindhi (ten males and eight females) were imported from Pakistan in the 1950's as a gift to the Australian government. The animals went into a program of crossbreeding with Jersey, with the aim of developing a tropical dairy breed of mixed taurus - indicus origin. This crossbreeding resulted in a population named the Australian Milking Zebu (AMZ). Results obtained in the early stages of the project were reported by Haymann (1974) and are presented in Table 5.6.1.

TABLE 5.6.1.
FIRST LACTATION PERFORMANCE OF JERSEY AND THEIR CROSSES WITH SAHIWAL AND RED SINDHI IN AUSTRALIA. Source: Hayman (1974).

The means in the table include also records of cows which failed to let-down milk when the calf was separated, these comprising about 70% of the crosses. Milk yields were low and lactations short for all crosses, and also for the purebred Jersey. The performance of F₁ was maintained and even improved in F₂ and F₃.

After initial generations of crossing and inter se mating among crosses the population was subjected to intensive selection of females for milk yield and milk let-down. During this stage Red Sindhi inheritance was largely eliminated. Later a progeny testing program was established with the aim of testing six bulls per year. Screening tests for heat tolerance and tick resistance were applied to young bulls before they were selected for breeding.

Records collected at a research station at Badgery's Creek (in the northern part of the country) indicate that Australian Milking Zebu is about equal to Jersey even under good manageent. The new breed has been widely accepted by the milk producers in the region. Breeding animals and semen have also been exported to other tropical countries like Malaysia and Sri Lanka. The proportion-of Bos taurus inheritance in the Australian Milking Zebu has been estimated at about 60%.

A few years after the initiation of the work leading to the development of AMZ another crossbreeding program commenced, based primarily on the Friesian on the Bos taurus side while the Bos indicus contribution was confined to Sahiwal. This program led to the development of a population named Australian Friesian Sahiwal (AFS). The breeding and selection procedure used was similar to that described for AMZ, except that heat tolerance tests were emitted (Rathie, 1981). Major emphasis has been on milk let-down and tick resistance, and both traits have been considerably improved.

Alexander et al. (1984) reported milk yields of the first three generations (Table 5.6.2). The production level in F₁ was maintained in subsequent generations, and the proportion of short lactations, which was very high in F₁, was considerably reduced in F₂ and F₃.

TABLE 5.6.2.
DAIRY PERFORMANCE OF FRIESIAN × SAHIWAI CROSSBRED COWS AT AYR RESEARCH STATION QUEENSLAND, AUSTRALIA. Source: Alexander et.al. (1984).

(1) Predicted from yield in 120 days (only cows milking at least 120 days are included).

Milk yields of AFS cows in cooperators' herds are set out in Table 5.6.3. AFS produced about 15% less milk and milk fat than their Bos taurus herdmates (breed not specified). The authors maintained that this comparison was biased in favour of Bos taurus, which benefitted from more intensive culling and selection, and from a smaller fraction of two year old cows.

TABLE 5.6.3.
MILK PRODUCTION OF AFS COWS AND THEIR BOS TAURUS HEPDMATES IN COOPERATORS' HERDS IN QUEENSLAND. Source: Alexander et al. (1984)

5.7 SUMMARY OF RESULTS REVIEWED

The vast majority of the reports reviewed are on crossbreeding between Bos taurus and Bos indicus cattle. The projects were carried out under widely different conditions, both climatically and otherwise. Experimental design and types of crosses differed, and the way of recording and analysis was inconsistent. Nevertheless it is felt that it is justified to try to summarize the findings in the various projects, in order to draw more general conclusions.

5.7.1. Comparison of various levels of Bos taurus inheritance.

The optimum proportion of Bos taurus genes has been an important issue in discussions on crossbreeding for milk production in the tropics. The reports reviewed are consistent in showing an improvement in almost all traits up to the 50% level, but further grading towards the taurus breed has given variable and often disappointing results.

In order to compare the various levels of taurus inheritance more precisely, records extracted from the reports reviewed were organized in data sets. Each set contained information on crosses between a given Bos taurus and a given Bos indicus breed (if reported separately for each breed) in a particular locality. Records or the parental breeds were also included, whenever available. Altogether forty-six sets of data were formed in this way (see Table 5.7.1.1.). All sets had records on F., while parental breeds and Bos taurus backcrosses were represented in about one half of the sets. The traits studied were age at first calving, milk yield, lactation length and calving interval. Lata on milk yield, lactatiop length and calving interval were usually for first lactation, but in some cases records had been pooled over lactations.

The effect of proportion of Bos taurus was estimated by fitting constants to data sets and genetic groups simultaneously. The various means were weighted by the number of observations included in the mean.

Adjusted means (least squares) for each of the genetic groups considered in the analysis are given in Table 5.7.1.2.

The means indicate a nearly linear improvement in all traits up to 50% Bos taurus inheritance (F¹). From this level onwards a slight increase in calving interval is observed, while no clear trend is seen in the other traits. F² is inferior to F¹ in all traits studied.

Estimates of the average genetic difference between Bos taurus and Bos indicus cattle used in these projects are set out in Table 5.7.1.3. The first line gives simply the difference between the means of the two groups (i.e. genetic groups 0 and 8) from Table 5.7.1.2., while estimates obtained by the multiple regression method are set out in the second line. The two sets of estimates are in close agreement, in part because they are largely based on the same data. The two groups differ by about one year in age at first calving, about one thousand kg in milk yield, and one to two months in length of lactation, while they are similar in calving interval.

The non-linear effect of level of Bos taurus inheritance demonstrated in Table 5.7.1.2 indicates that heterosis is important in these crosses. Table 5.7.1.4 shows the estimates of heterosis, computed first as the difference between F, and the mean of the two parental groups, and then by the multiple regression method. Again the agreement between the estimates obtained by the two methods is extremely good. Heterosis is substantial for all traits except lactation length. In proportion to midparent mean it amounts to about -14% for age at first calving, 28% for milk yield and -6% for calving interval. When milk yield is expressed per day of calving interval heterosis is increased to 37% of midparent mean.

TABLE 5.7.1.1.
SOURCES OF DATA USED TO COMPARE VARIOUS LEVELS OF BOS TAURUS INHERITANCE.

TABLE 5.7.1.2.
PERFORMANCE OF VARIOUS GENETIC GROUPS IN BOS TAURUS × BOS INDICUS CROSSBREEDING. SUMMARY OF RESULTS REVIEWED

(1) Group 0 to 8: Group code indicates proportion of Bos taurus

inheritance in parts of eight. All these groups were sired by purebred Bos taurus or Bos indicus bulls. Group 9: F₂, F₃ etc. (from inter se mating of halfbreds).

TABLE 5.7.1.3.
ADDITIVE GENETIC DIFFERENCE BETWEEN BOS TAURUS (P2) AND BOS INDICUS (P1) CATTLE

TABLE 5.7.1.4.
HETEROSIS IN BOS TAURUS × BOS INDICUS CROSSES

On the assumption of additive and dominance effects only, the performance of F₂ should be expected to be half-way between F₁ and midparent mean. The values observed for F₂ in these data agree well with expectation, except for milk yield, which was much lower than predicted from F₁ and parental performance. This disappointing yield of F₂ has been observed by several authors, and various explanations have been proposed (besides the reduction in heterozygosity):

1. F₁ animals were usually sired by highly selected progeny tested bulls, while the sires of F₂ (i.e. the F₁ bulls) were subjected to very little selection. However, the parents of F₁ were the grandparents of F₂, and the additive genetic value should therefore be similar in F₂ and F₁ even in the absence of selection among F₁ animals.

2. Deterioration in management over time. F₂ would necessarily be born later than F₁, unless steps were taken to produce additional F₁'s to be contemporaries with F₂. In many of the reports this effect of non-contemporaniety was accounted for in the statistical analysis, as year or period was included in the model. Furthermore, deterioration in management could not explain the difference between F₂ and the mean of the two backcrosses, as these three groups would belong to the same generation.

3. Maternal effects. F₁ calves were out of Bos indicus cows and were usually suckled by their dams, while F₂ calves had F₁ dams and were separated from their dams shortly after birth. Again the difference between F₂ and the mean of the backcrosses could not be explained in this way, as all had F₁ dams. It seems also unlikely that maternal effects should be important for traits expressed so late in the animal's life as milk yield.

4. A fourth possibility is that the relatively poor milk yield in F₂ can be ascribed to a partial breakdown of epistatic combinations which have been built up in the parental populations. These combinations will be intact in F₁, but will partly be lost in F₂ and subsequent generations. According to the model proposed by Dickerson (1973) only one half of the two-locus interactions present in the parents are retained in the gametes from which F₂ is produced.

5.7.2. Comparison of various Bos taurus breeds for crossbreeding with Bos indicus.

As is seen from the review several Bos taurus dairy breeds have been used for crossbreeding with Bos indicus. The present comparison was based on findings in projects in which two or more taurus breeds were used simultaneously. The data were extracted from the reports listed in Table 5.7.2.1. Each set of data contained information on crosses with a given Bos indicus breed in a particular locality. Almost all comparable results were for F₁, and the analysis was restricted to this cross. Unfortunately the only Bos taurus breeds represented in these sets were Friesian, Brown Swiss and Jersey. Friesian and Jersey crosses were present in all seventeen sets, while only six sets included Brown Swiss crosses. Traits studied were age at first calving, milk yield, lactation length and calving interval.

TABLE 5.7.2.1.
SOURCES OF DATA USED TO COMPARE VARIOUS BOS TAURUS BREEDS

Constants were fitted to sets of data and Bos taurus breeds simultaneously, thus accounting for any confounding between level of environment and taurus breed used. Adjusted means are presented in Table 5.7.2.2.

Jersey crosses were the youngest and Brown Swiss crosses the oldest at first calving, both differing significantly from Friesian crosses. Friesian crosses had the highest and Jersey crosses the lowest milk yield, and again all differences were signficant. In lactation length the only significant difference was between Friesian, which had the longest lactations, and Jersey, which had the shortest. Jersey had significantly shorter calving intervals than the two other crosses; Friesian and Brown Swiss crosses were not significantly different in this trait.

TABLE 5.7.2.2.
PERFORMANCE OF FRIESIAN, BROWN SWISS AM3 JERSEY Fl CROSSES WITH BOS INDICUS

The ranking was very consistent across date sets. Friesian crosses had the highest milk yield in all sets. Jersey crosses were the youngest at first calving in ten out of the thirteen sets in which this trait was recorded, and had the shortest calving interval in twelve out of fourteen sets.