Kennel club survey (Paper I)
A survey on breeding and health in pedigree dogs was designed and distributed in April 2017 by email to national KCs of 37 FCI countries and three non-FCI countries. These 40 countries were all with more than 2,000 puppies registered by KCs in 2015. Questions comprised eight subjects: 1) importance of issues related to dog breeding and health for the kennel club, 2) existence of government regulations, 3) management of data, 4) information recorded, 5) information provided to breeders and owners, 6) breeding strategies implemented, 7) breeding recommendations, requirements and restrictions, and 8) implementation of specific breeding tools. The survey was completely filled in by 11 European KCs and 4 non-European KCs. Statistical analysis was performed using R software (R Core Team, 2017) including: likert score analysis and heat map analysis.
Merging pedigree databases (Paper II)
Pedigree databases of four breeds (BMA, BMD, ESE and LBR) were provided by French, Swedish and British KCs and were merged across countries through detecting and removing duplicate dog IDs. Afterwards, using a FORTRAN package PEDIG (Boichard 2002), we calculated genealogical parameters (e.g., equivalent complete generations, inbreeding and coancestry coefficients, effective population sizes) and founder contributions according to the birth period, breed and registration country of dogs, to study underlying breeding practices and already existing gene flow between countries by breeds.
Estimation of genetic correlations (Paper III)
Pedigrees databases of GDR and LBR between France, Sweden and the UK were merged using the method of paper II. For French HD scores, the levels of HD (A, B, C, D, and E) were replaced with 1, 2, 3, 4, and 5, respectively. Data transformation of phenotypic records for Swedish and British HD scores were as in Malm et al. (2008) and Lewis et al. (2010), respectively. A trivariate analysis of HD records from the three countries was performed for each breed with mixed linear animal model using DMU software (Madsen and Jensen, 2013) and genetic parameters (e.g., heritability, genetic correlation, selection differential, accuracy of selection) were calculated. The fixed effects in the model differed somewhat across countries, but the random effects of animal (additive genetic effect) and litter were included for all countries. For Sweden, a random clinic-year effect was included as well.
Simulation of breeding strategies (Paper IV)
Specific functions needed in our study were added to the original simulation workflow of the FORTRAN program GenManSim (Windig and Oldenbroek, 2015), including generating breeding values and phenotypes for individual animals and selection on breeding values or phenotypes of HD. Eventually, we conducted a simulation study of national and international breeding programs to describe the potential changes in genetic progress in an example trait, HD, when moving from a national breeding program to an international one.
Increasing concerns of stakeholders toward dog health and welfare (Paper I and II)
Although there were large differences between KCs on how they regarded breeding policies and management, the awareness to improve breeding and health for pedigree dogs was strong among the surveyed countries. Along with having different views on important issues and facing different challenges, KCs
are currently making great efforts to collect and provide information related to breeding and health, as well as providing breeding strategies and guidance in their countries. This issue is even more complex owing to the wide diversity of traits potentially of interest for breeding, which differ according to breeds and national populations, as well as the growing number of genetic variants identified and commercialized gene tests, making it difficult to identify and prioritize breeding goals. Getting a consensus, in terms of constraints and priorities for breeding, is therefore probably one of the most challenging issues for KCs on the road towards implementing effective strategies for health and welfare that would be endorsed by clubs and breeders.
In addition to ‘inherited disorders’, national KCs also ranked ‘exaggerated morphological features’ and ‘inbreeding and genetic variability’ as issues of highest priority. These concerns seem to be in line with recent evolutions of mating practices by dog breeders from France, Sweden and UK, as the results from genealogical analysis in paper II showed a decrease of the proportion of matings between close relatives over the last years in most of the national breed populations studied.
Merging pedigree databases can increase pedigree knowledge and reduce inbreeding in dogs (Paper II)
Merging pedigree databases across countries and quantifying breeding practices and gene flow for pedigree dog breeds is possible and pedigree knowledge was improved after merging pedigrees. The breeding practice of matings between close relatives varied between France, Sweden and the UK. Moreover, a clear tendency of increased gene flow between countries over time was shown for all the breeds and countries in the study. More importantly, the potential to improve genetic variability within pedigree dogs through exchanging breeding animals across countries was confirmed by an increased effective population size when merging the populations of different countries, especially for small-sized populations.
International evaluation across countries and HD measurement scales is possible (Paper III)
The possibility of performing genetic evaluations of HD across countries was confirmed by the favorable genetic correlation estimated in our study between different measurement scales of HD in France, Sweden and the UK. The existence of connections between populations, also an important prerequisite for international evaluations, was illustrated by the number of common sires shared between countries. Enlarging the population size of selection candidates through combining data from different countries can increase the genetic progress of HD, particularly for small populations. The accuracy of selection for the most recent birth-year cohorts of male dogs was, however, only marginally improved by international evaluation compared to national evaluation.
Importing sires with high genetic quality is beneficial to improve HD status (Paper IV)
Our simulation study demonstrated that importing sires with high genetic merit can be an efficient breeding strategy to improve the status of HD in pedigree dogs, although the outcome depends on the accuracies, genetic correlations of HD across countries, selection intensities and use of imported dogs. When considering using of foreign sires with EBV rankings among Top 50%, which is suggested by British kennel club in practice, very strong genetic correlations (> 0.85) are necessary to guarantee the efficiency of importing sires based EBVs. While considering using sires with very high EBV rankings e.g., among Top 10%, moderate accuracies of EBVs (> 0.40) and high level of genetic correlations (> 0.70) are needed.
Table of contents :
1.1 Dog breeding
1.2 Inherited disorders in pedigree dogs
1.3 International breeding exchanges and collaboration
2 Aims of the thesis
3 Summary of studies
3.1 Materials and methods
3.1.1 Kennel club survey (Paper I)
3.1.2 Merging pedigree databases (Paper II)
3.1.3 Estimation of genetic correlations (Paper III)
3.1.4 Simulation on breeding strategies (Paper IV)
3.2 Main findings
3.2.1 Increading concerns of stakeholders toward dog health and welfare (Paper I and II)
3.2.2 Merging pedigree databases can increase peigree knowledge and reduce inbreeding in dogs (Paper II)
3.2.3 International evaluation across countries and HD measurement scales is possible (Paper III)
3.2.4 Importing sires with high genetic quality is beneficial to improve HD status (Paper IV)
4 General discussion
4.1 Breeding objectives
4.2 Data recording
4.3 Genetic evaluation
4.5 Genetic progress
5 Conclusion and perspectives