<p>The legally binding international declaration of the Convention on Biological Diversity (signed by over 180 countries) recently acknowledged the importance of conserving genetic diversity within livestock species. This study aimed to help Canada assess molecular diversity in its horse and pony (<i>Equus ferus caballus</i>) genetic resources. Here, 24 populations were examined, with special focus on the native Canadian, Mountain and Moorland, and Nordic pony populations, using two well accepted molecular tools. Additional horse breeds and feral populations were also included in this project as some may have influenced the development of the three equine groups of interest. Altogether, 821 individuals were genotyped at 38 microsatellite loci, and 280 individuals were sequenced using a 421 base pair portion of the mitochondrial displacement Hypervariable Region I.</p> <p>Results from the microsatellite analyses indicated that 13.33% of genetic diversity arose from breed differences, whereas 84.60% and 2.07% of diversity arose from within and among individuals respectively. The New Forest and Welsh breeds were found to be the most diverse while having the highest average effective number of alleles and allelic richness (4.31 and 6.01; 4.33 and 5.87 respectively). The Eriskay and Lac La Croix breeds were found to have the lowest average effective number of alleles and allelic richness (2.51 and 3.98; 2.83 and 4.01 respectively). Expected heterozygosities were lowest in the Lac La Croix (0.61) and highest in the Welsh and New Forest (0.74) breeds, whereas observed heterozygosities were highest in the Kerry Bog (0.77) and lowest in the Exmoor (0.57) breeds. The genetic structure and admixture analyses suggested that the most probable number of unique genetic clusters was 21 as opposed to the 24 predefined populations.</p> <p>Results from the mitochondrial sequence data revealed that there were 36 informative sites producing 62 haplotypes, 20 of which were previously unreported. The Connemara was found to have the highest haplotype diversity of the pony breeds (0.89); however, the Highland pony was found to have the highest nucleotide diversity and pairwise difference (0.16 and 6.73 respectively). In contrast, the Fell pony had the lowest haplotype diversity (0.22), and the feral Sable Island population had the lowest nucleotide diversity and pairwise difference (0.01 and 0.29 respectively). Multiple phylogenetic trees were reconstructed and produced similar topologies. In general, the Mountain and Moorland and Nordic breeds were spread among the clades, whereas native Canadian populations were most frequent in the D and E clades. Interestingly, a large portion of ponies were found within the rare E clade as opposed to horses.</p> <p>Information gathered from this project can be incorporated with other available data into pre-existing conservation/breeding programs currently managed by the various breed societies to ensure that the most optimal and sustainable strategies are in place.</p>

<p>The legally binding international declaration of the Convention on Biological Diversity (signed by over 180 countries) recently acknowledged the importance of conserving genetic diversity within livestock species. This study aimed to help Canada assess molecular diversity in its horse and pony (<i>Equus ferus caballus</i>) genetic resources. Here, 24 populations were examined, with special focus on the native Canadian, Mountain and Moorland, and Nordic pony populations, using two well accepted molecular tools. Additional horse breeds and feral populations were also included in this project as some may have influenced the development of the three equine groups of interest. Altogether, 821 individuals were genotyped at 38 microsatellite loci, and 280 individuals were sequenced using a 421 base pair portion of the mitochondrial displacement Hypervariable Region I.</p> <p>Results from the microsatellite analyses indicated that 13.33% of genetic diversity arose from breed differences, whereas 84.60% and 2.07% of diversity arose from within and among individuals respectively. The New Forest and Welsh breeds were found to be the most diverse while having the highest average effective number of alleles and allelic richness (4.31 and 6.01; 4.33 and 5.87 respectively). The Eriskay and Lac La Croix breeds were found to have the lowest average effective number of alleles and allelic richness (2.51 and 3.98; 2.83 and 4.01 respectively). Expected heterozygosities were lowest in the Lac La Croix (0.61) and highest in the Welsh and New Forest (0.74) breeds, whereas observed heterozygosities were highest in the Kerry Bog (0.77) and lowest in the Exmoor (0.57) breeds. The genetic structure and admixture analyses suggested that the most probable number of unique genetic clusters was 21 as opposed to the 24 predefined populations.</p> <p>Results from the mitochondrial sequence data revealed that there were 36 informative sites producing 62 haplotypes, 20 of which were previously unreported. The Connemara was found to have the highest haplotype diversity of the pony breeds (0.89); however, the Highland pony was found to have the highest nucleotide diversity and pairwise difference (0.16 and 6.73 respectively). In contrast, the Fell pony had the lowest haplotype diversity (0.22), and the feral Sable Island population had the lowest nucleotide diversity and pairwise difference (0.01 and 0.29 respectively). Multiple phylogenetic trees were reconstructed and produced similar topologies. In general, the Mountain and Moorland and Nordic breeds were spread among the clades, whereas native Canadian populations were most frequent in the D and E clades. Interestingly, a large portion of ponies were found within the rare E clade as opposed to horses.</p> <p>Information gathered from this project can be incorporated with other available data into pre-existing conservation/breeding programs currently managed by the various breed societies to ensure that the most optimal and sustainable strategies are in place.</p>

Population-level dynamics are affected by temporal variation in individual vital rates of survival and reproduction, which are in turn influenced by habitat-specific processes. Variation in habitat quality within a population’s range can drive movement of individuals between different areas, and so there may be a relationship between variation in vital rates and spatial heterogeneity in population growth (λ). I investigated this relationship for the feral horses (Equus ferus caballus) of Sable Island, Nova Scotia, Canada, from 2008−2010. The horses (n = 484 in September 2010) form a closed population that is free from human interference and predation. I analyzed annual population growth using age-structured projection matrix models parameterized with survival and fertility data collected from almost every female (98.7% of females). I found some evidence of temporal variation in growth during the two years I studied the population (λ2008−2009 = 1.065, λ2009−2010 = 1.117). Age structure appears to have converged to a stable age distribution, suggesting this growth rate has been sustained in the years leading up to the end of my study. Variation in vital rates of adult fertility and foal survival made the largest contribution to annual variation in population growth. Future growth is predicted to be most influenced by proportional changes in adult survival, which remained relatively unchanged between 2008 and 2010. The population can be stratified into three spatially distinct subunits found across a west−east longitudinal gradient of water resources (access to permanent ponds vs. ephemeral water sources and holes dug in sand). I assessed the existence of source-sink dynamics to determine if individual movements between subunits could explain spatial heterogeneity in population growth. I found that spatial heterogeneity in growth appears to be most influenced by immigration and emigration events between subunits. Evidence suggests that current growth of the overall Sable Island horse population is made possible by individual emigration from more productive into less productive subunits; in particular, a source presented in the west of the island where permanent water ponds are located.

Body size is an important determinant of reproduction in capital breeding animals, including large mammals. However, it is not always practical to hand-measure body size of free-ranging species. In recent years, parallel-laser photogrammetry has been used to obtain remote estimates of body size for some animals, though it remains unknown how well this technique might capture variation in curvilinear body features or if the distance between parallel-laser calipers is altered when projected onto a curved surface. In this thesis, I describe a photogrammetric system that may be useful for obtaining body-size measurements from unrestrained large mammals that permit approach, using domestic horses (Equus ferus caballus) as a model (Chapter 2). I then apply this technique in the field to a wild (feral) population of horses at Sable Island National Park Reserve, Nova Scotia, Canada, where I include body size measurements as variables in a detailed analysis of factors affecting reproduction in females (Chapter 3). Using my parallel-laser photogrammetric system, I show how curvilinear hand-measurements (e.g., across the barrel of a horse) are stongly correlated with their respective linear photogrammetric estimates (R2 ≥ 0.998), and most photogrammetric estimates using my system had high reliability. Using three variables of body size, photogrammetric estimates and hand-measurements explained 86.0% and 96.2%, respectively, of the variation in body weight of a sample of domestic Newfoundland ponies. On Sable Island, Nova Scotia, I examined the relationship of numerous variables (including skeletal body size and body condition) with the probability of yearly reproductive success for female Sable Island horses (years 2008–2012), where I define reproductive success as production of an offspring surviving to one year of age. Age class was a dominant factor predicting reproductive success, as expected from trends previously associated with body size or reproductive experience iii in other populations. Age-class specific energy budgets or social and sexual behaviour caused a more pronounced relationship with body condition at parturition in sub-adults, and body condition at conception and stability of consort relationships were associated with reproductive success in adults. In addition, relationships with local density suggested limited forage around the time of conception and limited water during lactation might also influence reproductive success in adult females. Although relationships were evident for age class, which is correlated with body size, reproductive success was not related to skeletal body size, past reproductive experience, age of primiparity, or band structure. The capital breeding strategy and year-round social associations seen in horses make their reproductive ecology a combination of patterns observed for large ungulates and social primates.