A COMPARISON OF LIFETIME PRODUCTIVITY IN EWES EVALUATED BY ANIMAL-MODEL WITH THE TRADITIONAL METHOD OF PRODUCTIVITY AND NURSING INDEX CALCULATION

Lifetime productivity of ewes evaluated by animal model (AM), and life time productivity assessed by the traditional method of calculating nu rsing arid productivity indexes have been compared. Lifetime records o f ewes were employed for the calculations. Depending upon the ewes' lo ngevity, ewe records from the years 1983 to 1992 were used. The ewes o f Sumava (S) breed were evaluated, as well as groups of hybrids with t he Romney Marsh (K) breed, Romanov (R) sheep, Finnsheep (F), Improved Wallachian sheep (ZV) and Texel (T) breed. A total of 902 individuals was evaluated, 502 of them had the known productivity and fixed effect of environment was determined in them. These reproductive indicators were analyzed: number of born lambs and nursed lambs, productivity ind ex (IP) and nursing index (IO) (IP = lifetime number of born lams/age- 1, IO = lifetime number of nursed lambs/age-1). Systematic effects of environment and individual breeding value (total breeding value, aggre gate utility value) were considered in the analysis of calculatims. Al l genetic relations were used to estimate individual breeding value. S eparate estimates of each trait were calculated. The data were process ed by animal model (M i s z t a l , 1988). Correlations between breedi ng values, fixed effect of environment and utility values were also ca lculated with respect to productivity and nursing indexes. The levels of the effects of age and utility year are statistically significantly different from each other. As for the comparison of age categories, f ive-year ewes have approximately by 21 % more born lambs than two-year ewes, and by 14 % more lambs than three-year ewes. The highest values for the number of nursed lambs were found in four-year ewes, which ha ve by 17 % more lambs than three-year ewes and approximately by 26 % m ore lambs than two-year ewes. Standard deviations of breeding values f or the investigated traits are lower seven times by order than the rec orded values of productivity. The variability of the estimated fixed e ffect of environment is lower than the variability of breeding values. The aggregate utility value, which is a prediction of the individual' s future performance, has the standard deviation approximately by 50 % higher than the variability of breeding values amounts to. The value s adjusted according to the systematic effects of environment have by almost a third lower standard deviations than the original recorded va lues of the individuals. Standard deviations of the original recorded values of the individual have the twice higher values than the residua l standard deviations adjusted according to all effects. The values of the relations between breeding and utility value are high and statist ically highly significant for the investigated traits. The correlation s between breeding values, fixed environment and utility values with r espect to productivity and nursing indexes show the statistically high ly significant values. In this case the coefficient of determination a mount to 0.57 to 0.84, indicating the reliability of the individual's breeding and utility value prediction if productivity and nursing inde xes are used. Hence it is apparent that prediction based on the progen y's performance using the breeding value determined by means of linear model (BLUP, AM) is much exacter.