National Kennel Clubs are major stakeholders in the governance and regulation of dog breeding. As such, they have been the targets of major criticism related to dog health issues. It is therefore interesting to investigate to what extent health and welfare is a priority for kennel clubs (KCs), and what are the capacities and actions implemented to deal with those issues.
A survey was sent in 2017 to 40 KCs with 15 answers received from 11 European (Austria, Belgium, Denmark, France, Germany, Ireland, Latvia, The Netherlands, Norway, Sweden and the UK) and 4 non-European countries (Australia, Mexico, Uruguay, and the USA) aimed at describing and comparing information across countries in dog breed health management (Wang et al. 2018).
First, in order to determine the population of dogs under the responsibility of KC, the percentages of all dogs being registered as ‘pedigree’ dogs were estimated considering the 15 surveyed KCs, as well 35 other countries, using sources such as the FCI online statistics. Across countries, the average and median percentage of the entire dog population that were registered pedigree dogs, respectively was 20% and 14%. However, there was a large variation across countries, with European Nordic countries showing, in general, a larger proportion of pedigree dogs (see Figure 1). This aspect is of importance, since it is expected that the responsibility toward general dog health, as well as the capacity to improve the situation, relates to the proportion of dogs that are at least to some extent under the influence of the KCs.
When asked about the current challenges, KCs ranked exaggerated morphological features and inherited disorders as the most important issues, showing those two problems are now clearly identified as priorities (Figure 2). By contrast, issues such as economic constraints to breeding were rarely viewed as problematic for dog breeding. Kennel clubs also commented on challenges related to the difficulty to find balance between increased regulation and the risk of losing members; to achieve consensus and compliance of breeders and clubs toward breed health strategies; as well as lack of capacity regarding information provision and education.
Surveyed countries showed great diversity in terms of information management, implementation of breeding strategies, recommendations, requirement, restriction and tools. Most KCs indicated that information on genealogies, breed standards and dog shows were recorded in their data base for most, if not all breeds; however, health information (e.g. screening examinations, genetic tests) was more sparsely recorded and provided to the public, both for breeds within countries and across countries (Figure 3). For instance, KCs from Austria, Australia, Denmark, France, Germany, The Netherlands, Norway, the UK and the USA provided health information status on pedigrees and in online data bases, but in general, not all breeds were covered. When considering implementation of breeding strategies, six countries indicated that there were no breeding strategies implemented by any breed clubs, while in three countries (Austria, Sweden and the Netherlands) it was reported that each breed club had its breeding strategy.
Several countries indicated that they were planning to develop breeding tools and provide health information to users, and for instance, France and Belgium reported having ongoing work to develop tools to provide online pedigree with health information or estimate breeding values for complex disorders such as hip dysplasia.
Although limited by the relatively low number of countries considered, this survey showed that despite large differences in their approach to breeding policies and management, the awareness to improve breeding and health of pedigree dogs was strong among the surveyed Kennel Clubs. The dog breeding world is increasingly global in scope. The understanding of both the diversity of health initiatives and the potential for coordinated actions internationally is key to further efforts to promote dog health and welfare.
There is probably still a lot of progress to be made in term of information provision and collection, as well as planning breeding strategies considering dog health. In particular, finding a consensus in terms of constraints and priorities for breeding, is expected to be particularly challenging for Kennel Clubs and breed clubs in order to implement those strategies. Although the situations differ across countries, exchanges of experiences may surely help to find the most adequate solutions toward improvement of health and welfare.
Wang, S., Laloë, D., Missant, F. M., Malm, S., Lewis, T., Verrier, E., ... & Leroy, G. (2018). Breeding policies and management of pedigree dogs in 15 national kennel clubs. The Veterinary Journal. https://doi.org/10.1016/j.tvjl.2018.02.018
Published a few months ago, a study of Marsden et al. (2016) used whole sequence data of 90 canids to investigate the importance of population bottleneck, inbreeding and artificial selection of the health of the dog.
In order to investigate this issue, the authors compared sequences from breed dogs, village dogs, and gray wolves, measuring (i) the proportion of amino acid changing variants, as an indicator of genetic load, and (ii) the number of Mendelian disease genes, considering genes already identified in dogs and human. Several interesting results were found.
Based on their measurement of amino acid changing variants (corresponding in general to mutations with deleterious effects), it has been estimated that dogs carry between 2 and 3% more deleterious alleles than wolves, which is comparable to the genetic load in the non-African human population (around 1-3%). This seems surprising, as dog breeds may appear to have undergone more severe bottlenecks during their demographic history than human populations. Also, the authors found that bottlenecks during domestication and breed creation were more responsible of the current genetic load in dogs than recent inbreeding.
On the other hand, it appears that genomic regions under selection (also called selective sweeps), were enriched in disease-related genes. This enrichment could be either explained by the fact that genes controlling artificially selected traits in dogs could also confer Mendelian diseases, or to a linkage between genomic regions where disease genes and genes under selection are.
On the basis on these different results, the authors concluded that artificial selection incidentally reduced dog fitness. Also, given the fact that repeated population bottlenecks and small effective population size seemed to have more impact on accumulation of weakly deleterious variation than recent inbreeding (i.e., mating between close relatives), the focus should be more on maintaining large population size, than avoiding inbreeding.
Given the number of studies that showing the impact of recent inbreeding of fitness and selected traits in dog (Ólafsdóttir and Kristjánsson, 2008; Urfer, 2009, Leroy et al. 2014) or other domestic species (Leroy 2014), I would nevertheless disagree with this last recommendation. It would be therefore be interesting to investigate why the results of this genomic study differ from more phenotype-oriented ones. There are several potential explanations, one being that the traits investigated in those latter studies are under different genetic mechanisms that the one studied here. In any case, we may expect that further studies will confirm, refute or further clarify these first interesting results.
Leroy, G. 2014. Inbreeding depression in livestock species: review and meta-analysis. Animal Genetics 189, 177-182.
Leroy G., Phocas F., Hedan B., Verrier E., Rognon X. 2015. Inbreeding impact on litter size and survival in selected canine breeds. The Veterinary Journal.
Marsden, C. D., Ortega-Del Vecchyo, D., O’Brien, D. P., Taylor, J. F., Ramirez, O., Vilà, C., ... & Lohmueller, K. E. 2016. Bottlenecks and selective sweeps during domestication have increased deleterious genetic variation in dogs.Proceedings of the National Academy of Sciences, 113(1), 152-157.
Ólafsdóttir, G.Á., Kristjánsson, T., 2008. Correlated pedigree and molecular estimates of inbreeding and their ability to detect inbreeding depression in the Icelandic sheepdog, a recently bottlenecked population of domestic dogs. Conservation Genetics 9, 1639-1641.
Urfer, S.R., 2009. Inbreeding and fertility in Irish Wolfhounds in Sweden: 1976 to 2007. Acta Veterinaria Scandinavica 51, 21.
I remember in the early months of my PhD to have assisted with a seminar by Raymond Coppinger on dog behaviour. At the beginning of his talk, he told the assembly that all comments and contributions would be welcome, providing the participant did not begin with “my dog did…”. Over the last years, I have begun to think that this comment from Coppinger is the perfect illustration of one the biggest misunderstandings between dog breeders/owner and scientists, that I would call the sampling effect.
Coppinger's meaning was the following: you cannot make a generalization for the whole species, based on the sole example of your own dog. I think this applies not only to behaviour, but also to a lot of traits related to health.
With complex traits, to be able to make generalizations, you need to analyse data on a very large number of dogs… I will take the effect of inbreeding on longevity as an example. How many people have I heard saying “I have known many dogs with high inbreeding that have lived a healthy and long life”?
Well, guess what, they are surely right…
The preceding figure plots the longevity and inbreeding from more than 6,000 Britanny Spaniel (data from Leroy et al. 2015). This plot is not really impressive by itself. If you consider the five dogs on the right side, with inbreeding close to 40% (remembering that 25% corresponds to a mating between full-sibs), one dog died at the age of 9 years, and four died between 15 and 17 years, the average longevity being around 11 years.
So, based on this example, people might conclude that after all, if they can see such inbred dogs live, for most of them, such a long life, then maybe inbreeding does not affect longevity.
Well, guess what, they would be completely wrong…
Wrong, because of the sampling effect, wrong because when considering complex phenomenon such as inbreeding and longevity, you cannot generalize from a small number of examples. You need a lot of data, not on a few dozen… You need at least several hundred, and if possible several thousand.
Applying a simple linear regression of inbreeding on longevity to the same data, the impact is clear: for each 10% increase in inbreeding, longevity decreases by about 1 year. More elaborate models, taking into account the specific distribution of data, breeder or animal effect, confirmed this negative impact (Leroy et al. 2015). But the thing is that, as illustrated by the two figures, longevity is affected by a large number of other factors, related to genetics or environment. This is why it requires a lot of dogs to have an estimation of the global effect of inbreeding on the trait. The individuals cited above were inbred and lived relatively long lives. But they must be considered outliers, in this example, and their experience is not enough to impact the findings for the whole population.
By providing this example, my intent is not to criticize the expertise of dog breeders. The message would be more to underline the importance of confronting and sharing one's own experience, to be open to contradictory opinions, and try to gather data and analyse results in all objectivity. In the dog world, we still have a lot of work to do in that extent.
Leroy, G., Phocas, F., Hedan, B., Verrier, E., & Rognon, X. (2015). Inbreeding impact on litter size and survival in selected canine breeds. The Veterinary Journal, 203(1), 74-78.
As in other domestic species, a large number of traits of selective interest are of complex inheritance, meaning the phenotype expressed by a given individual is determined by an undetermined number of genes, and more or less impacted by its environment. The efficiency of a selection programme on those traits is depending on several factors, including the heritability of the trait, i.e. the part of phenotypic variation which can be related to genetic variation. In livestock species, statistical approaches have been developed and are used to provide estimated breeding values (EBVs), i.e. estimates of the genetic level of an animal, regarding a specific trait. For this, phenotypic measures are taken and combined among related animals to differentiate impact of genes from effect of environment.
In dogs, a classical example of complex trait is the extent of hip dysplasia, probably the most common disorder within the species, with incidence above 10% in 45% of breeds according to OFA. A recent literature review of Soo and Worth (2015) investigated the diagnosis and approaches against hip dysplasia across the world. Currently three radiographic schemes are used as clinical diagnosis, namely the FCI one (based on the objective measurement of subluxation using the Norberg angle) in Europe, the BVA (British Veterinary Association) one in UK, and the OFA one in USA, based on subjective criteria. The figure below provides estimates of heritabilities reported in the above study. Whatever the system chosen, heritabilites have been estimated above 0.15 in most breeds, indicating that adequate breeding programmes should improve the status of breeds relative to hip dysplasia.
What about efficiency of breeding programmes and EBVs?
Despite the fact some EBVs have been already developed a long time ago in Germany for the German shepherd dog, most of the selection has been based until now on the basis of radiographic screening, i.e. the phenotype of the dogs. Different retrospective studies have observed only limited progress over the last years, despite more than 40 years of screening. Using the OFA data base, Hou et al. (2013) estimated that between 1989-2003, a limited genetic improvement has been observed, equivalent to 16.4% of the phenotypic standard deviation. For the UK, Lewis et al. (2010) has observed a 13% improvement in hip scores over the 10-year study period. However, pre-screening by owners, effectively removing information from the worst-affected animals and resulting in only better radiographs being shared/ recorded, has been viewed as a strong bias, being one of the main factors incriminated for such a limited progress. [Editor's note: this is less of a problem in registries where hip dypslasia screening is mandatory for all dogs used in breeding, e.g. Sweden and Finland. See link below.)]
This has led specialists to recommend the use of EBVs instead of phenotypic evaluation solely, as the EBV allows a more accurate identification of dogs with more healthy genetic status. Different studies (Lewis et al. 2010, Malm et al. 2013) has estimated that a selection based on EBVs should improve the genetic progress on hip dysplasia by around 20%. Consequently, in the last years, several countries such as Sweden and UK have proceeded to the implementation of EBVs in their breeding scheme. In that extent, the recent blog post of Katarina Mäki on the genetic progress against hip dysplasia since the implementation of EBV in Finland in 2002 is quite instructive.
And other traits?
Hip dysplasia is certainly not the only trait for which EBVs could have some use. For elbow dysplasia, for instance, EBVs are also computed in the Mate Select tool of UK Kennel club. The recent study of Lappalainen et al. (2015) on intervertebral disc calcification in Dachshunds in Finland provides another interesting example of this. With heritability around 0.5, there is clear indication of potential genetic progress on the trait, the improvement over the last years having been limited, in relation to the fact that only 10% of dogs have screening. It appears that the Finnish Kennel Club hopes to develop EBV on the disease for its breeders, which may be expected to increase the genetic progress relative to this disease.
Regarding behavioural traits, use of EBVs could be a way to improve the selection of dogs. However, as underlined in a previous post, the generally low heritability of those traits, rarely higher than 15%, may constitute a serious limitation to genetic progress.
And in the future?
Of course, one important opportunity in the future is related to the use of genomic information in relation to complex traits. In livestock species, a growing number of traits, usually estimated through classical EBVs, are now estimated through genomic selection, largely improving accuracy. This requires however a large reference population and a good knowledge of the genome of the species. Taking the example of hip dysplasia, several candidate genes have been identified with potential impact on hip dysplasia (Soo et al. 2015). In the future, it may be expected that genomic information could be used either to support classical EBV estimation (marker-assisted selection) or even replacing it (genomic selection). This would allow, among other benefits, to provide an estimate of breeding values at an earlier age, differentiating litter mates based on their respective genome. The technic would however need further improvement before any implementation.
Hou, Y., Wang, Y., Xuemei Lu, X., Zhang, X., Zhao, Q., Todhunter, R. J., & Zhang, Z. (2013). Monitoring hip and elbow dysplasia achieved modest genetic improvement of 74 dog breeds over 40 Years in USA. PloS one, 8(10), e76390.
Lappalainen, A. K., Mäki, K., & Laitinen-Vapaavuori, O. (2015). Estimate of heritability and genetic trend of intervertebral disc calcification in Dachshunds in Finland. Acta Veterinaria Scandinavica, 57(1), 1-6.
Lewis, T. W., Blott, S. C., & Woolliams, J. A. (2013). Comparative analyses of genetic trends and prospects for selection against hip and elbow dysplasia in 15 UK dog breeds. BMC genetics, 14(1), 16.
Malm S, Sørensen AC, Fikse WF, Strandberg E. (2013) Efficient selection against categorically scored hip dysplasia in dogs is possible using best linear unbiased prediction and optimum contribution selection: a simulation study. Journal of Animal Breeding and Genetics 130, 154–64.
Soo, M., Sneddon, N. W., Lopez-Villalobos, N., & Worth, A. J. (2015). Genetic evaluation of the total hip score of four populous breeds of dog, as recorded by the New Zealand Veterinary Association Hip Dysplasia Scheme (1991–2011).New Zealand veterinary journal, 63(2), 79-85.
Wilson, B. J., Nicholas, F. W., James, J. W., Wade, C. M., Tammen, I., Raadsma, H. W., ... & Thomson, P. C. (2012). Heritability and phenotypic variation of canine hip dysplasia radiographic traits in a cohort of Australian German shepherd dogs. PloS one, 7(6), e39620.
For most dog breeders, behavior represents a major challenge. It is indeed one of the main fields of interest in terms of breeding. In a survey we made with French breeders a few years ago (Leroy et al. 2007), behavior was ranked as a breeding goal after morphology, but before health and work. Yet, although the influence of genes on behavioral traits, which are illustrated by the large behavioral differences between breeds, cannot be denied, those traits are also largely impacted by education and environment, which makes selection of those traits difficult.
A recent study (Hradecká et al. 2015) tried to make a reappraisal on the heritability of behavioral traits in dogs, i.e. of the proportion of behavior variation that can be attributed to genes. For this 1763 heritability estimations from 48 studies were gathered in a meta-analysis study, traits being classified into five categories: Environment interaction, herding, hunting, play and psychical characteristics. As showed in the following table, whatever the categories, heritabilities were found to be low, ranging between 9 and 15%. In other words, according to this study, depending on trait categories, around 9-15% of within breed observed variability could be explained by gene differences.
Heritability of traits for different categories according to Hradecká et al. (2015) meta-analysis
This low heritability has several consequences relative to breeding and welfare issues.
First, it underlines the importance of education. It is of course not a surprise, but when coming to behavior, breeders should rely more on their capacity at providing an environment optimal for the behavioral development of the puppies than to genetics.
Secondly, improving the behavior of a breed through selection is not going to be an easy task. Two shortcuts may, however, be considered. First, if the improvement of behavior of the breed is an absolute necessity, maybe the genes of interest can be found outside the breed. Heritabilities are, indeed, in general estimated within breeds, so they do not take into account differences across breeds. It may also be wondered if genomics could be of some help. Given the complexity of behavioral traits, it is expected that even the small proportion of variability which is genetic involves a large number of genes, possibly with small and indirect effects. Yet in the last years, several studies have been able to identify gene variants impacting behavior. Some examples:
In Belgian Shepherd breed, Lit et al. (2013) identified a mutation in dopamine transporter gene, single copy carriers showing seizures, loss of responsiveness to environmental stimuli, episodic aggression, and hyper-vigilance.
When considering disorders related to behavior, several genes have been identified to underlie some inherited forms of epilepsy (see Ekestedt et al. 2013).
A recent comparative analysis between pointing and non-pointing dog breeds allowed researchers to identify two gene candidates that could contribute to pointing behavior in hunting dogs (Akkad et al. 2015).
In German Shepherd Dogs, a Chinese study (Yang et al. 2015) has linked the genotypes for some olfactory receptors to the olfactory ability of those dogs.
Although interesting, it may be argued that some of these findings (e.g., for instance, the last one) do not directly relate to behavior. But actually, the behavior is connected to a wide range of morphological and physiological mechanisms, and it may be expected, for instance, that olfactory capacity impacts the environment interaction and hunting capacities of dogs. It is, however, important to underline that results should always be replicated and confirmed, especially if there are conflicting findings. Also, results may only be of interest or applicable to some breeds or populations. Nevertheless, it may be expected that in the next years, some gene test(s), more or less relevant, will be developed for selection on behavior. As for all other developments in diagnostic molecular genetics, the test(s), their validity, potential impact and application within breeding strategies will need to be carefully considered.
Akkad, D. A., Gerding, W. M., Gasser, R. B., & Epplen, J. T. (2015). Homozygosity mapping and sequencing identify two genes that might contribute to pointing behavior in hunting dogs. Canine Genetics and Epidemiology, 2(1), 1-13.
Ekenstedt, K. J., & Oberbauer, A. M. (2013). Inherited epilepsy in dogs. Topics in companion animal medicine, 28(2), 51-58.
Hradecká, L., Bartoš, L., Svobodová, I., & Sales, J. (2015). Heritability of behavioural traits in domestic dogs: A meta-analysis. Applied Animal Behaviour Science, 170, 1-13.
Leroy, G., Verrier, E., Wisner-Bourgeois, C., & Rognon, X. (2007). Breeding goals and breeding practices of French dog breeders: results from a large survey. Revue de Médecine Vétérinaire, 158(10), 496.
Lit, L., Belanger, J. M., Boehm, D., Lybarger, N., Haverbeke, A., Diederich, C., & Oberbauer, A. M. (2013). Characterization of a dopamine transporter polymorphism and behavior in Belgian Malinois. BMC genetics, 14(1), 45.
Selection on exaggerated morphological features is probably one of the most important problem facing purebred dogs, one of the difficulties being to identify precisely how those morphological traits are affecting the health of the dogs. The recent study of Packer et al. (2015) however provides a very interesting example and suggestions on what could be done relative to the brachycephalic issue.
It indeed illustrates nicely how Brachycephalic Obstructive Airway Syndrome (BOAS), a chronic debilitating syndrome seen frequently in brachycephalic dogs, is connected, in a non-linear manner, with the shortening of muzzle length. According to this study BOAS occurred only in dogs whose muzzle length was less than half the cranial length. A majority of dogs which had this craniofacial ratio lower than 0.2 were found to be affected. The study found also that thick neck girth, obesity and neutering increased the risk for BOAS.
With this craniofacial ratio, breed clubs now have a good indicator that could be used to reduce the incidence of this condition. It is clear that some good work could be made through the promotion of dogs with longer muzzle or even, inserting some condition on this ratio into the standard. As shown in the article, for some breeds, such as the Japanese Chin, it would probably be difficult to improve this ratio to needed values relative to BOAS but efforts have to be made any way, using crossbreeding if necessary. If stakeholders from the dog world do not move on this, it is clear that, at some point, some one else will, and it won’t necessarily be for the good of their breeds.
Asher, L., Diesel, G., Summers, J. F., McGreevy, P. D., & Collins, L. M. (2009). Inherited defects in pedigree dogs. Part 1: Disorders related to breed standards. The Veterinary Journal, 182(3), 402-411.
Packer, R. M., Hendricks, A., Tivers, M. S., & Burn, C. C. (2015). Impact of Facial Conformation on Canine Health: Brachycephalic Obstructive Airway Syndrome. PloS one, 10(10), e0137496. http://www.plosone.org/article/metrics/info:doi/10.1371/journal.pone.0137496
Credit picture: I. Horvath
A few weeks ago, Tom Lewis and his colleagues published what is, up to now, the largest pedigree analysis regarding to the number of populations analyzed, with the 215 breeds recognized by the UK Kennel Club. In the same time, individual breed reports have been made available in the KC website, with accessible infographics on the phenomena behind inbreeding such as effective population size or popular sire effect. Just for this, this work should be saluted, and we can hope that other national kennel clubs will follow this example. Actually, even people most critical to the KC breeding policy have recognized that it should be congratulated for this study. Yet it has also been stated that the global message of the paper, indicating that “initial excessive loss of genetic diversity has latterly fallen to sustainable levels” was probably too much optimistic. After all, it can be indeed considered that 25% of breeds with effective population size under 50 is not so sustainable.
Is it so? And actually what can be deduced from the paper and the analysis?
Evolution of inbreeding
From far away, the most striking result is related to this stabilization (and in several case decrease) of inbreeding occurring in most breeds since the years 2000. As a consequence when computed on recent years, effective population size have increased in multiple breeds. Actually, in case when inbreeding was decreasing, it was not even possible to compute effective population size. But does this really mean that inbreeding stopped to increase? Probably not. First because, in theory, inbreeding can only increase. As the number of generations increases, more and more ancestors in common between both parents will be found, adding their contribution to overall inbreeding. Therefore, as underlined by the authors, this change in inbreeding trend is probably largely explained by the fact that importation increased largely in the years 2000, after relaxation of UK quarantine law. This has probably led to the registration of foreign dogs, with limited pedigree knowledge, and therefore supposedly unrelated. In practice, it is certain that on the beginning, those dogs were less related that the average of national dogs. Yet, after a few years of importations, it is also probable that imported dogs are actually bringing more inbreeding that what is hypothesized based on their limited pedigree knowledge. It could therefore have been interesting to correct genetic parameters such as effective population size by considering this pedigree knowledge.
But is this so simple? Actually, the study of Lewis also provides some results indicating that the recent evolution of inbreeding could be also linked to real inbreeding reduction, although temporary. Indeed, inbreeding is the sum of the reduction of within population genetic variability (which can only decrease), added to factors inherent to the population, such as existence of subpopulations, or mating between close relatives. The extent of inbreeding due to population structure can be assessed by comparing population inbreeding with the degree of coancestry between individuals, or between their parents. This is what Lewis and his colleagues have done, when referring to expected inbreeding. In most of their breeds, real observed inbreeding appears larger than expected one, indicating existence of substructure. Yet in the majority of cases, this difference has clearly decreased over the last years, which probably indicates that breeders have improved their breeding practice, eventually toward less mating between close relatives.
Reduction of popular sire phenomenon
One of the interrogations that I have with this paper is related to the fact that they may have been some change in the use of popular sires. Indeed, a more balanced use of reproducers is expected to reduce (but not inverse) the increase of inbreeding. However, as previously stated, it is difficult to identify such tendency on the basis of inbreeding evolution alone, as it is already impacted by importations and change of substructure. As suggested by the authors, the analysis of the progeny distribution overtime may bring some information on this issue. Yet it has to be stated that this distribution is highly dependent on breed demographics: in a population with 100 annual births, it is quite easy to have a sire producing 25% of the puppies; it is hardly possible in a population with 10,000 births. In most of the breeds studied, registration have either largely increased or decreased over years, so the direct interpretation of progeny distribution evolution should only be made in breeds with stable registrations. Eventually a statistical analysis of the results over the 215 breeds could bring interesting results on this evolution.
So, on my point of view, if there is still effort to be made in the management of genetic variability within UK dog breeds, the article point out some clear progress. Increased importation probably brought some variability within national population, even if those improvements should not be overestimated: imported dogs of yesterday are probably related to imported dogs of today. They have been also probable improvements in mating practice, which is also good news. It is rather difficult for me to identify if there has been change in popular sire phenomenon, and further analysis could bring information on this. In term of communication, it is highly appreciable that messages on better management are brought to the breeders.
References: Lewis T.W., Abhayaratne B.M., Blott S.C. (2015) Trends in genetic diversity for all Kennel Club registered pedigree dog breeds.Canine Genetics and Epidemiology20152:13
Credit picture: I. Horvath
Whatever the domestic species, breed is a central concept of artificial selection. This is especially true for dog, where, as stated in earlier posts, purebreeding is often view as a paradigm. However, it can be legitimate to interrogate this concept of breed.
If, in general, most of the current dog breeds can be well-differentiated, based on phenotypic or molecular analysis, this is not always the case, and among the breeds currently recognized, some actually correspond to the same population, differentiated only on the basis of simple morphological features (for instance long-haired and smooth faced Pyrenean sheepdogs). In other case, cultural factors may contribute to the recognition of a given dog sub-population as a breed. Here is a draft of a text I shared three years ago with the FCI scientific commission, proposing a definition for dog species.
Here are the two definitions given by Clutton-Brock (1999) and FAO:
Clutton-Brock (1999): “A group of animals that has been bred by humans to possess uniform characteristics that are heritable and distinguish it from other animals within the same species.”
FAO (2012): “either a sub-specific group of domestic livestock with definable and identifiable external characteristics that enable it to be separated by visual appraisal from other similarly defined groups within the same species or a group for which geographical and/or cultural separation from phenotypically similar groups has led to acceptance of its separate identity.”
There are several notions behind those definitions:
Homogeneity of the population
Distinctness in comparison to other individual/groups within the same species
Transmissibility of these particular characteristics.
The fact that these characteristics can be defined and identified (FAO)
Socio-cultural context (FAO). It is in my opinion important to remind that behind a breed, there is always breeders.
Therefore I have two propositions: either we use Clutton-Brock definition or we use a sentence such as the following one, which is actually very close to both previous definitions.
Definition: “A population of the canine species sharing definable and inheritable phenotypic characteristics, after a breeding process raised by a group of humans over a given period of time, allowing animals to be distinguished from other defined populations within the same species.”
Clutton-Brock J. (1999) A natural history of domesticated mammals. Cambridge Univ. Press 1999, Cambridge. FAO (2012) Report of a consultation on the definition of breed categories. FAO, Roma.
Credit picture: I. Horvath
When considering implementation of breeding strategies in relation to health of welfare, breed associations and Kennel clubs constitute the first actors susceptible to implement adequate actions. There are however multiple problems in the implementation of efficient strategies, including the fact that, even once efficient clinic or genetic diagnosis have been developed for a specific disease, clubs often don’t know how to proceed then. Efficiency and potential side effects may indeed be difficult to assess, while some choices maybe considered as unpopular, and therefore rejected by breeders. Moreover, in some case, national legislation does not permit the implementation of specific action.
However, when thinking about it, there is a large diversity of initiatives that can be implemented by clubs to promote the health of their breeds, with specific advantages and limits. In the following table, I tried to list some examples, considering those actions can be sorted according to the fact they are based on restrictions on the use of reproducers, promotion of healthy dogs, or information provided to owners and breeders.
Carrot and sticks strategies
The first categories of actions are related to constraints or incentive that can be set to limit the use of dogs that can be deleterious to the health of the breed (note those dogs can be healthy by themselves but carriers of deleterious defects), and promote on contrary healthy dogs. Restrictive measures aiming at banning some dogs from reproduction, or some specific mating, are probably the most evident. Classically it may be decided, for a specific disease, that only non-affected or non-carrier dogs may reproduce. Also, the fact that mating between merle dogs is forbidden in numerous countries relates to this kind of measure. Such measures have the advantage of efficiency, as they may unsure that no product will be affected by the targeted condition (or eventually carry it). A major drawback is that they can be viewed as more or less constraintful by breeders. This is why, before their implementation, it is important to communicate clearly the reason why such action is taken. Note that in some countries, as in France for instance (as well as in USA from what I know?), it is, in theory, not allowed to ban a dog from reproduction.
It can be also decided, as a variant, that to obtain a given award, dogs must be non-affected or non-carriers. In France for example, several clubs impose restriction on health status for the title of national champion. Such measures have the advantage to be less binding; they will be probably also less efficient.
Finally, Clubs can also decide to reward dogs with healthy status or breeders putting effort on breed health, by promoting them in their website for instance. Of course, the efficiency of those measure depend largely if the reward as some impact in term of reproduction. On the other hand, it is also important to assess the consequence of a given measure on breed diversity and health. The constraints and incentive should not restrict the number of reproducers to a few sire or dams, otherwise it could led to the opposite effect to the desired one (see this previous note for more information).
Increase awareness and provide information
When there is a given health problem within a dog breed, a classical reflex is sometime to hide the problem and make as everything was ok, while the best thing to do should be to take action before it’s too late. Similarly for health strategy, an interesting option should be provide information to owners and breeders on the diseases and health and dogs to allow them to make the best choices. Of course, this requires to have access to the information, as well as the right to use it.
Despite the potential constraint that may exist for breeding strategy implementation, given the diversity of potential actions that can be taken for health improvement, there should be not excuse for immobility. If Breed clubs want to remain legitimate stakeholders in dog breeding, they have to take their responsibility and act to the best of their ability to insure a future to dog health and welfare.
Credit picture: I. Horvath
In a previous post, we investigated how a breed population can be affected by fashion effects, showing that the evolution of the popularity of a given breed is the combination of various factors. Legislation is clearly one of those factors. In this post, we will illustrate how some French dog breeds have been affected by two specific regulations, namely the France’s dangerous dogs legislation, and, in relation to ear cropping, the European Convention for the Protection of Pet Animals.
After a serie of tragic incidents that was particularly mediatized, a law was passed in 1999 in France in order to restrict the attacks on humans. Similarly to equivalent legislation in other countries, such as UK, this law targets dog breeds or breed types considered as dangerous, with two different categories. The first category includes dogs with phenotypic appearance of American Staffordshire terrier (pitbulls type), Mastiff (boerbulls type), and Tosa, not registered as purebred. Those dogs are banned from public areas, they have to be neutered and cannot be sell or given. The second category includes purebred American Staffordshire terrier, Rottweiler, and Tosa. Those dogs must be kept on leashes and muzzled in public areas. Possession of dogs of the two categories must be declared and is restricted. Those different constrains are binding for owners, and may undoubtedly have an effect on dog popularity.
European Convention for the Protection of Pet Animals (1987) states that surgical operations for the purpose of modifying the appearance of a pet animal or for other non-curative purposes shall be prohibited, in particular for the docking of tails and the cropping of ears. As a consequence, French government published the convention in 2004 for national implementation. Several measures have been then passed to ban ear cropping. Note that tail docking remains however unrestricted in France, considering potential tail injury that may experience hunting dogs moving through dense vegetation. Several breeds have been concerned by the legislation on ear cropping, including Boxer, Doberman, Great Danes, American Staffordshire terrier, Cane Corso, Beauceron, or Dogo Argentino. The potential impact on breed population is related to change in appearance of the breeds, which may prevent owner used to dogs with cropped ear.
We may suppose that popularity of the breeds targeted by those legislations has been affected (especially for American Staffordshire terrier, concerned by both), but is it so simple? In the following graph, we consider the evolution of French registrations over the 1995-2014 period for targeted breeds (results were restricted to breeds with more the 1000 annual registrations over the last years).
On the one hand, for six breeds out of eight, national registrations have experienced sharp decreases, ranging between 10 and 60% over the 10 years following implementation of the decision (in 1999 for Rottweiler and in 2004 for the other breeds). In most cases, registrations stabilized after a few years.
On the other hand, growing popularity of American Staffordshire terrier and Cane Corso breeds did not seem to be much impacted by those legislations. Eventually a stabilization of registrations could be observed after the publication of the European Convention for the Protection of Pet Animals, while considering dangerous dogs legislation, it could be hypothesized that the law had actually a positive effect on the popularity of purebred American Staffordshire terrier. Indeed as non-pure bred American Staffordshire terrier were classified in the first category, it is probable that a large proportion of owners decided to get purebred ones instead, as the constraints related to the second category were much less binding. In the case of Cane Corso, it is interesting to relate the lack of effect of the ban on ear cropping to the fact that the breed was unknown in France until the end of the 90s. On contrary to more traditional breeds, French owners were not used to cropped ears, which may explain that the change was not perceived by most of owners. Effect on the dog esthetic and how it has been perceived by public may also have differed according to breeds own phenotypes.
These examples illustrate quite well how breed demography is affected more or less directly by national and international legislation. In the next years, we may expect that future legislations, in particular those related to health and welfare, will impact as well dog breeds, leading to the eventual ban of some of them, as it has proposed over the last years by Nordic countries for some cattle breeds. This is one reason why it is important for breeders and kennel clubs to work on the health of their breeds, before national and international legislation take drastic decision for them.
Credit picture: I. Horvath
Here is an article I wrote with Zeev Trainin for FCI a few months ago, The original draft can be found here.
Over the last 20 years, genomics has evolved from a promising field to the main area of study for most animal breeders. Dog has been one of the first mammal species whose genome has been completely sequenced, already in 2005. In parallel, an increasing number of genomic tools have been developed to assist dog breeding. Without going into details about their developmental processes, eventually our interest should be focused more on the current and potential use of these tools, not only from the aspect of efficacy, but rather from the economic and ethical point of view.
First, it should be reminded that variation within the genome is not always linked to diversity in the phenotype. Some mutations that occur may not have a specific impact. When those mutations are not located close to a selected gene, they are not under selection and therefore are considered as neutral. The importance of its mentioning relies on the fact that genomic tools can be differentiated into two categories, focusing either on neutral variation, or on functional one. While the aim of analyzing functional variation is to detect or predict phenotypic traits such as coat color or occurrence of diseases, the goals of investigating neutral variation is related more to the assessment of differentiation and diversity among and within populations and individuals.
Neutral Variation Tools and tests related to neutral variation are generally based on molecular markers (locations on a chromosome where some variation can be observed through an automatized procedure), which have been chosen to be as representative as possible of the whole genome of the population studied. Increasing the number of markers used will increase the precision (and the cost) of the analysis, while the required precision depends on the aim of the tool:
Genetic identification and parentage testing is by far the most common genetic test in dog breeding. It requires in general only a limited set of markers to insure differentiation among individuals, and compatibility between two parents and their putative offspring. The international ISAG (International Society for Animal Genetics) panel for dog identification includes around twenty microsatellites markers. Most important is the sharing of common marker sets among the various laboratories across the world, in order to be able to standardize all results.
Considering that genetically related dogs share more alleles than unrelated ones, initiated the development of various tools measuring the relation between breeds and individuals. In the last few years, some laboratories have commercialized genetic tests that are able to identify breeds at the origin from a given mixed-breed dogs. The efficacy of such test is dependent, of course, on the number of markers used (which should be much more important than the number for genetic identification), as well as to the genetic background of the dog. It is relatively easy to unravel the origins of a first generation cross, but things become more complicated when the ancestors of the mixed dog come from rare breeds ten generations back.
Following the same principle, multiple additional uses of neutral variation can be considered. For instance, breed assignation could be used to check whether a dog without pedigree, showing the appearance of a certain breed, really belongs to that breed. This could be of interest for opening registries to lines that are no longer registered, enlarging (therefore) the genetic variability within the breed.
At the population scale, neutral markers have also been used in multiple studies to assess genetic relations across breeds, in relation to their history for instance (vonHoldt et al. 2010). In practice, this could be used, to determine the breed to be used in order to reintroduce some diversity within a given breed (Farell et al. 2015). For international breed recognition, FCI is now requiring a genetic study for assessing the degree of genetic originality of a candidate breed.
Finally, it can be of interest to measure the level of genetic variability shared by a population or an individual, considering, for instance, the links between genetic variability and health. However, the precision of such analysis relies strongly on the choice of markers and individuals, while until now, no specific procedure exist for such a purpose. Moreover, the links between genetic variability and dog health are still unclear, this question requiring further studies.
Functional variation The main interest of clubs, breeders and owners should probably be focused on the tests considering functional variations. Here the precision is in general not related to the number of markers used, as most of these tests consider only a single area of the genome. Yet, the marker used does not always correspond to the mutation of interest, but maybe located in an area close to the mutation. Also, the trait of interest can be of complex inheritance mode (several genes, influence of environment). These different parameters can reduce the efficacy of the various functional genomic tools.
Issues and concerns During the last decades, hundreds of mutations were identified in dogs, concerning morphological, health or other traits. A growing number of commercial tests have been developed following these discoveries, and recently, testing packages have been proposed to breeders and owners, involving more than 100 genetic tests in one panel. This raises multiple questions for dog breeding.
First, given the number of existing tests, it may be difficult for breeders and owners, to identify the specific ones necessary for their own context. As the incidence of disorders may vary across subpopulations, the use of a given test can be recommended for a given national population, while in another country it may be of no interest, as the same mutation is not being under segregation there. There is a need for guidelines in that extent, being concomitantly updated, according to health situation or to availability of new tests.
Also, priorities should be set at the breed scale, as (i) inadequate selection programs against a given disease may have deleterious consequence on the breed health, increasing incidence of other diseases, and (ii) those genetic test may have a non-negligible cost, and imposing too much the use of genetic tests may in some cases be counter-productive, pushing breeders away from the breed. The test(s) to be used, as well as the constraints or incentives to be set, are depending on the population size, as well as the incidence, inheritance mode and health consequence of the disease.
Hence, breeding strategies should be developed according the context specific to each breed (Leroy and Rognon 2012). In order to answer these various concerns, the establishment of guidelines for clubs and breeders is of upmost importance. This is one of the purposes of the DogWellNet platform launched during the second Dog Health Workshop in February 2015. The brain trust which could initiate such guidelines should be comprised of researchers, veterinarians, welfare associations, breeders, clubs and of course FCI, which should have a major role in their coordination and implementation.
Beyond the growing number of genomic tools and their consequence for dog and breeding, one crucial question remains concerning the role of the breeder in future years. Genomics will be, and actually already is, one key component in dog breeding. Yet, dogs cannot be reduced to a pack of genes, which are combined at will. A lot about genetics remains undiscovered, and in most traits breeding environment will continue to have a role in shaping dog health, morphology and behavior. Genetic tools are, and will continue to serve as complementary instrument to the breeders. We still will need breeder’s skills to work on improving and raising our dogs.
References: Farell L.L., Schoenebeck J.J. et al. (2015) The challenges of pedigree dog health: approaches to combating inherited disease. Canine Genetics and Epidemiology 2015, 2:3
Leroy G., Rognon X. (2012) Assessing the impact of breeding strategies on inherited disorders and genetic diversity in dogs. The Veterinary Journal 194(3) 343-348.
vonHoldt B. M., Pollinger J.P. et al. (2010) Genome-wide SNP and haplotype analyses reveal a rich history underlying dog domestication. Nature 464, 898–902.
Credit picture: I. Horvath
Tests of genetic diseases evoke most interest, of course, using functional variation tools for dog breeding. In the simplest case, such test allows the identification of carriers and potential affected individuals, which is, for breeders, more than useful, as it enables the right choice of their reproducers. Hence, it may help clubs to apply adequate breeding strategies. Yet, it has to be stated that the situation is not always that simple. There are examples where, for instance, two forms of a given disease are presented, having two different mutations behind (and only one has a test). Moreover, in multiple cases, in which a more complex inheritance of the disease occurs (incomplete penetrance, several genes involved) the efficacy of the tool may be affected. For instance, in case of incomplete penetrance, only a given proportion of individual carrying the deleterious allele(s) express the disease: the genetic test has then to be considered as a susceptibility test.
In some cases, some diseases (hip dysplasia or patellar luxation for instance) are due to a combination of environmental and genetic factors (Farell et al. 2015). Here also, some specific tools could be used based on the principle of genomic evaluation, combining genomic, phenotypic and environmental information, to provide an index estimating the genetic susceptibility for an individual to develop a given trait.
Of course, all identified functional mutations are not necessary related with inherited diseases. Several tests have been developed in relation to coat colors, and some mutations are related to dog size or skull shape. These tests may be of some interest for the choice of candidate reproducers. As some mutations have been identified to be associated with abnormal behavior, we may expect, in years to come, commercialization of genetic tests focused on behavioral traits.
I take here the opportunity of the recent (and excellent) posts of Katariina Mäki on crossbreeding, to give some personal insight on this burning issue. As underlined in a previous post, in dogs, pure breeding has been viewed as a paradigm, probably more than in any other domestic species. However, since the lack of genetic diversity (and behind it the fact that most registries are closed to crossbreeding) is considered as one of the most important causes of current health problems in dog, the crossbreeding issue is controversial, if not explosive. This can be illustrated by the recent debates around the integration of crossbreds within British Kennel Club programmes, which have been pointed out as the reason of the replacement of its Chairman.
It is important to state that, if rarely applied, crossbreeding is theoretically not banned from official cynophilia, even once a breed is recognized. First, FCI encourages, in theory, crosses between closely related breeds and breed varieties. In France, in Braque Saint-Germain for instance, there are still official outcrossing with English Pointer implemented by the club. In a survey in 2001, Dumortier (2002) indicated that 41% of French breed clubs reported that some outcrosses had occurred in their breeds. In USA, Farrel et al. (2015) described the outcross made in Dalmatian with a Pointer to improve the health status of the breed relative to hyperuricosuria. Yet, in 2011, even 13 generations after the cross, acceptance of the registration of the products as Dalmatian was still not accepted by some breeders.
Among National Kennel Clubs, the Finnish KC is probably the most proactive in relation to the issue, with specific instructions for crossbreeding implementation approved in 2013 (see the different posts of Katariina Mäki on the topic). Those instructions aim to provide some guidelines, in relation to the reason why crossbreeding should be implemented, breeds and individuals to be chosen, number of litters to be produced, and following of the mating plan. One can argue that those instructions just provide general principles, and remain rather vague for practical implementations. Yet, given the absence of practical and theoretical studies assessing how to implement efficient breeding crossbreeding plans, it is already great to have such principles. Also their different programmes may constitute interesting practical cases for the future.
Lack of knowledge is probably one of the reasons for the controversies. For a breeder, purebreeding is comforting, as s/he may, more or less, have a good idea of what the offspring will look like, what will be their behavior, potential health problems and so on… By contrast, crossbreeding can be viewed as a step into the unknown, and is a process that cannot be taken lightly… Also, crossbreeding with a given breed can be viewed as a strong political decision toward the orientation of breeding goals. Impacting a specific trait or reintroducing genetic diversity at a breed scale, through crossbreeding, requires much more than the use of one single sire to produce one or two litters. Therefore, the consequences on the genetic pool, and behind that, on the morphological, behavioral, and health features can be important.
It is therefore not always easy to find a compromise among actors involved, on the breed to be used, as well as the extent of the crossbreeding programme. An article of Lauvie et al. (2008) on Flemish Red rare cattle breed illustrates that even in other species, stakeholders may have divergent position about the opportunity to use crossbreeding. In dogs, one important question remains about who should decide the use of crossbreeding: e.g. national breed clubs or kennels clubs, country of origin, FCI?
More generally, there is an urgent need to have people working on the issue, providing practical recommendations and political guidelines for dog clubs and breeders.
Dumortier E. (2002) Bilan de la diversite ethnique canine et des pratiques de gestion genetique dans les clubs de races canines en France. These Veterinaire. ENVA.
Farrel L.L., Schoenebeck J.J., Wiener P., Clements D.N., Summers K.M. (2015) The challenges of pedigree dog health: approaches to combating inherited disease. Canine Genetics and Epidemiology. doi:10.1186/s40575-015-0014-9
FCI (2012) Breeds and Varieties: Crosses. FCI Circular www.fci.be/medias/SCI-REG-CRO-RAC-VAR-613.pdf Accessed April, 24 2015.
Lauvie A., Danchin-Burge C., Audiot A., Brives H., Casabianca F., Verrier E. (2008) A controversy about crossbreeding in a conservation programme: The case study of the Flemish Red cattle breed. Livestock Science 118: 113–122.
Credit picture: B. della Rovere AgroParisTech
A recent study, published a few weeks ago, investigated measures that can be used to limit genetic erosion within dog breeds, using simulation of the dutch Golden retriever population. It is not the first time that the main author, Jack Winding, interests himself on questions around management of genetic variability in dogs. He wrote among others with Kor Oldenbroek a book on dog breeding (in Dutch), and build also an interesting monitoring tool for kennel clubs interested in the management of genetic variability. This last article aimed at assessing efficiency of the measures that can be taken by breed clubs to limit the increase of inbreeding over time in their breeds.
Before talking about the result of this study, it appears important to remind what are the factors involved in the erosion of within population variability.
First, genetic variability within a given population is basically related to the number of reproducers used at each generation. It may be easily understand that in a population where less than 10 sire or dams are used each generation, the chance of inbreeding will occur much quickly in comparison to another population with 1000 reproducers.
Not independently, the disequilibrium in the use of reproducers may also lead to increased genetic erosion. Even if you count several hundreds of reproducers within a given population, if 50% of the mating is made by 2 or 3 sires, diversity is going also to fall quickly. This means also that sex ratio is important for the management of genetic variability, and it has been showed that genetic variability will be better preserved in a population with 20 sires and 100 dams, than with a population with 10 sires and 10,000 dams.
The choice of the individuals to be used as reproducers is also very important, and you may use as many sires and dams you want, if in the next generations the reproducers are chosen among the offspring of only a couple of them, you may expect a strong genetic erosion in a few generation.
Finally, the generation interval is also a parameter to be considered, as inbreeding increase over time will be all the more quick since reproducers are rapidly replaced by their offspring.
So, just suppose you have a limited number of reproducers: to limit genetic erosion as far as possible within the population, 50% of those reproducers should be males and 50% female, each of them should have one offspring male and one female used as reproducers, and those replacement offspring should be produced as late as possible. Of course, such measures may be very binding for breeders, removing a lot of the interest of breeding. It appears therefore of paramount interest that kennel clubs propose breeding rules allowing an efficient management of genetic variability, without restricting too much the choices of breeders. A classical measure is to limit the number of litters authorized for a given reproducer, which is recommended by FCI breeding rules and applied in countries like Finland, where this option is included in national health PEVISA programmes. In an early publication on similar topic (Leroy and Rognon 2012), we showed the impact that the limitation of the number of offspring per reproducer could have on genetic variability, as illustrated below.
Other measures can be imagined, aiming at increasing the number of reproducers used, limiting the number of offspring allowed to breed per reproducers (focusing on males as they use is generally more intensified than females), or restricting the relatedness of potential reproducers within each other or with the current generation. In their article, Windig and Oldenbroek tested those different strategies on a population simulated, showing that the best strategies were related to the limitation of the use of popular sire, and to the restriction of breeding to parents with low relatedness to the population. By contrast, approaches aiming at limiting the inbreeding of reproducers themselves or relatedness between parents to be mated were found less efficient. Relative to the use of the threshold to limit the number of litter per sire, it was showed in the study that it was more efficient to consider a fixed number of litters per year than per reproductive life, as the replacement of sire may lead to a reduction of generation interval, which may reduce the impact of the measure. Also, choosing less related reproducers to be mated has only short term effect on inbreeding. For improving management of breed diversity, it appear better to choose reproducer as less related as possible to the current population.
These guidelines are of clear interest for dog clubs aiming at implementing breeding strategies for the management of within-breed genetic variability. Of course, some practical questions remain, such as the thresholds to be chosen for a given breed, once a given measure has been decided. In our early publication, we showed that the same threshold recommended by FCI (the number of offspring per dog should not be >5% of the number of puppies registered in the breed population during a 5 year period) could be not applicable in some dog breeds with large population size, as this threshold was already higher than the maximum number of puppies per reproducer observed (for instance the FCI threshold was found to be equal to 1366 puppies in Epagneul breton), and probably too binding in some small breeds. In Braque Saint Germain breed, for instance the FCI threshold was found to be 14 puppies, meaning than dogs should not be able to produce no more than 2-3 litters to respect the recommendation. This illustrates the difficulty to find a general rule and the necessity to develop threshold specific for each breeds. In that aspect, the simulation program proposed by Windig and Oldenbroek can be very useful.
Of course, simulation approaches are not without weakness, as any simulation process is based on simplification hypothesis. For instance, it was not taken into account here that in real genealogies, pedigree knowledge is not the same for any individual, and a policy aiming at choosing reproducers unrelated to the current population, may actually favor the use of dogs with no pedigree. Strategies considering genealogical relatedness (coancestry, kinship…) should therefore take into account this difference, on one way or another. It should be also interesting to see how the use of original reproducers could be supported at the expense of those who are more related to the current population (i.e. from more famous origins), which is clearly against the usual practices in dog breeding... This kind of measure would therefore need a lot of communication with breeders, to explain issues at stake, but may be necessary, especially for breeds with limited population size.
Leroy G., Rognon X. 2012. Assessing the impact of breeding strategies on inherited disorders and genetic diversity in dogs. Vet J. 194(3):343-348.
Windig J.J., Oldenbroek K. 2015. Genetic management of Dutch golden retriever dogs with a simulation tool. J Anim Breed Genet. DOI: 10.1111/jbg.12149 .
Credit picture: I. Horvath, SCC
In a last post, we saw how different breeding parameters are impacted by the size and the weight of the dogs. It has yet to be stated that these parameters may also, more or less directly, impact breed demographics, in relation to fashion trends which will make some dog breeds popular or not.
The figure below illustrates the evolution of population size in France according to the average female weight of the breeds. Large breeds (and in a lesser extent giant ones) have experienced a clear decline in popularity over the last twelve years, as they accounted for 42% of total French registration in 2002, while this proportion has dropped to 33% in 2014 (note the situation appears stabilized over the last four years). By contrast, medium breeds registrations have increased steadily over the period, ranging from 33% in 2002, to 40% in 2014. It is also interesting to remark that the proportion of registrations from small breeds has reached a peak in 2011 (26%), and has gradually declined since.
Of course, such results are only national ones and should not be extrapolated outside France, as local trends may differ across countries. Also, if dog size can by itself be an explanatory factor, as owner may change their taste and consider large dogs are less convenient (in relation to feed expense for instance), other parameters may also influence dog popularity over time. For instance, legislation on ear cropping has impacted dramatically the popularity of some breeds: in France, registrations of Doberman and Beauceron plunged between 2002 and 2009 from 2,338 to 786, and from 3981 to 2656, respectively. Note also that some breeds may constitute perfect counter-examples of global tendencies: the Cane Corso, a breed of large size with ear traditionally cropped, experienced an impressive increase of its registrations from 1,232 to 4,687 dogs over the 2002-2014 period. Another well-known factor is of course the popularization of dog breeds through movie (and other media) featuring them. A few months ago, a study investigated the extent of the phenomenon in USA, confirming the influence of movie on breed popularity (Ghirlanda et al. 2014).
As showed in the figure above, it appears that the influence of movies on dog popularity although clear, has surprisingly declined over the last 70 years. Indeed, in comparison to past movies, (the 101 Dalmatians, Lassie …) recent ones have only minor influence on the registrations of the breeds featured. Also, they can be exceptions to this tendency, as illustrated by one of the authors in his blog (see below). A previous study of the same authors (Ghirlanda et al. 2013) was not able to relate popularity and fluctuations in popularity with breed behavioral features, leading them to conclude that breed popularity is more determined by fashion than function. Actually as illustrated by these different results, breed demographics appear impacted by a wide diversity of factors.
In relation to this issue, one interesting question remains how popularity impacts the health and welfare of dog breeds. It is generally believed that huge popularity is a bad thing for breed welfare, as some producers will focus more on increasing the number of puppies produced, and less on health improvement. One of the aforementioned studies (Ghirlanda et al. 2013) investigated also the relation between popularity of a given breed in USA and UK and its health, considering the number of disorders reported within the breed. In both countries, a positive correlation was found between both parameters. Yet, such result does not necessarily prove that there is a causal relationship between popularity and breed health. Indeed, within a given breed, the number of dogs that are monitored medically is proportional to the size of the population; it is therefore not surprising to identify a larger number of disorders in large populations. Interestingly, authors also found that breeds with more disorders reported tend to decrease in popularity over a ten year period, which has been interpreted as the fact that owners may have started to avoid breeds considered as unhealthy.
Actually, there is, to my knowledge, a lack of data to prove an eventual deleterious impact of popularity on dog health. When considering evolution of genetic variability, it is of course expected that when comparing two breeds of similar size, the first having experienced a recent increase in popularity, it will probably have a narrow genetic base in comparison to the second one. Yet, in any case, for the first breed, situation will be better than if it population size had not increased at all, as in future, the number of reproducers used within the breed is expected to be higher. Therefore, from a genetic variability point of view, increased popularity is a good thing. The risk is probably more related to the fact that breeders could make inadequate choices of reproducers. Eventually, we may hypothesize that increased popularity of a given breed may decide new breeders (either inexperienced or more interested by profit than dog welfare) to breed. Case-studies may provide eventual insight on the eventuality of such risk, providing an idea of cost-benefits behind a sudden increase of population size.
Ghirlanda S., Acerbi A., Herzog H.A., Serpell J.A. (2013) Fashion vs. function in cultural evolution: The case of dog breed popularity. PLoS ONE 8: e74770.
Ghirlanda S., Acerbi A., S., Herzog, H.A. (2014) Dog movie stars and dog breed popularity: a case study in media influence on choice. PLoS ONE 9(9): e106565
Thanks to Alberto Acerbi for information provided from his blog.
Credit picture: I. Horvath
A few days ago, I have been sent this graph published in 2014 by the excellent Information is Beautiful website. In this infographic, popularity (based on AKC 2011 registrations) was plotted against a synthetic index based on various information related to health, longevity, cost and convenience (data can be found there). Of course, the choice on variable and data source can be discussed (as always), yet I was surprised, at first sight, to see how large or giant breeds where generally poorly rated in comparison to other breeds.
When thinking on variables chosen, such results should however not be surprising, as large dogs are indeed expected to eat more, to cost more, and to have shorter life expectancy (we will come to that later). Large dogs have also predispositions to specific disorders such as hip and elbows dysplasia, but actually there are also conditions associated with small body size (Asher et al. 2009). Anyway, it is amazing to see how health and breeding parameters are impacted by dog size. A funny example of this is related to a preliminary study we made with colleagues of dog canine genetics unit of Rennes on dog prolificacy (see below). We tried to identify genes that could impact the litter size, in a genome wide association study covering 37 breeds. We actually identified several genome areas significantly related to difference in litter size. Yet, actually all those genome areas were already known to be involved in dog size, both parameters being largely correlated, in relation to increased size of uterus for instance. Note that independently to dog size, FCI breed groups appeared to have a significant impact on litter size, in agreement with Borge et al. (2011).
Similarly, there is the well-known relationship between breed size and average longevity, although not as clear as for prolificacy, as illustrated below. In a recent study, Kraus et al. (2013) discussed extensively the reasons why breeds with large body size show shorter lifespan. According to those authors it seems there is a strong correlation between size and aging rate (i.e. increased risk of dying when dogs grow older). It may be eventually due to the fact that age-related diseases are delayed in small breeds, or to specific lethal conditions in large dogs.
Anyway, this increased aging rate may impact dog reproductive carriers, and in a survey on breeding practice (Leroy et al. 2007), breeders from the molossoid FCI 2nd group (including breeds of large size with short lifespan as illustrated by the previous figure) reported ending the reproductive career of their dams and sires earlier (6.1 and 7.3 against 6.7 and 8.4 years on average, respectively). On the other hand, as small dogs reach adult weight earlier (Hawthorne et al. 2004), terrier and companion dogs were reported in the same survey to begin their reproductive career earlier (for dams: 2.2 against 2.4 years on average, for sires 1.6 against 2.1 years on average). In relation to an earlier maturity and longer lifespan, smaller dogs are expected to have longer reproductive career, this, and the low prolificacy illustrated earlier will explain why on average in small dogs breeds, male and female reproducers may be expected to produce more litters during their life.
Selection in dog has led to breeds with an incredible variation in body size, which has important consequences for their health and welfare, as well as for the way they are bred. It could be also interesting to investigate, in a next post, how the perception of the public on dog size may evolve over time and impact the breeds.
Asher L., Diesel G., Summers J.F., McGreevy P.D., Collins L.M. (2009) Inherited defects in pedigree dogs. Part 1: disorders related to breed standards. The veterinary Journal 182(3), 402-411.
Borge, K.S., Tonnessen, R., Nodtvedt, A., Indrebø, A., 2011. Litter size at birth in purebred dogs – A retrospective study of 224 breeds. Theriogenology 75, 911–919.
Kraus, C., Parvard, S., Promislow, D.E.L., 2013. The size-life span trade-off decomposed: Why large dogs die young. American Naturalist 181, 492–505.
Hawthorne, A. J., D. Booles, P. A. Nugent, G. Gettinby, and J. Wilkinson. 2004. Body-weight changes during growth in puppies of different breeds. Journal of Nutrition 134:2027S–2030S.
Leroy, G., Verrier, E., Wisner-Bourgeois, C., Rognon, X., 2007. Breeding goals and breeding practices of French dog breeders: Results from a large survey. Revue de Médecine Vétérinaire 158, 496–503.
Leroy G., Hedan B., Andre C. (2013) Genetic trends and genome wide association study for litter size in dog. 7th International Conference on Advances in Canine and Feline Genomics. Boston.
Thanks to Gull as well as to Knowledge is Beautiful, whose latest publications can be found here.
Credit picture: I. Horvath / SCC / AgroParisTech
Contrary to what could imply the title, instead of dealing with dog shows, in this post I will talk about population genetics...
With development of genomics in the last fifteen years, population geneticists have been able to study the genetic structure of domestic species, investigating among other things the genetic relationships among breeds. Phylogenetic trees constitutes classical tool used for representing such relationships graphically. However those tree assume that once lineages have diverged, no cross may ever occur, which is obviously untrue when considering domestic populations. Alternate graphical representations have therefore been developed such as neighbornet network, which consider all possible relationships among breeds. It appears interesting to compare the results that have been found depending on the species considered (knowing however that those results also depend on the breeds included within the analysis).
In horse for instance (Leroy et al. 2009), some analysis have clustered horse breeds mainly according to their use, with a differentiation from heavy horses to race and riding breeds, pony breeds being found intermediate.
By comparison, neighbornet network in cattle are shaped by the differentiation between European cattle and Zebu breeds (Martinez et al. 2012).
In goat species also, it appears the genetic differentiation is based on the genetic origin of breeds (Lenstra et al. 2005).
When considering dog the situation, phylogenetic networks appear shaped like a kind of star (Leroy et al. 2009). This star-like pattern could suggest that in dog, breeds emerged from an undifferentiated gene pool. This is obviously untrue: dog types existed largely before the standardization of breeds in the 19th century, and other studies have been able to underline closer genetic relationships existing among herding dogs, mastiff like dogs or spaniels for instance (VonHoldt et al. 2012). More likely, this particular shape underlines that genetic differentiation at the breed level is more emphasized in dog, in relation to the fact that, more than in any other species, purebreeding is considered as a paradigm. As a consequence, amount of geneflows accross breeds is in general very limited in dog, especially in comparison to other domestic animals.
Indeed, in cattle, improvement of genetic trait has favored introgression from the more productive breeds to improve the performance of other populations, and for instance, the Hostein breed has been used all around the world to improve local dairy breeds. In horses, gene flows are frequent and regular across breeds, as illustrated below (Pirault et al. 2013). Even in cats, a large part of the breeds are actually varieties which are intercrossed.
The limited amount of geneflows, combined with important prolificacy in dog species and relative short generation intervals (around 4 year), have probably led to important founder effects in dog breeds, explaining the genetic structure observed at the specie level. The problem is that, while on the one hand, purebreeding is one of the bases of modern selection, whatever the species, on the other hand, excluding any form of introgression can be damageable, especially for the purpose of management of genetic variability. Indeed, genetic bottlenecks related to this strong founder effect reduces the genetic basis of each breed, increasing the risks linked to inbreeding and dissemination of inherited disorders.
In the cases when some group of breeders decides to undertake a crossbreeding programme, such initiative is often view as controversial by other breeders. Of course, crossbreeding should not be decided by anybody, but on the other hand, it is rather unclear at which level the decision should be taken: by the local kennel club, by the kennel club of the country of origin, by the FCI? Also, until now, there has been a lack of studies investigating when, and how, crossbreeding should be used. This will be the subject of another post.
Lenstra J.A., The Econogene Consortium (2005) Evolutionary and demographic history of sheep and goats suggested by nuclear, mtDNA and Y-chromosomal markers. In: International Workshop on “The role of biotechnology for the characterization of crop, forestry, animal and fishery genetic resources”, Turin. Available at: http://www.fao.org/biotech/docs/lenstra.pdf.
Leroy G., Callède L., Verrier E., Mériaux J.C., Ricard A., Danchin-Burge C., Rognon X. (2009) Genetic Diversity of a large set of horse breeds raised in France assessed by microsatellite polymorphism. Genetics Selection Evolution 41, 5.
Leroy G., Verrier E., Mériaux J.C., Rognon X. (2009) Genetic diversity of dog breeds : (2) relations between breeds, breed assignation and conservation approaches. Animal Genetics 40, 333-343.
Martínez AM, Gama LT, Cañón J, Ginja C, Delgado JV, et al. (2012) Genetic Footprints of Iberian Cattle in America 500 Years after the Arrival of Columbus. PLoS ONE 7(11): e49066. doi:10.1371/journal.pone.0049066.
Pirault P, Danvy S, Verrier E, Leroy G (2013) Genetic Structure and Gene Flows within Horses: A Genealogical Study at the French Population Scale. PLoS ONE 8(4): e61544. doi:10.1371/journal.pone.0061544.
Volholdt B.M., Pollinger J.P. Lohmueller K.E. Han E. et al. (2010) Genome-wide SNP and haplotype analyses reveal a rich history underlying dog domestication. Nature 464(7290):898-902.
Credit picture: I. Horvath / SCC / AgroParisTech
It the previous post, we have seen the parameters to be taken into account for the establishment of a breeding strategy. It does however not say which measure should be taken in what circumstance, which is not surprising given the number of factors involved. A few general guidelines however exists such as the ones developed by Animal Health Trust, the problem being that each situation is rather unique.
The “What If” procedure: In 2012, we developed a method to assess the potential impact of breeding strategies in simple cases, considering the genealogy specific to a given population (Leroy and Rognon 2012). The principle of the procedure is to investigate “what” would have happen “if” a given breeding strategy had been applied a few years earlier, considering that the impacts would be similar if the strategy would be applied nowadays. In that purpose, we simulated the segregation of an allele a (representing the deleterious allele involved in a monogenic recessive disorders) within the genealogies of the breed until a time t, corresponding to the implementation of the breeding strategies. From time t, genealogies were modified, with some parents removed from reproductions and replaced by reproducers used the same year. Different strategies were tested considering, the mutation identified and each reproducers tested, with the following scenarios:
Scenario erA, in which, from the first year of the programme, dogs affected by the disorder (i.e. homozygote aa) were removed from the reproductive pool
Scenario erI, an ‘intermediate’ policy in which, from the first year of the programme, dogs affected by the disease (i.e. homozygote aa) were also removed from the reproductive pool; heterozygote dogs (Aa) were allowed to reproduce, but their carrier offspring (i.e. heterozygote Aa or homozygote aa) were removed from the reproductive pool;
Scenario erC, in which, from the first year of the programme, carriers (i.e. heterozygote Aa or homozygote aa) were removed from the reproductive pool.
Two allele frequencies were tested (20 and 50%) on four dog breeds with contrasting population sizes: Braque Saint-Germain BSG (283 dogs registered over the 2006-2010 period), Berger des Pyrénées BRP (3,630 dogs registered), Coton de Tulear COT (10,784 dogs registered), and Epagneul Breton EPB (27,325 dogs registered).
The impact on genetic variability was assessed through the evolution of average genealogical coancestries (ɸ) either observed on real genealogies, or modified under the different scenarios. The average coancestry (also called kinship) corresponds to the basal inbreeding of the population, the coefficient of inbreeding of a given individual corresponding to the coancestry between its parents.
The figure 1 illustrates quite well the contrast that a given breeding strategy may have according to the incidence of the disease and to the demographic situation of the breed. For instance, in the Epagneul Breton breed, the different breeding strategies did not impact much the genetic variability (with exception of scenario erC with allele frequency around 50%). By contrast, the impacts were more important for breeds with lower population size. Also genetic variability was much more impacted by the different breeding strategy when the disease incidence was high, and for example, implementing erC strategies (i.e. removing of all carriers) with allele frequency around 50% would have doubled the coancestry, i.e. the basal inbreeding, within Braque Saint German breed.
Of course, as illustrated by the figure 2, the time taken to reduce the allele frequency of the deleterious allele differs across strategies (it is however equivalent across breeds), and while the direct removal of carriers (erC) allow a prompt elimination of the disease, limiting the strategy to the removal of homozygous affected individuals (erA) will have only limited effects over time, intermediate policy (erI) allowing the elimination of the disease over a 10 years period.
These examples illustrate clearly the fact that strategies should be defined according to the situation of each population. While in small populations, the removal of carriers from reproduction should be avoided unless the incidence of the disease is very low, such strategy can be implemented in large populations of the size of Epagneul Breton without affecting too much its genetic variability, if the frequency of the deleterious allele is not to high (under 20%). Of course, it is important to remember, this diagnosis tool is based on a number of simplifications. For instance, it does not take into account the selection made by the breeders in reaction to the disease (even if some variants have been tested within the article). Nevertheless, it may provide a diagnosis specific to each breed having enough pedigree information. The tool could probably be adapted to some more complex situations, considering more complex inheritance mode or coexistence of two different diseases.
In the future, other initiatives could help breeders and clubs to design their breeding strategies. Especially, the sharing of experience and case studies on DogWellNet would allow a better assessment of the efficiency of available measures.
Leroy G., Rognon X. (2012) Assessing the impact of breeding strategies on inherited disorders and genetic diversity in dogs. Vet J, 194: 343-348.
Paper available - https://www.vin.com/Proceedings/Downloads/TuftsBG2013/90649_BreedStrategies_A6_Leroy_TuftsBG2013.pdf
Credit picture: I. Horvath
In relation to the concerns regarding pedigree health, as well as to the growing number of genetic tests available, there is an increasing demand from breed clubs about guidelines regarding the management and selection against inherited diseases. Indeed, the commercialization of a given test does not always mean it allows to indentify all carriers of the targeted disease. Also, removing all carriers may create genetic bottlenecks and can, in some cases, be the surest way to favor the emergence of other diseases. It is therefore very important for breed clubs to have a clear view of their situation, before deciding a given strategy. Here are some elements that should be taken into account for clubs confronted to such interrogation.The first step should be to assess the priorities within the breeds, by answering some questions relative to the disease and capacities to handle it.
What is the impact of the disease on dog welfare?
It is first important to assess the consequence of the disease on the animal'health. For instance icthyosis, which cause some mild skin problem in Golden Retriever, should not be considered with the same importance as cerebellar ataxia in American Staffordshire Terrier, which often leads to premature euthanasia on the affected dogs.
What is the incidence of the disease within the breed at a national or international scale?
The importance that should be given to a given disorder is fully dependent on its incidence within the breed. It is one of the reasons why regular health surveys have to be implemented within dog breeds. It is also important to take into account the international context, in order to know if imported dogs may bring a risk, or on contrary constitute an opportunity relative to the fight against the disease.
What is the inheritance mode of the disease?
The inheritance mode of a disease can vary from very simple (one single allele, recessive or dominant, with full penetrance and early expression) to extremely complex (several allele and genes involved, incomplete penetrance, late expression, and variable incidence across subpopulations). The way the breeding strategy should be implemented will of course depend on this inheritance mode.
The existence of genetic test can greatly simplify the task for breeders and clubs, especially in relation to recessive and/or late expression disorders. It is however important to verify if the efficiency of the test has been assessed in scientific literature for the population of interest. Also for more complex diseases (with incomplete penetrance for instance) a genetic test can only indicate an increasing risk for the dog to develop the disease, which can be interesting anyway, especially if the disease has a heavy effect on health.
What are the demographics of the breed?
Depending on the population size of the breed, the capacity to handle a given breeding programme relative to its impact on genetic variability is not the same at all. The selection pressure should be much less important if dealing with a small breed than if considering an international breed with thousands of litters produced every year.
Are there other diseases or problems to be considered within the breed?
Many breeds have to deal with more than one inherited disorder, with consequences that are more or less important for welfare. Of course, breed clubs may want to select also traits that are not related to breed's health, such as behavior, working capacities or morphology. Breeding strategies may represent constraints for breeders; for instance imposing a multiplication of genetic tests for large number of diseases may be counterproductive and push breeders away from the breed. It is of paramount important to assess which level of priority should be given to which selected trait.
What is the timeframe considered for the breeding programme?
Elimination of the disease is not always reachable, especially for complex diseases. Therefore clubs should decide what is a realistic objective (elimination or a given reduction of incidence) and how long should it take to reach it.
Once the situation and objectives are assessed, it is important to consider the measures that could be used to reach those objectives. Basically, these measures can be classified according to the fact they constitute constraints or incentives for breeders. They are also largely dependent of local Kennel club rules and national legislation; for instance, AKC cannot ban a dog from breeding. These contrasts across countries may appear therefore problematic when considering breeding programmes implemented at an international scale.
Removal of individuals (affected or carriers individuals for instance) from breeding will be by far the most simple (and probably the most efficient) measure to implement, especially if adequate genetic tests are available. The question on if the affected or carriers should be removed and at which timeframe, is dependent of the context (examples will be provided in the next post). In absence of a genetic test, it is possible for instance to remove parents of affected individuals, or eventually to implement genetic indexation, as it was made for hip dysplasia in Sweden or UK for instance, and to base the breeding choice on estimated breeding values. The problems with reproducer removal are related to the fact that, as stated earlier, local rules do not always allow its implementation. Also, as a constraint, it can be unpopular among breeders.
Others measures can be taken to encourage breeders to select against the disease, such as through awards, communication and education. As an example, in France dog are classified in a selection grid specific to each breed, including six levels according to the performance and health status of the dog and its offspring. The efficiency of such measures is in practice quite difficult to assess. It also depends a lot on the participation of breeders.
This post shows the different factors that have to be taken into account for the choice of a given strategy. There is, however, a need for more developed guidelines and case studies. In the next post, we will show some simple illustrations…
Credit picture: I. Horvath
Is there a reliable genetic test available?
The biological causes usually advanced for the welfare problem experienced by dog breeds are either related to selection on traits indirectly deleterious to health, or to reduction of genetic variability, through dissemination of inherited disorders (popular sire effect), or inbreeding depression. It is however not simple to measure if one of those causes is preponderant on the others, which could be of interest to prioritize the efforts to be made for improving breeds health. In that purpose, a recent study of Jansson and Laikre (2014) assessed the level of genetic variability of 26 dog populations raised in Sweden, in relation to their health status, based on pedigree information, while in 2013, Mellanby et al. investigated the relationship between molecular variability (using low density panel of markers) and prevalence of inherited conditions in a set of 13 UK breeds.
In the Swedish study, the classification between 11 healthy and 13 unhealthy breeds was based on veterinary cost reported to insurance companies, unhealthy breeds including typically molossoid type breeds, while healthy breeds included mostly Nordic spitz and hunting dogs. Genetic variability was assessed considering various genealogical indicators including measures of average inbreeding and kinship, or effective population size.
Surprisingly, genetic variability levels were found higher in unhealthy breeds, in comparison to healthy ones, with, for instance, effective population size estimated in 2010 around 147 for the first, and 94.4 for the last. Several explanations can be brought to this unexpected result. Among others, authors advanced pedigree file used did not trace back to the real founders of populations studied, and that genealogical indicators might not reflect actual levels of genetic homozygosity. It has also to be stated that the breed sampling could be more adequate: the healthy breeds included 11 local populations from Nordic countries, such as Norrbottenspitz or Schiller Hound, which may be expected to be of limited effective population size, in comparison to international molossoid breeds, such as French Bulldog or Rottweiler.
Nevertheless, this result could indicate that health status of breeds is probably more related to specific morphological features than recent breeding history. This does not mean that no attention should be taken regarding the current management of genetic variability within dog breeds. In their 2009 studies, Asher, Summers, and their colleagues reported for instance a much larger number of non-conformation linked inherited disorders in comparison to those related to breed standards (312 and 84 respectively). It is currently not possible to assess if the best-in-show of today is not the carrier of a disease that will poison the welfare of its breed tomorrow…
In the UK study, Mellanby et al. did not find any correlation between low molecular variability (using low density panel of markers) and prevalence of inherited conditions for the 13 breeds studied. Here again, it is difficult to make conclusion as the breed sampling and the number of markers are limited. Such relationship should be definitively assessed on a larger number of breeds and using dense genotyping analysis. This would allow to consider in the same time, impact of morphological features and genetic variability on breed health, and would confirm that genetic variability should be a concern at a population level as well as the individual one.
Asher L., Diesel G., Summers J.F., McGreevy P.D., Collins L.M. (2009) Inherited defects in pedigree dogs. Part 1: disorders related to breed standards. The veterinary Journal 182(3), 402-411.
Jansson M., Laikre L. (2014) Recent breeding history of dog breeds in Sweden: modest rates of inbreeding, extensive loss of genetic diversity and lack of correlation between inbreeding and health. Journal of Animal Breeding and Genetics 131, 153–162.
Mellanby R.J., Ogden R., Clements D.N., French A.T., Gow A.G., Powell R., et al. Population
structure and genetic heterogeneity in popular dog breeds in the UK. The veterinary Journal 2013;196(1):92–7.
Summers J.F., Diesel G., Asher L., McGreevy P.D., Collins L.M. (2009) Inherited defects in pedigree dogs. 2: Disorders that are not related to breed standards. The veterinary Journal 183, 39-45.
Credit picture: I. Horvath
If you ask people on the streets about the main problem of purebred dogs, you can bet the first word coming from their mouth will be "inbreeding". Inbreeding makes dogs ill, inbreeding makes dogs stupid, inbreeding makes dogs aggressive... This oversimplification eludes a large part of what is behind by inbreeding and what is the part taken by inbreeding in the current issue of dog health.
To begin with, inbreeding may have several meanings. It can indeed refer to either a measure of shared ancestry, global reduction of genetic diversity or a given system of mating. Those meanings are of course not independent from each other. Here we will essentially focus on the definition of inbreeding as a measure of shared ancestry, i.e. as the probability that two alleles on a given gene have been inherited from a common ancestor.
Inbreeding also depends on which ancestors you consider, and how far back in earlier generations you will get to find common ancestors. Indeed any individual, including mongrels, can be considered as inbred: it just depends on the number of generations you are looking at. Of course in practice, for a given dog you only know a limited number of ancestral generations. Also, inbreeding occurring 3 generations away will probably impact the dog's health differently than inbreeding occurring 30 generations away. However, It is very important when considering the inbreeding of a given dog, to consider the number of generations on which it has been computed.
Finally, inbreeding can be considered either at an individual or at a population scale. Considering inbreeding as a measure of shared ancestry, the situation of dog (or cat) breeds are not the same when considering individual (average inbreeding) or population (average coancestry) level. Coancestry corresponds to the level of shared ancestry between two individuals; inbreeding of a given individual being equal to the coancestry of its parents. Averaged over a population, coancestry will correspond to the baseline inbreeding of the breed. If reproducers within the breed are mated randomly, average inbreeding and coancestry are expected to be almost similar, which is for instance not the case of dogs, as illustrated in the following table.
When considering the average genealogical inbreeding coefficients measured on various cattle, horse, sheep, dog or cat breeds raised in France, the inbreeding level appears clearly higher for dogs and cats compared to livestock species, by contrast to average coancestries. This difference is essentially explained by the breeding practice among those two species, and more precisely, by the mating between close-relatives. In fact, in a study on seven different dog breeds, we estimated that around 5% of litters showed a coefficient of inbreeding larger than 12.5% (equivalent to a mating between half sibs). This shows that a large part of inbreeding observed among dog breeds is easily avoidable, by changing some mating practices.
Let´s talk briefly about the effects of inbreeding on dog health. In general, inbreeding consequences are divided into two categories, according to whether the alleles involved have a substantial impact on breed health (i.e. with reference to inherited disorders) or are mildly deleterious (i.e. with reference to inbreeding depression). Inbreeding may be related to the expression of inherited disorders, and more specifically to those having a recessive inheritance mode (the most common one in dogs), as their expression requires that the dog has inherited the allele expressing the disease twice. In that extent, either at population or individual level, higher inbreeding coefficient multiplies the risk of expressing inherited disorders of any kind, especially if that inbreeding is related to a limited number of ancestors (popular sire effect). Note that independently from any form of inbreeding, selection on morphological features deleterious to breed health has a large part of responsibility in the current prevalence of inherited disorders among dog breeds.
Inbreeding depression refers to the reduction of the mean phenotypic value for traits that can be quantified in one way or another. In livestock species, evidence of inbreeding depression has been showed either for reproduction, growth, conformation or production traits, with an average reduction of 0.14% of the mean of the trait per 1% of inbreeding. This means that for average inbreeding measured by breeds (around 2% according to the previous table) consequences of inbreeding depression are clearly limited. Also the different studies on livestock species have not demonstrated a clear impact of inbreeding on behavioral traits until now. In dogs the consequences of inbreeding on traits related to reproduction, the occurrence of some specific diseases (hip dysplasia), or the longevity (see Institute for Canine Biology webpage) have been reported previously. When considering inbreeding depression, litter size and longevity constitute two interesting life history indicators in relation to their link to prenatal and postnatal survival. As illustrated below, we were clearly able to show the consequence of inbreeding of the litter and its dam on the litter size, as well as impact of individual inbreeding on dog longevity.
On the one hand, if considering the average level of coancestry among dog breeds (around 2%), and summing the impact of both dam and litter inbreeding on litter size (-0.5% each per percent of inbreeding), baseline inbreeding impact only moderately prolificacy in dog, with a reduction of litter size that can be approximated to 2%. On the other hand, in the Britanny Spaniel and German Shepherd Dog for instance, there was an average decrease of longevity of more than 1 year between dogs with inbreeding coefficients <6.25% and those with inbreeding coefficients >12.5%. This illustrates that given the current level of baseline inbreeding (a few percent) reported in most dog breeds, breeders can easily manage inbreeding depression by avoiding mating between close relatives.
So, does inbreeding make dogs ill? Undoubtedly yes, even if it is a predisposing factor, which means that it´s not because your dog is inbred that his health will be compulsorily bad.
Does inbreeding make dogs stupid or aggressive? Until now, we have no proof of that. We may suppose that, as the inbreeding impacts welfare of dog, it might as well affect their behavior. Yet the effect is probably minor, and if you want a culprit for your dog aggressiveness, first blame its environment, i.e. the way he has been raised, which includes yourself.
Should we try to reduce inbreeding increase as far as possible? Clearly yes. Because inbreeding, in its different forms, is one of the key component of the health problems currently occurring in dog breeds. It is not possible to avoid any level of inbreeding, particularly in purebreeding systems. Yet, its extent can be largely minimized through better management practice aiming for instance at reducing close-breeding practice or limiting popular sire effect. There is clear progress to be made in that.
Leroy G., Mary-Huard T., Verrier E., Danvy S., Charvolin E., Danchin-Burge C. (2013) Methods to estimate effective population size using pedigree data: examples in dog, sheep, cattle and horse. Genetics Selection Evolution 45, 1.
Leroy G., Vernet E., Pautet M.P., Rognon X. (2013) An insight into population structure and gene flow within pure-bred cats. Journal of Animal Breeding and Genetics 131(1) 53-60.
Leroy G. (2014) Inbreeding depression in livestock species: review and meta-analysis. Animal Genetics 45(5) 618-628.
Leroy G., Phocas F., Hedan B., Verrier E., Rognon X. (2014) Inbreeding impact on litter size and survival in selected canine breeds. The Veterinary Journal.
Credit picture: I. Horvath, C. Hermeline/LOOF, SCC, AgroParisTech
So who is responsible?
Of course, breeders will probably be the first to be blamed, because they are those who are at the origin of the selection and evolution of dog breeds… also because, in comparison to other domestic species, dog breeders have almost no constraints to make their own selection. Legal background for breeding and reproduction is indeed more permissive than in livestock species, while, as breeding remains a hobby for most, breeders have a large liberty in choosing their reproducers.
Yet the blame is easy…
If more and more dogs have been produced with exaggerated morphological features, it is because there has been a demand on it… Public (I am including me in it) likes those spectacular animals, those looney tunes Bulldogs and other pocket Chihuahuas. And the reason why is that there is still a lack of public awareness on the issue. Here we join again the question of education and information, and also the responsibilities of media in it. Indeed, if media may have a salutary role in the denunciation of the dog loving world excesses, they can as well play a major role in the public enthusiasm toward spectacular dogs.
Coming back to the dog world, breed clubs and judges also share their part. Breed clubs, because the responsibilities have rarely been taken to proceed to real measure for the improvement of dog health, either for personal interest, political reasons or lack of awareness. Judges, because of their own interpretation of standards, complying more with the requirement of the star system, than to the interest of dogs. In that extent, there is an effort to be made in judge formation.
Finally, scientists and veterinary professionals have also frequently failed in their role to provide adequate information to dog owners and breeders. Outside of the Internet, a veterinarian is logically the main person to which a dog owner will turn to about dog health. Meyers-Wallen underlined in 2003 that increasing information from genetic research would be useful in improving canine health only if practitioners have the knowledge and skills to use it ethically and responsibly. Again there is urgent need is that extent.
To conclude, it appears clear that the responsibility about dog health is shared by the different actors of dog world. Development of education and information around dog breeding are probably among the main tools needed to improve the situation. It is actually one of the main purpose of this website - DogWellNet.com and other initiatives such as the Institute of Canine Biology.
Meyers-Wallen V.N. (2003) Ethics and Genetic Selection in Purebred Dogs Reprod Dom Anim 38, 73–76.
Credit picture: I. Horvath
As a starting point of this blog, I find interesting to get back to the reasons why we have such health problems in companion animals. I have got the inspiration to write about it, because of a speech made about hypertype by the former member of Standard Commission of FCI Raymond Triquet in 2013, underlining the shared responsibilities behind this old problem… Enlarging the question to all the inherited conditions deleterious to dog health, this appraisal appears for me as an interesting entry to future discussion topics.
The dog star-system
No need to look far away to find the main cause of current dog health problems. Dogs are basically victims of the star system of dog world, related to the increased competition between breeders/handlers to provide the bests dogs for competition. By itself, there should be no reason to criticize a system rewarding the effort of breeders. Yet there are three consequences of this, impacting selection and reproduction, and having indirect harmful effects on dog health:
First, such a system will tend to overrate a “spectacular” individual, i.e. hypertype dogs showing exaggerated features of the breed type. This is something of a paradox, as dog shows should on the contrary reward the individual most ordinarily close to the standard.
Secondly, intensive competition brings the need for breeders to provide quickly new dogs able to compete, showing morphological traits best ranked at the moment in Shows. This means putting more emphasis on morphology among breeding goals and paying less attention to functional traits, related for instance to health and behavior.
Last but not least, the popular sire phenomenon, is also a direct consequence of this competition, as the dog succeeding in dog shows and working trials is susceptible to transmit its qualities to his offspring… as well as his defects… Each dog (and each human) is carrier of inherited disorders. Knowing some dogs may sire more than 1,000 puppies, it is therefore not surprising that popular sire phenomenon is considered as one of the most important problems for dog health.
Standards by themselves Sadly some disorders reported in dogs are more or less directly related to the standard of the breed, whatever the interpretation that is made from it. An article of Lucy Asher and her colleagues in 2009 listed multiple traits leading to predisposition to welfare problems, such a screw-tails, leading to spina bifida, or shortened faces, which predispose, among others, to multiple eye conditions. Note that even basic traits can be more or less related to health problems. Taking the example of size, large dogs are in general more prone to health problem such as hip dysplasia or dog bloat, having also reduced longevity, while small dogs may suffer from various troubles, such as dental problems and hypoglycemia. Yet it is difficult to deny the responsibility of standards in the current disorders that dog breeds experience. This has led the UK Kennel Club to perform a large review on breed standards. These change, if applied by judge and breeders, may have indeed positive impacts on breed health, however it probably remains a lot to do on this aspect to improve dog health.
Conflicts The dog world is far from being a peaceful world. There are several reasons for that, among others the combination of affective and financial considerations, as well as the concurrence previously raised. This concurrence is however not only about results in shows but also about the fact that purebred dogs for sale are produced by a wide diversity of people, from individuals mating their female once in a life time, to professional breeders, living from the sale of their puppies and considering (or not) selection as part of their activities.
There are positive sides to have a diversity of opinions around breeding, and especially breeding goals, as it offers opportunities to maintain a diversity of families and lines within a given breed. On the other side, opinion divergence may have some consequences when dealing with breeding strategies and plans to be held for health improvement, as in general, the success of a breeding plan is highly related to the involvement of all breeders.
Advanced treatments of dogs The best companion of man is also probably the species benefiting from the better medical monitoring after man himself, which may have side effects when considering breed health. As direct consequence of better healthcare, dog longevity has been improved. However in the same time, incidence of disorders specific to old dogs has dramatically increased. Also, the same empathy that makes us sensible to the current suffering of purebred dogs, led breeders to provide treatments to diseased or weak pups, allowing them to survive and reproduce, and by consequence transmitting the alleles responsible of their initial state. Note that treatment is understandable from an ethical point of view, yet the reflex should be taken by breeders to not bred dogs having experienced potentially inheritable health issues.Not completely independently, the advantages brought by veterinary procedures, from the sanitary, logistic and comfort aspects, has led to a systematization of some medical and surgical treatments in breeds prone to specific problems. A classic example is the generalization of caesarian in small molossoid breeds. From the selective point of view, this has been counterproductive, postponing the health problem and allowing a degradation of the situation, as in the same time, incidence of deleterious condition could be indirectly increased when correlated to “desirable” conformation (Asher et al. 2009).
Outdated consideration on dog breeding… Old habits die hard, and it is true for breeding practice. In current days, no serious breeders give credit anymore to some tales, like the idea that the sire used for a first litter could impact the morphology of puppies produced later by the female. Yet many dog lovers still rely on principles that are part or even complete misconceptions. One fact to consider purebreeding as a paradigm, more than in any domestic species,. The use of crossbreeding, even planned and concerted, is still considered as an heresy by some dog stakeholders, and remains a underused tool, especially in comparison to other companion animal species, such as cats and horses. Inbreeding as a mating practice is a extremely controversial topic for dog breeders. One thing I remember clearly during some interviews in the first months of my thesis is to have heard breeders considering either that “inbreeding was incest” or that “it’s not possible to perform any selection without inbreeding”. We could get back later to the pros and cons behind the use of inbreeding (this will be the topic of a future post). However there are many reasons to believe that this practice is basically detrimental to dog welfare.
Lack of information and formation Last but not least, one of the main reasons of the current situation is probably related to a lack of information and education around dog breeding. Development of the internet has probably increased the accessibility of knowledge to dog lovers, yet it is not always easy to distinguish pertinent information from complete nonsense. For instance, there are currently dozens of genetic tests available on which only a few are of interest for a given breed in a given country. There is therefore strong progress to be made in the field of information and communication around dog health. We could probably continue and find other reasons why dog welfare is currently such an important issue, but those ones are probably the best known, and the responsibilities are shared, as it will be discussed on the next post.
Asher L., Diesel G., Summers J.F., McGreevy P.D., Collins L.M. (2009) Inherited defects in pedigree dogs. Part 1: disorders related to breed standards. The veterinary Journal 182(3), 402-411 .
Triquet R. (2014) La lutte contre les hypertypes, le point de vue d’un vieux cynophile. Ethnozootechnie 95.
Credit picture: I. Horvath
In the last two centuries, since the creation of first kennel clubs, an incredible number of new dog breeds have emerged, thanks to the work of dog lovers. The first companion of man has also been able to provide new services, among others because selection work on behavior allowed some breeds to fulfill specific tasks. Yet, this incredible phenotypic and behavioral diversity generated at the species level did not occur without side effects.
There is an increasing concern about the welfare of dog breeds in relation to their genetics. Today, between 500 and 600 inherited disorders are reported in dog species (Nicholas 2011). The question of banning some breeds particularly prone to health problems appears more and more frequently in the media, and for an increasing number of people, the breeding activity itself is considered as unnatural, unhealthy and immoral.
It appears easy to judge breeding with an anthropomorphic point of view, and people easily forget that without breeding, domestication would probably not have occurred, and the dog would actually not exist. Yet it is not possible to deny anymore the actual damage that breeding has made to dog health. It is high time that the dog community takes action around improvement of dog health.
It is the purpose of DogWellNet, it is the purpose of this blog.
My name is Gregoire Leroy, I am a population geneticist, who has been working for several years on topics related to characterization, conservation and management of domestic breeds, especially with dog species.
In this blog, I will try to provide opinion and information on the existing issues and possibilities to improve dog health, based on existing literature, personal experience or shared data. I also hope to give voice, occasionally, to some experts (searchers, breeders, dog lovers…) in different domains.
I do not intend to give in this blog neither the solution to the current problems that dog breeders experience, as it is probable that no single solution exists, nor to provide the scientific truth to all questions around dog genetics, as there is still a lot to discover and several topics still create debates in the scientific community. I just hope to provide input that may interest and be useful to people interested or working on the improvement of dog welfare.
Credit picture: I. Horvath