Epsilon Archive for Student Projects

Abstract

In 1996 a hunting test for British gun dogs was started in Sweden, based on a similar one al-ready in place in Norway. The aim of the new test was to describe the hunting ability of the dog and to use the results as a basis for selection. However, up to now, no genetic study of these data has been done. Therefore the aim of this study was to estimate genetic parameters for the traits measured in the hunting test and to investigate whether the traits can summarized into fewer overall measures.
Data from trails of British gundogs between 1996 and 2005 in Sweden were used to estimate genetic parameters and environmental effects in hunting performance. British gundogs consist of five different breeds: Pointer, English setter (ES), Gordon setter (GS), Irish red setter (Irl S) and Irish red and white setter. The study used 7251 competition results were dogs and bitches had equal hunting performance. The pedigree data was collected from the Swedish kennel club. Because the data set was very small for Irish red and white setter, no calculations were made regarding genetic correlations and heritability.
The hunting performance included 13 different traits. Due to the few observations for five of the traits, only eight traits were used in the analysis: speed, style, hunting eagerness, bird find-ing, independence, seeking width, ability to work in the field, and cooperation. Factor analysis of these traits extracted two factors, named energy and an independent ability. Seven different traits, with the exception for the trait bird finding, were evaluated using a mixed animal model using the Restricted Maximum Likelihood method (REML). In the model, adjustment was made for year, season, class and type of performance test.
Earlier studies made by Vangen and Klemetsdal (1988) regarding hunting potential in English setters have shown that some hunting traits have a rather low to moderate heritability: 0.22 for hunting eagerness, 0.18 for field work, and style and speed, and 0.09 for cooperation. They also showed that some of the traits correlated very strongly. Vangen and Klemetsdal (1988) also indicated that hunting traits were heritable and that better systems of assessing such traits might lead to an even higher heritability which leads to a greater potential progress in selec-tion (Serpell, 1995).
The results suggested large difference in breeds regarding heritability. The estimated herita-bility for pointer was low for all traits (0.02-0.09). Gordon setter (GS) was the only breed that showed a high heritability 0.51 (independence). Except for the trait cooperation (0.01) all traits had moderate heritability in the breed GS. However, GS had few data in the study, and for that reason no firm conclusion could be drawn. Irish setter had moderate heritability (0.15-0.27) for four traits (speed, style, hunting eagerness and seeking width), whereas independ-ence, ability to work in the field and cooperation all showed low heritability (0.01-0.08). Al-most the same results were found in English setter: 0.13 (speed), 0.11 (style) and 0.12 (hunt-ing eagerness). The remaining traits showed very low heritabilities.
The repeatability was very variable (0.03-0.51), where GS had the highest values. A high re-peatability can partly be explained by a high additive variance, but also from the fact that the performance test and the judge are relatively reliable.
The genetic correlations were positive and very high for almost all traits. Correlations close to 1.0 were found in ES and GS (hunting eagerness and independence), but also between the traits seeking width and independence (ES). The phenotypic correlations were lower in com-parison with the genetic correlations and showed almost the same results for all breeds: speed, style and hunting eagerness correlated very strongly. Regarding permanent environmental correlations, GS showed high negative values in some traits, possibly owing to too few dogs in the study.
The new traits created based on the factor analysis, named energy and independent ability, showed heritabilities of 0.14 (ES), 031 (GS), 0.22 (Irl S) and 0.06 (pointer), and 0.1 (ES), 0.18 (GS), 0.15 (Irl S) and 0.07 (pointer), respectively. The correlations between the new compo-site traits were all positive, moderate to high.
Because of the low to moderate heritabilities in nearly all traits the evaluation of breeding values should be based on progeny and sib testing instead of performance testing only. The animal model, Best Linear Unbiased Prediction (BLUP) would be a useful method when breeding for hunting traits. Furthermore, more records have to be collected before any further conclusions can be drawn.