CLUSTERING OF DIFFERENT TYPES IN A CHAROLAIS HERD

Measurements were made in a nucleus herd of Charolais cows on pasture in early summer 1998. The average age and body weight of the cows were 6.8 years and 600 kg, respectively. Simultaneously to weighing, the following body parameters were measured with traditional measuring equipment (measuring rod and tape measure): height at withers (132.2±3.93 cm), rump width II (distance between the femoral necks) (52.1±2.74 cm), slanting body length (177.2±8.09 cm) and chest girth (194.5±8.50 cm). The average values in the group of pure-bred cows (n=17) were as follows: height at withers (137.2 cm), rump width II (55.6 cm), slanting body length (185.3 cm), chest girth (203.2 cm). In the group of cross-bred cows with a 50–69% Charolais gene ratio, the body size of the cows was poorer for each characteristic, as compared to the data of pure-bred individuals (height at withers: –5.3 cm, rump width II: –3.5 cm, slanting body length: –6.8 cm, chest girth: –7.3 cm). Two clusters were determined in the group of pure-breds: one was made up of young, moderately wide, small cows with low body weight and short body (1), and the other of young, but markedly wide, tall cows with high body weight and long body (2). The average values of the individuals in cluster 2 were significantly (P<0.05) higher than those of animals in cluster 1, except for age.Cows with a 90–99% Charolais gene ratio were ranked into 4 clusters as follows: cluster 1: old cows with average body weight and size; cluster 2: average aged cows with low body weight, average height and short body; cluster 3: old cows with high body weight and long, wide body; cluster 4: young cows with low body weight and below-average body parameters. Comparing cows in clusters 2 and 4, cluster 2 was found to be superior for each body measurement and this difference was statistically significant (P<0.05). Obviously, the average of the measurements in cluster 4 was significantly smaller than the data of cluster 3. Cows with an 80–89% Charolais gene ratio were also grouped in 4 clusters. These data suggest that applying cluster analysis to cow groups with similar gene ratios is a suitable method for ranking individuals of a certain population and as such for observing the variance within varieties.