LICENTIATAVHANDLING ZOOLOGISKA INSTITUTIONEN Stockholms Universitet S Stockholm

Transcription

1 LICENTIATAVHANDLING ZOOLOGISKA INSTITUTIONEN Stockholms Universitet S Stockholm ISSN ASSESSING INBREEDING AND LOSS OF GENETIC VARIATION IN CANIDS, DOMESTIC DOG (CANIS FAMILIARIS) AND WILD WOLF (CANIS LUPUS), USING PEDIGREE DATA Författare: Mija Jansson Ämne: Populationsgenetik Löpnummer: 2012:3

2

3 Arbetets titel: ASSESSING INBREEDING AND LOSS OF GENETIC VARIATION IN CANIDS, DOMESTIC DOG (Canis familiaris) AND WILD WOLF (Canis lupus), USING PEDIGREE DATA Författarens namn: Mija Jansson Handledare: Linda Laikre, Nils Ryman Forskningsämne: Populationsgenetik Granskningskommitté: Mats Amundin, Bengt Karlsson, Hans Temrin Datum för seminarium:

4 Assessing inbreeding and loss of genetic variation in Canids, domestic dog (Canis familiaris) and wild wolf (Canis lupus), using pedigree data Mija Jansson Division of Population Genetics, Department of Zoology, Stockholm University, SE , Stockholm, Sweden LIST OF MANUSCRIPTS... 5 ABSTRACT... 6 INTRODUCTION... 7 Conserving genetic diversity... 7 Pedigree analysis... 7 The genetic potential of a population... 8 NATIONAL AND INTERNATIONAL CONSERVATION POLICY... 9 OBJECTIVES STUDY SPECIES The domestic dog The wolf MATERIALS AND METHODS Paper I: the mped converter program Paper II: inbreeding and health in dogs Paper III: genetic effects of wolf hunting RESULTS Inbreeding and health in dogs (Paper II) Genetic effects of hunting the wild Swedish wolf population (Paper III) DISCUSSION AND CONCLUSIONS REFLECTIONS ACKNOWLEDGEMENTS REFERENCES APPENDIX 1 CONSERVATION GENETIC CONCEPTS APPENDIX 2 INBREEDING DISTRIBUTION, SCHILLER HOUND APPENDIX 3 INBREEDING DISTRIBUTION, BULLDOG APPENDIX 4 MEAN KINSHIP DISTRIBUTION, SCHILLER HOUND APPENDIX 5 MEAN KINSHIP DISTRIBUTION, BULLDOG

5 LIST OF MANUSCRIPTS Paper I: Paper II: Jansson, M., Ståhl, I., Laikre, L. mped: a computer program for converting a text file into a pedigree file used by the PMx-software for conservation genetic analysis. Submitted. Jansson, M., Laikre, L. Lack of correlation between conservation genetic status and health in dog breeds in Sweden, Manuscript. Paper III: Laikre, L., Jansson, M., Allendorf, F. W., Jakobsson, S., Ryman, N. (2012). Hunting threatens achieving Favourable Conservation Status for an isolated, highly inbred wolf population. Conservation Biology, in press. 5

6 ABSTRACT Breeding strategies based on conservation genetic principles need to be established to retain genetic diversity within species of small population size. Genetic variation is necessary for wild populations to maintain the ability to genetically adapt to environmental changes and for domestic populations to retain potential to change selection traits. For populations where the relationships between individuals are recorded in studbooks, conservation genetic strategies are typically based on statistical pedigree analysis. Traditionally, pedigree analysis for conservation management has focused on zoo populations of threatened wild animals, and available software have been developed in that context. For instance, Population Management x (PMx) is a free software for estimating genetic parameters including inbreeding, kinship, founder allele contribution and survival. PMx is an accessory program to the zoo studbook platform SPARKS and is not easily applied outside this platform. I wanted to apply conservation genetic approaches typically used for captive populations of threatened species to domestic and wild populations. Specifically, I wanted to use PMx for pedigree analysis of domestic dog and wild wolf pedigrees that were not available in the SPARKS format, and we therefore developed a converter program (mped) for making pedigrees of any studbook format fitting the input requirements of PMx (Paper I). I then used mped to modify 26 domestic dog (Canis familiaris) pedigrees (Paper II) and the pedigree of the Swedish wild wolf (Canis lupus) population (Paper III) to permit PMx conservation genetic analyses of these pedigrees. Paper II: Many dog breeds exhibit physical problems that affect individual dogs through life. A potential cause of these problems is inbreeding and loss of genetic variation that is known to reduce the viability of individuals. I investigated the possible correlation between inbreeding and health problems in dogs by comparing the conservation genetic status of dog breeds in Sweden classified as healthy and unhealthy. The classification was based on statistics on the extent of veterinary care obtained from Sweden s four largest insurance companies for pets. We found extensive loss of genetic variation and moderate rates of inbreeding in all the 26 breeds examined, but no strong indication of a difference in these parameters between healthy and unhealthy breeds. Paper III: The wolf (Canis lupus) is classified as Endangered in Sweden. The present population descends from five individuals and is isolated and highly inbred, with individuals being on average more related than full siblings. Hunts have been used to reduce the population size to 210 and keep the population on this level. Using pedigree analysis we showed that continued hunting of the population will make it less likely to reach genetically based Favourable Conservation Status (FCS) criteria that are required by biodiversity legislation within the European Union. In this thesis I will that, when possible, pedigree analysis is a useful tool of evaluating conservation genetic strategies as well as predicting outcomes of such strategies. 6

7 INTRODUCTION Conserving genetic diversity Conservation of genetic diversity is particularly complicated with respect to small populations (Franklin 1980) such as many captive populations of threatened species, domestic breeds, or threatened wild populations. Small populations inevitably suffer from inbreeding and loss of genetic variation which can cause inbreeding depression (negative effects of inbreeding) and loss of evolutionary potential (i.e. the ability to adapt to changes in the environment). Small populations are also largely affected by demographic stochasticity (Allendorf and Luikart 2007). Many old, domestic breeds, which are not used in large scale, commercial production, typically have a small population size, and many are considered threatened (Lannek 2007). Domesticated animals are under strong selection. Selective breeding results in loss of genetic variation (Johansson and Rendel, 1968), and during recent years increasing conservation genetic focus has been devoted to domestic animal populations. This attention includes both scientific efforts and international and national policy work, including the Global Plan of Action for Animal Genetic Resources adopted in Interlaken, Switzerland (FAO, 2007). Breeding strategies based on conservation genetic principles need to be established to retain genetic diversity within breeds of small population sizes (Frankham et al. 2004). Traditionally, population management software has handled zoo animals. But the use has extended and there are many pedigreed populations, not kept in zoos, where a conservation genetic analysis might be interesting. For populations where the relationships between individuals are recorded in studbooks such strategies are typically based on statistical pedigreeanalysis (Ballou et al. 1995). Populations of wild animals kept in zoos and domestic populations typically have studbooks (e.g. Paper II), and construction of pedigrees for wild populations has become more common in later years (Koch et al 2008, Naish and Hard 2008). Guidelines and several software (e.g. exist to aid in the conservation genetic management of populations bred for conservation in zoos, and have only to a limited extent been incorporated in the management of domestic populations. For example, Population Management x (PMx; Ballou et al. 2011; Lacy et al. 2011), a software for conservation genetic analysis of studbook data has been used for aquarium and wild fish populations (Leus et al. 2011, Naish and Hard 2008). Pedigree analysis In statistical pedigree analysis, exact computations of genotypic probabilities are often too complex to compute even for modern day personal computers. Computer simulations thus become a valuable tool (Ballou et al. 1995). For example, the computer simulation application Gene dropping (MacCluer et al. 1986) is frequently used to analyze loss of genetic variation over time in populations bred for conservation purposes in zoos. Gene dropping simulates how alleles 7

8 from individual founders are spread to descendants in the pedigree, and may be used to address a series of questions relating to allele diversity retention and genotypic similarity among various groups of individuals in the pedigree (Lacy 1989; Geyer et al. 1989). Examples of the use of gene dropping include (MacCluer et al. 1986); estimating inbreeding coefficients and the amount existing genetic variation in the population, as well as predicting the risk of future loss of genetic variation. The genetic potential of a population The founders of a population set the genetic potential of a population. The amount of genetic variation in a population cannot exceed that contributed by the founding individuals. The variation in dog traits seems to be driven by artificial selection (Vilà et al. 1999). Today s modern breeds have closed gene pools and 99 % of 414 dogs from 85 breeds are correctly assigned to their breed in a cluster analysis. This results in reduced population size and an overall increase in genetic drift among domestic dogs (Wayne & Ostrander 2007). Because of the closed gene pools, founder effects are responsible for several breed related diseases (Ubbink et al. 1998). It is possible that the intense inbreeding when creating the breeds made the deleterious recessive alleles widely spread already when founding each breed (Vilà et al. 1999). The increase in genetic drift results in loss of genetic diversity within breeds and greater divergence among them. In some breeds genetic variation has been further reduced by catastrophic events such as World War II (Wayne & Ostrander 2007), that had decimated the number of pure bred dogs during that time. There are currently five founders of the present Scandinavian wolf population. Population structure corresponds to a pattern of preferential mating within a subgroup of the population. For many wild species it is natural to identify the subgroups within a geographical region, but this is less natural for purebred dog populations, mainly due to the effective use of popular sires (Calboli et al. 2008). The wild wolves of Scandinavia suffer, like many other threatened wolfpopulations in the world, from geographic isolation and fragmentation (Liberg and Sand 2009). Assessing inbreeding levels Indications of inbreeding depression has been shown in both domestic dogs and the wild wolfpopulation of Scandinavia (Olafsdottir and Kristjansson 2008; Liberg et al. 2005). Many pedigree-dogs have high coefficients of inbreeding (Calboli et al. 2008), and the wild wolves of Scandinavia are, on average, more related than siblings. Inbreeding levels increase over time given a situation with no immigration (and finite population size). Pedigree analysis is useful to estimate loss of genetic variation of due to inbreeding increases (Calboli et al. 2008). 8

9 NATIONAL AND INTERNATIONAL CONSERVATION POLICY Both wild and domestic animals are mentioned in conservation policies. The United Nations Convention on Biological Diversity (CBD; and the National Swedish Environmental Objectives explicitly states that both wild and domestic populations should be conserved. Other policies handle either wild (The Habitats Directive, 92/43/EEC, available at: or domestic (Global Plan of Action for Animal Genetic Resources, available at: populations. The CBD aims at conserving biological diversity, sustainable use of the components of biological diversity and fair and equitable sharing of the benefits arising out of the utilization of genetic resources. CBD mainly refers to wild species diversity, but the importance of conserving genetic variability of domestic populations of animals and plants is becoming increasingly recognized. The CBD, as well as the National Swedish Environmental Objectives ( explicitly states that domesticated animals and the genetic resources they represent are part of the biological diversity that should be conserved, monitored, and sustainably used. Domesticated animals are, unlike wild animals, mentioned as an indicator for assessing biologic diversity trends ( and Target 13 of the new Strategic Plan for explicitly focuses on genetic variation of domestic populations ( The Habitats Directive (92/43/EEC) is the central biodiversity legislation within the European Union, which obliges all member countries to promote a Favourable Conservation Status (FCS) of certain listed habitats and species, including the wolf (except for a Spanish and a Greek population). FCS of a species is defined in Article 1i of the Habitats Directive as: conservation status of a species means the sum of the influences acting on the species concerned that may affect the long-term distribution and abundance of its populations within the territory referred to in Article 2; The conservation status will be taken as favourable when: population dynamics data on the species concerned indicate that it is maintaining itself on a long-term basis as a viable component of its natural habitats, and the natural range of the species is neither being reduced nor is likely to be reduced for the foreseeable future, and there is, and will probably continue to be, a sufficiently large habitat to maintain its populations on a long-term basis The Interlaken Declaration on Animal Genetic Resources for food and agriculture has been signed of 109 countries, including Sweden. The declaration recognizes that there are significant gaps and weaknesses in national and international capacities to inventory, monitor, characterize, sustainably use, develop and conserve domestic animal genetic resources, and needs to be addressed urgently. It also calls for mobilization of substantial financial resources and long-term support for national and international animal genetic resources programs. Wolf conservation is a controversial issue in Sweden, heavily and constantly debated. 9

10 OBJECTIVES The aim of my work is to generate knowledge that will contribute to a sound conservation genetic management of my study species, the domestic dog and the wild wolf. For this thesis I used pedigree data to address questions relating to rates of inbreeding and loss of genetic variation (measured in terms of founder alleles; Lacy 1989). The major objectives were: To develop a converter that can transform a studbook from a text file (.txt), to a input file for the pedigree analysis software PMx (.ped; Paper I) that can be analyzed in the free software Population Management x (Papers II & III). To examine levels and rates of inbreeding and degree of retention of genetic variation in dog and wolf pedigrees (Papers II & III) with the specific objective of addressing the following questions: Is there a difference with respect to inbreeding levels and retention of genetic variability in dog breeds that are healthy versus those that are unhealthy that might imply that recent genetic management affects health status? Were levels of inbreeding, kinship and retention of founder genetic variation in the Scandinavian wild wolf population affected by the hunt performed in 2010? 10

11 STUDY SPECIES The model organisms used in this thesis are the domestic dog (Canis familiaris) and the wolf (Canis lupus; Figure 1). I provide a brief description of these species relating to the present work. Figure 1. Study species of this thesis, left wolf (Canis lupus; photo by Lovisa Häggström) and right domestic dog (Canis familiaris) as exemplified by a Norwegian Elkhound, grey (photo by Sannse and obtained from en.wikipedia.org). The domestic dog Domesticated animal breeds have had profound effects on the evolution of human societies and on the course of human history, although they account for only a small number of species (Ruane 2000). The dog is considered to be the first domestic animal, and its wild ancestor, the wolf, was probably domesticated by mobile hunters-gatherers rather than by settled farmers (Savolainen 2007). The domestication of the dog occurred at least 14,000 years ago (Sundqvist et al. 2006). Up until 200 years ago, dogs were primarily selected for breeding based on practical use for hunters and herders, but dogs have also long been used for other practical purposes such as pulling sledges, guarding property, and as lapdogs to provide warmth (Beilharz 2007). During the last centuries morphology has become the primary focus of selection. Sundqvist et al. (2006) suggest that most modern dog breeds have a recent origin, probably less than 200 years ago. They also show that there was an unequal contribution of sexes in the origin of modern dog breeds with fewer males than females contributing genetically. 11

12 Dog breeds are, according to phylogenetic studies, organized into a distinct evolutionary hierarchy with the following primary groups (Wayne and Ostrander 2007; Vilà and Leonard 2007, Table 1, materials and methods, below): Herding Mastiff (including e.g. some terriers) Modern European (from the 1800) Mountain (including e.g. German shepherd and some spaniels) There are associations of dog haplotypes with wolf lineages which indicate admixture between wolves and dogs which could have been an important source of genetic variation for domestic dogs (Vilà et al. 2003). Some North Scandinavian/Finnish dog breeds were recently proven to have a different origin from the rest of domestic dog. These breeds are the results of wolf-dog crossings a few hundred to a few thousand years ago rather than from one single domestic event (Klütsch et al. 2009). The breeds were Finnish spitz, Norwegian elkhound (grey), Norwegian elkhound (black), and Finnish lapphund. With respect to domestic animal populations, the Swedish Board of Agriculture has identified a number of traditional Swedish breeds of particular conservation concern, including ten dog breeds (Lannek 2007). Of the dog breeds that are Sweden s conservation responsibility we have worked with four breeds, three scenthounds and one hunting spitz: Hamilton Hound, Småland Hound, Schiller Hound, and Norrbottenspitz. The wolf The wolf (Canis lupus L.) is a native species of Sweden, since the last ice age, but it is currently classified as Endangered (Swedish Species Information Centre; and was hunted to extinction during first part of the 20 th century, and in the mid1960s only occasional individuals were observed and the species became protected, but the populations last individuals disappeared. In the winter , a breeding pair established in the border region between the Province of Värmland and Norwegian Hedmark and provided the foundation for the return of the species to the Swedish fauna (Liberg and Sand 2009). These two individuals, together with a male wolf that immigrated from Finland/Russia around 1990, constitute the primary genetic basis for the current population of over 200 wolves inhabiting the central part of the Scandinavian Peninsula. For over 18 years, none of the at least ten wolves that immigrated via Finland to Northern Scandinavia were able to reach the population in Mid-Scandinavia and contribute genes to the population (Liberg and Sand 2009). New genes were added in 2008 when two unrelated immigrating male wolves reproduced with females from the Scandinavian population (Åkesson and Bensch 2010). Since 2008 five more wolves have migrated into Scandinavia. Only two of them had contact with the Scandinavian wolf population including a female that was first observed in the northern part of Sweden in the Province of Norrbotten and later in the Province of Jämtland in 2010, where she caused damage on reindeers and was subsequently moved south three times but she just walked back north again. Now (in 2012) she has find a male and they have settled down near Junsele. Thus, there are currently five, and one potential, founders of the existing Swedish wolf population (Figure 2). This is a low number from a conservation genetics 12

13 perspective because the amount of genetic variation in a population cannot exceed that contributed by the founding individuals. Figure 2. The location of the breeding sites of the five founders of the Swedish wolf population. Nyskoga refers to a mating couple from Nyskoga, all other founders are males. The precarious situation of the Swedish wolves has been pointed out for decades (e.g. Laikre 1999, Liberg et al 2005, Hagenblad et al 2009). Rapid population growth and gene flow from neighboring populations are critical for achieving long-term population viability. Nevertheless, population levels have been kept low, and in 2009 the Swedish Parliament temporarily decided that the population should be kept below 210 individuals (Swedish Government 2008/09:210). In order to achieve this, 28 wolves were culled in 2010 and 19 in In addition, almost 30 animals were shot during these two years for other reasons (e.g., causing damage to livestock), 16 were killed in traffic and two died because of being tagged with radio transmitters. 13

14 MATERIALS AND METHODS Paper I: the mped converter program In Paper I, I presents the mped program which we developed to convert studbook data into a format that can be used in the software Population Management x (PMx; Ballou et al. 2011; Lacy et al. 2011). PMx is a free software for analysis and conservation management of pedigreed zoo populations. It provides tools for optimal genetic management of populations for which the primary goal is to conserve genetic diversity. It can handle quite large data sets and its output includes several parameters relating to inbreeding and loss of genetic variation. PMx is most easily used as an accessory program to the SPARKS software for studbook management, but it can also be used as a stand-alone program for population management if the demographic and genetic data are first prepared in a specific format (.ped, the PMx input file). It is quite tedious to create.ped-files from other than SPARKS s studbooks, with many manual steps. To simplify the creation of new ped-files from databases that were not originally constructed for PMx we developed a converter called mped (make ped-file) in the programming language C. Paper II: inbreeding and health in dogs In total, individuals from 26 pedigrees (breeds) of domestic dog were analyzed (Table 2) in Paper II that focuses on investigating potential correlation between inbreeding and health in dog breeds in Swedish pedigrees record by the Swedish Kennel Club. I selected populations for analysis by first identifying healthy and unhealthy breeds based on information from insurance companies. Statistics reflecting the extent of veterinary care per dog breed were obtained from Sweden s four most important insurance companies for pets (Agria; Folksam; If; and Sveland; The companies Agria, Folksam and Sveland use six categories (six being the highest costs for veterinary care per dog, and one is the lowest). The If insurance company uses eight categories (where eight represents the highest costs for veterinary care per dog, and one the lowest). I ranked dog breeds based on the classifications from the four companies, respectively, and defined unhealthy breeds as those classified as most expensive with respect to veterinary care by at least three of the four companies. The opposite was done to identify healthy breeds (breeds classified in the category of lowest veterinary care expenses by at least three of the four companies). I identified 15 unhealthy and 11 healthy breeds. These breeds are presented in Table 1 together with the classification of each breed with respect to type of dog made by the international kennel club FCI ( Parker et al. (2004), and Wayne and Ostrander (2007), respectively. The number of individuals per pedigree is shown in Table 2. 14

15 Table 1. Unhealthy/healthy breeds were chosen from data from insurance companies. The type of dog based on the nomenclature from the international kennel club (FCI) and the final classification was based on analysis in the publications of Parker et al. (P. et al. 2004) and Wayne and Ostrander (W. and O. 2007). We have used the categories of W. and O., namely: Ancient breeds (A), Herding (H), Mastiff (M), Modern European (E), Mountain (X). In Paper II only a certain type of Bull Terrier - the Miniature Bull Terrier, was studied. Type of breed Classification Class Breed (FCI Breeds nomenclature) P. et al. W. and O. Unhealthy Bull Terrier Bulltype terrier M M Unhealthy Bulldog Molossioid breed M M Unhealthy Bullmastiff Molossioid breed M M Unhealthy Dogo Argentino Molossioid breed - - Unhealthy German Boxer Molossioid breed M M Unhealthy Great Dane Molossioid breed E H Unhealthy Mastiff Molossioid breed M M Unhealthy Neapolitan Mastiff Molossioid breed - - Unhealthy Rottweiler Molossioid breed M X Unhealthy Shar Pei Molossioid breed A A Unhealthy French Bulldog Small Molossian type M M Unhealthy Bernese Mountain Dog Mountain Dog M X Unhealthy Dobermann Pinscher E E Unhealthy Deer Hound Sighthound - - Unhealthy Irish Wolfhound Sighthound H H Healthy Coton de tuléar Companion and Toy dog - - Healthy Hamilton Hound Scenthound - - Healthy Schiller Hound Scenthound - - Healthy Småland Hound Scenthound - - Healthy Finnish Lapphund Spitz type - - Healthy Finnish Spitz Spitz type - - Healthy Norrbottenspitz Spitz type - - Healthy Norwegian Buhund Spitz type - - Healthy Norwegian Elkhound, black Spitz type - - Healthy Norwegian Elkhound, grey Spitz type E E Healthy Siberian Husky Spitz type A A 15

16 Bulltype terriers are terriers that originally were used for fighting (Bull terriers and Staffordshire terriers). Molossers are solidly built, generally quite large dog breeds generally used for guarding people or livestock. Mountain dogs are molossoid breeds but more of a farm dog. Pinschers (and schnauzers) are continental dogs of the type that in Great Britain are called terriers but unlike the terriers the pinschers (and schnauzers) were farm dogs. Sighthounds hunt primarily by speed and sight. Companion and Toy dogs are small in body size and are primarily used as companions. Scenthounds hunt primarily by scent. Table 2. The number of individuals in the pedigree for each of the 26 dog breeds included in Paper II. The type of breed is based on Federation Cynologique Internationale (FCI) Breeds nomenclature. Status Full name No. in ped. Healthy Coton de tuléar 3563 Healthy Hamilton Hound Healthy Schiller Hound Healthy Småland Hound 7118 Healthy Finnish Lapphund 6465 Healthy Finnish Spitz Healthy Norrbottenspitz 8176 Healthy Norwegian Buhund 4123 Healthy Norwegian Elkhound, black 994 Healthy Norwegian Elkhound, grey Healthy Siberian Husky Unhealthy Bull Terrier 4870 Unhealthy Bulldog 5830 Unhealthy Bullmastiff 3749 Unhealthy Dogo Argentino 1295 Unhealthy German Boxer Unhealthy Great Dane Unhealthy Mastiff 1678 Unhealthy Neapolitan Mastiff 1207 Unhealthy Rottweiler Unhealthy Shar Pei 2475 Unhealthy French Bulldog 8933 Unhealthy Bernese Mountain Dog Unhealthy Dobermann Unhealthy Deer Hound 1555 Unhealthy Irish Wolfhound

17 The health problems occurring in the breeds classified as unhealthy have been noted by the Swedish Kennel Club (SKC). Attention to these health issues caused by exterior exaggerations has been noted for instance in the publication Special Breed Specific Instructions (BSI) - regarding exaggerations in pedigree dogs for the individual dog and for the development of the breed as a whole. BSI is handed out to every judge at Swedish dog shows and the aim is to identify areas of risk and to prevent possible future problems. 10 of the 15 unhealthy breeds in this study are noted in the BSI, i.e.: Bull Terrier (and Miniature Bull Terrier), Bulldog, Bullmastiff, German Boxer, Great Dane, Mastiff, Neapolitan Mastiff, Shar Pei, French Bulldog, and Irish Wolfhound. The origins of the dog breeds exterior are to be found in their original tasks even though the interpretations of the breed standards are not always the same today as it has been before. The molossoid and bulltype breeds are typically heavy and solid and when this is exaggerated, the exterior becomes unhealthy. The pincher and sighthounds included in the study are very large in body size for their group of breeds. Typical molossoid breed diseases are shown in Table 3. Even though molossoid breeds show common problems (all have skin problems), it is unknown whether this is due to the same genetic background. 17

18 Table 3. Typical health problems of some of the breeds classified as unhealthy. Data were obtained from Breeds Specific Strategies available from the Swedish Kennel Club ( [in Swedish]). Breed type classification is based on Parker et al (P et al) and Wayne and Ostrander 2007 (W & O); Mastiff (M) and Mountain (X). Breed P et al W & O Health problems Bull Terrier M M immunological diseases (mainly skin diseases), patellaluxation, heart and kidney problems Bulldog M M Bullmastiff M M German Boxer M M Mastiff M M Rottweiler M X HD, ED, epilepsi, thyroid diseases French Bulldog M M Bernese Mountain Dog M X weak contractions, HD 1, a large number of skin diseases, eye problems, problems with circulation and breathing, epilepsy cancer (lymphoma), skin diseases, eye problems, ligament injuries, HD, ED 2, aortic stenosis HD, osteoarthritis, PNP (kidney disease), heart problems, skin problems, epilepsy, Cryptorchidism, spondylosis, weak contractions ear and eye problems, obstetric complications, HD, ED, cruciate ligament injuries, gastric torsion/bloat stomach, skin problems respiratory problems, eye problems, ear infections/problems, allergies (including skin), patella dislocationtumors of the mouth/throat sarcoma lymphon, back pain, epilepsy, othematom, problems in the lower urinary tract HD, ED, PNP (kidney disease), umbilical hernia, tumors, pyometra, cruciate ligament injuries 1 Hip dysplasia 2 Elbow dysplasia 18

19 To address the issue of possible temporal trends in average inbreeding and retention of founder genetic diversity we analyzed levels of inbreeding and loss of founder genetic variation at three points in time including dogs alive at 12/31/1980, 12/31/1995, and 12/31/2010, respectively. The pedigrees of December 31, 2010, represent the full pedigree of each breed (Table 2). The number of individuals per full pedigree varied from 994 (Norwegian elkhound, black; Table 1) to (Norwegian elkhound, grey). Paper III: genetic effects of wolf hunting Paper III is based on a studbook of the Swedish wild wolf population that has been generated by the SKANDULV Research Project ( The Swedish wolf population has been monitored closely since the establishment in the 1980s. Tracking data in combination with molecular genetic analysis of collected blood, hair, and other biological material including tissue from dead animals has resulted in an almost complete pedigree of the population that is maintained by the Skandulv Project (Liberg et al. 2005). We analyzed the genetic effects of the 2010 wolf hunt by using the pedigree data as of November 2010 obtained from the Skandulv Project. We used the Population Management 2000 software ( to compute inbreeding coefficients, mean kinship, and to analyze founder contribution and loss of founder genetic variation. Paper III also relates wolf hunting and the genetic situation of the Swedish wolf population to existing international and national conservation policy, particularly the EU Habitats Directive ( the UN Convention on Biological Diversity ( and the Swedish national environmental goals (Swedish Government Bill 2004/05:150, Environmental Quality Objectives - A Shared Responsibility, adopted by the Swedish Parliament in November 2005). RESULTS The mped converter that we developed (Paper I) proved to be useful to transform studbook data into a pedigree file (a.ped file) that can be used by Population Manager x. We had several problems with some dog pedigrees obtained including too extensive data for PMx (Table 2) which typically has an upper limit of about individuals, but also depending on the complexity of the pedigree. mped was constructed to reduce the file by deleting dead individuals that have no descendants in the living population. Also, data on e.g., date of birth is sometimes missing in the SKC studbooks, and mped can help estimate birthdates in such situations. Similarly, the SKC databases do not include dates of death but mped can provide estimated dates. There are also other ways that mped can modify pedigrees, and in addition to providing input files for PMx, mped can produce input data to the simulation program Vortex Population 19

20 Viability Analyses Software ( mped was successfully used to provide pedigree files used for Papers II and III. Inbreeding and health in dogs (Paper II) We found extensive loss of genetic variation and moderate rates of inbreeding in all the 26 breeds examined, but no strong indication of a difference in these parameters between healthy and unhealthy breeds (Table 4). Thus, we conclude that recent breeding history with respect to inbreeding levels and maintenance of founder alleles does not appear to be a main cause of poor health in some dog breeds. Table 4. Summary of results from Paper II regarding healthy and unhealthy breeds at the three different points of time. Mean values, for all breeds, of founder, founder genome equivalents (FGE), mean kinship (MK) and inbreeding (F). Health status Year Founders FGE MK F Healthy Healthy Healthy Unhealthy Unhealthy Unhealthy Appendix 2-3 shows an example of distribution of F for a healthy and an unhealthy breed and Appendix 4-5 shows MK for the same breeds. MK typically varies less than F when the mean kinship is the theoretical F for the next generation. Figure 3 includes a figure of inbreeding level and one for genetic variation for all breeds during the years grouped in unhealthy and healthy breeds. The inbreeding Figure (3a) illustrates the lack of a linear trend in inbreeding. 20

21 Figure 3. Left (a) - Mean inbreeding for breeds, classified as healthy (light) and unhealthy (dark). The classification of healthy or unhealthy was based on statistics on extent of veterinary care obtained from Sweden s four largest insurance companies for pets. Right (b) - The Loss of Genetic Variation measured as Founder Genome Equivalents (FGE) per Founder for each breed, grouped as unhealthy breeds (dark) and healthy breeds (light). Genetic effects of hunting the wild Swedish wolf population (Paper III) Prior to the 2010 hunt the wild Swedish wolf population consisted of 209 individuals. Fourteen of these wolves were protected from hunting because they represented the territories of the two males that immigrated into Sweden in 2007/2008. This implies that 195 animals were subjected to hunting. The mean inbreeding coefficients among the 195 animals was F=0.29. Among the 28 wolves that were killed, the average inbreeding was F=0.26. This was significantly less than what would be expected if the 28 killed animals had been selected at random. Average F was at 0.27 after the 2010 hunt. If pedigree data had been used to identify the most inbred individuals, average F could have been reduced to However, reducing the average level of inbreeding should not be the only objective of genetic management in a case like this; equally important is to maintain as much as possible of the remaining allelic diversity from the five founders of the population. 21

22 Maximizing the retention of their alleles includes reducing further loss of genetic variation and striving to spread the genes of the two most recent immigrant males so that their genetic contribution becomes similar to that of the three original founders. During the 2010 hunt, offspring from the two immigrant males were protected. Thus, the genetic contribution from these founders was not reduced, but the proportion of lost variation measured as founder allele survival as calculated from the pedigree increased from 18% to 20% for the two original founders and from 4% to 5% for the male that immigrated in DISCUSSION AND CONCLUSIONS I and my colleagues were able to develop the converter program mped (Paper I) that allows for a broad range of pedigrees outside the zoo community to be easily analyzed using already available software specializing in conservation genetic management. This includes domestic populations such as rare breeds subjected to conservation breeding programs in Sweden (Wennerström 2009) and wild populations for which pedigrees are becoming available for an increasing number of populations (Naish and Hard 2008). We then used mped to address conservation genetic issues using domestic dog (Paper II) pedigrees and the pedigree of the wild Swedish wolf population (Paper III), the estimated inbreeding, mean kinship and loss of founder genetic variation in these populations. With respect to our specific questions, we found no difference regarding inbreeding levels or retention of genetic variation in dog breeds which are classified as unhealthy versus those classified as healthy. Thus, inbreeding and loss of variation over the last few decades do not seem to be associated with the health problems in many dog breeds. It is possible that the initial inbreeding and founder effects associated with the creation of separate dog breeds are the major causes of these health issues. The fact that the same type of defects occur in closely related breeds supports this idea. The temporal tendency of reduced inbreeding levels and increasing number of founders indicate that these Swedish dog populations are not closed but that dogs from other pedigrees of the same breeds in other countries are added to the majority of these populations. First, inbreeding and mean kinship coefficients might be underestimated in cases where imported dogs have ancestors in the Swedish population but further back than three generations (which is what is registered). Imported dogs might be regarded as unrelated to the Swedish population while, in fact, they are not. Second, founder statistics can be affected; if an imported dog does not have Swedish ancestors in its pedigree three generations back, up to eight new founders will be added to the Swedish population of the breed. The exact number depends on potential relationships among the dogs of the three generations back pedigree of the imported dog. If one dog occurs in several places in this pedigree the number of added founders will be less than eight. This issue on the data is still unsolved. There might be a way to handle it, but, it will include a lot of manual steps. One can, in the present version of PMx, specify the kinship of any animal 22

23 (including founders) to any other animals, by manually entering a matrix of empirical kinships for those animals. Thereby probable non-founders would not be considered as founders. My future plan is to use an Excel spreadsheet of existing kinships for a living population and then save it as a text file for use by PMx. What I probably will need to do for "founders" imported to Sweden is to assign kinships that make them related to all the animals in Sweden at the time that the breeding lines are back in Sweden. Bob Lacy, developer of PMx (personal contact) thinks that the best way to do this probably is to assign kinships that give them relationships to each of the (true), early, founders of the Swedish population. This will cause the latter imported founders also to be related to all of the descendants of those Swedish founders. With respect to the wolf pedigree, the 2010 hunt resulted in 28 animals being killed that were on average less inbred than expected from a random sample among the 195 wolves subjected to the hunt. On average two percent of the remaining genetic variation from the original Nyskoga founders (G1-83, D85-01; Figure 2) were lost during the hunt and one percent of the reaming variation from the Gillhov male founder (G1-91; Figure 2). We conclude that the official hunts decided by the Swedish government have not contributed to the genetic health of the population. Further, the hunt was not carried out in a manner that agrees with adaptive management because available pedigree information was not used in the best possible way. For example, a common conservation genetic recommendation when genetically managing populations with a known pedigree is to identify and remove individuals with the highest mean kinship (Lacy 1995). Relating the genetic situation of the Swedish wild wolf population and the hunt to international and national conservation policy, we conclude that hunting to reduce wolf numbers in Sweden is currently not in line with national and EU policy agreements and will make genetically based FCS criteria less achievable for this species. We suggest that to reach FCS for the wolf in Sweden the following criteria need to be met: i) a well-connected, large, subdivided wolf population over Scandinavia, Finland, and the Russian Karelia-Kola region must be reestablished, ii) a genetically effective population size (Ne) is in a minimum range of Ne= , iii) Sweden to harbor a part of this total population that substantially contributes to the total Ne, and that is large enough to not be classified as threatened genetically or according to IUCN Red List criteria, and, iv) average inbreeding levels in the Swedish population of <

24 REFLECTIONS My project was from the beginning only on the domestic dogs and there was a lot of data available (without any field or lab work at all), since the SKCs data base is a treasure when working with pedigree analysis, which also has a good reputation foreign researchers while being of such a high quality. One could think that the hard part would be to understand the extensive pedigree analysis literature and that it would be quite an easy job to transpose those dog pedigrees to useable input files for conservation genetic management software s. It was not. At my division, Population genetics, pedigree analysis have been done for a long time and my advisors thesis concerned pedigree analysis (Laikre 1996). In my magister thesis, I evaluated PM2000 (the earlier version of PMx) for conservation genetic monitoring of dog breeding (Jansson 2008), creating the input file structuring the data in MS Excel and Access (programming in SQL), with some programming help from SKC and my boyfriend, James Dickson (SKC provided data from a relation data base where an individual s information existed in several different files and PMx demands all information on one raw). Creating the input file took me a month. When starting my PhD-project I first started to work on was Paper II. I soon realized that the challenge was to develop a method fast enough to do this in dbase would take me forever. Instead I spent years trying to develop a computer program, while realizing more and more problems with the data along the way. I started off programming in SQL where I created a lot of small programs for different steps that I identified. It worked, but it took days to run. Then we contacted a professional programmer associated with our division Ingvar Ståhl. Ingvar and I have had a very interesting, challenging, and fun time working on creating the counterpart to all of my small SQL-modules in one program mped (C). I m very glad for my candidate exam in (computer) system science, which I honestly think I couldn t have done this without, where a lot of focus was on transforming information from experts on the context (in this case pedigree analysis) to the programmer. To explain things correctly and understandable, are true challenges. The main advantages with mped are compared to earlier methods: mped runs for minutes or hours, not days or weeks reducing the manual steps fewer error causes by humans mped can take different types of pedigrees for input, I used it for both the dogs and the wolves in this thesis When working on the actual biological concepts of this thesis, the main challenge might have been to try to communicate the concept of inbreeding (which is often misunderstood) outside the society of researchers and in conservation biology we, of course, like the research to be applied and have a conservation impact. Inbreeding is a common concept in the discussion on dog breeds breeding and keeping wild wolfs in Sweden, both in the Swedish Kennel Club ( and in associations supporting wildlife (Ring 2011) as well as in the popular media (Laikre et al 2010), The former Swedish environment minister, Anders Carlgren, stated that it was a good idea to shoot wolves to reduce the mean inbreeding level, acting as genetic diversity in itself is not an 24

25 issue, only inbreeding level matters (Laikre et al. 2010) and recently current environment minister, Lena Ek, have made statements showing similar opinion. As proposed by wildlife scientist Guillaume Chapron (and popular, Soran Ismail) one should ask other questions, like What are the ecological service provided by reindeer and sheep herding? - How could we plan for a wolf corridor? The burden of adaption should not always be on nature but on humans. I will keep this in mind in my future work with wolves. This thesis contributes to the knowledge of how human selection affects inbreeding and loss of genetic variation and hopefully it will contribute to a more sound conservation genetic management of small populations; whether the species is wild or domestic, the genetic concepts are the same. Genetic variation is needed for evolutionary potential; that includes potential selection for future needs. A writer writes not because he is educated but because he is driven by the need to communicate. Behind the need to communicate is the need to share. Behind the need to share is the need to be understood. (Leo Rosten) 25

26 ACKNOWLEDGEMENTS Tack till alla, nuvarande och tidigare, kollegor på avdelningen för populationsgenetik: Först och främst, tack till mina handledare, Linda Laikre och Nils Nisse Ryman, för att ni givit mig den här möjligheten och för hur generöst delar med er av er kunskap och er tid. Utan diskussioner med er hade denna avhandling helt enkelt inte sett dages ljus. Den här resan är något av det intressantaste jag hittills gjort i mitt liv och jag är glad att jag fått göra den med just er av både proffessionella och personliga skäl. Jag uppskattar er mycket. Karin Tahvanainen ditt lugn och till synes ständigt goda humör är ovärdeligt, tack för det och för koncentrerad korrläsning. Ingvar Ståhl ditt bidrag till denna avhandling är så mycket större än jag kunde ha hoppats på när du fick uppdraget. Detta är jag väldigt tacksam för, liksom för de tillfällen vi har träffats på olika parkeringar i Stockholmsnatten för att lösa mina akuta (det känns ju så just då) datorproblem. Lovisa Wennerström, jag hade blivit galnare utan dig! Tack för din vänskap och mänsklighet men kanske framför allt för ditt sällskap, för privatundervisningen i statistik och för all trevlig samvaro på pubar, gym, bokklubb och jag vet inte allt. Jo, S:t Johannes kyrka hade jag antagligen inte hittat till utan dig. Inte heller hade jag hittat till fejan. Tidigare doktorandkollegor tack för trevlig social samvaro både på jobbet och på fritiden. Särskilt tack till Lena Larsson för att du tittat över min amerikanska engelska i denna kappa, Anna Palmé för att du förklarade vad en kappa överhuvudtaget är och Johan Charlier för ett gott mentorskap såväl formellt som informellt. (PS. Johan, i mina ögon vann du utan tvekan popgen-doktorandernas halv-åtta-hos-mig även om alla kvällarna var speciella och trevliga.) Jag vill också passa på att hälsa vår nytillkomna doktorand, Anastasia Eklund, välkommen i gänget! Ett par tidigare examensstuderande, Peter Guban och Rebecka Salamon, har på ett värdefullt sätt bidragit till denna avhandling genom aktivt lyssnande på vägen och trevlig social samvaro. Tack också till medförfattare utifrån : Fred W. Allendorf (University of Montana) och Sven Jakobsson (Tovetorps forskningsstation, Zoologiska institutionen, SU). Alla medarbetare på Zootis tack för en härlig miljö att arbeta i och tack för att ni är som ni är. Varmt tack till Thomas Wink och Håkan Ericson på Svenska kennelklubben för det arbete ni lagt ner för hundmanuset, Thomas genom att försörja mig med data och information och Håkan som SKK-allvetare och kynologisk ordbok. Mari Edman tack för väl utfört arbete med att testa mped och stort tack för all tid du lagt på att korrläsa mina texter i jobbet och privat, jag har lärt mig mycket. Min vän och vapendragare Helene Wallskär tack för allt! I det här sammanhanget självklart mest för ditt implicita bidrag till denna avhandling. 26