In this dissertation, I report new information that is necessary for future mating system studies in a little studied species, the marbled salamander (Ambystoma opacum). I studied female mating behavior, sexual selection, and the consequences of polyandry for individual females and salamander populations. I also compared the performance of several statistical approaches for analyzing genetic mating system data. The first chapter summarizes the characteristics of several novel microsatellite DNA loci as well as cross-amplified loci for marbled salamanders and mole salamanders that may be used for future studies. In the second chapter, I report estimates of sire number for 13 marbled salamander clutches based on microsatellite data from 32 hatchlings per clutch. Females mated with as many as three different males as indicated by conservative techniques. Less than half of females mated with multiple males. Based on comparative analyses, I recommend the parental reconstruction approach with the computer program GERUD for assessing multiple paternity. The third chapter describes an experiment designed to study sexual selection. As expected, in breeding mesocosms, the potential for sexual selection was much higher for males than for females. Size was unrelated to variance in male reproductive fitness. Only opportunity for selection and Morisita’s index conformed to theoretical expectations of the relationship between operational sex ratio and the potential for sexual selection among males. Because opportunity for selection has intuitive links to formal sexual selection theory, I recommend its continued use. In the fourth chapter, I compared polyandrous and monandrous females to explore the potential fitness consequences of multimale mating. No fitness measure at the egg or hatchling stage (clutch size, hatching success, hatchling size, etc.) differed between the two types of clutches. Size of metamorphs was not different, but polyandrous clutches had significantly higher survival to metamorphosis. In the fifth chapter, I analyzed effects of increased polyandry and male availability on genetic diversity, effective population size (Ne), and fitness of experimental populations. Although no analyses were significant, some effects were moderate to high in size. Ne was higher when estimated from hatchlings than with metamorphs.

O presente estudo teve como objetivo avaliar a estrutura populacional, estimar parâmetros (coeficientes de herdabilidade e correlações) e determinar as tendências genéticas e fenotípicas de características de crescimento e adaptação em bovinos da raça Nelore, linhagem Lemgruber. O banco de dados utilizado nas análises constituiu-se de 39.290 animais no arquivo de pedigree e de 24.353 animais no arquivo de produção. Os parâmetros populacionais foram obtidos por meio dos softwares POPREP (GROENEVELD et al., 2009) e ENDOG v 3.2 (GUTIÉRREZ e GOYACHE, 2005). As estatísticas descritivas e os parâmetros genéticos para cada característica estudada foram estimadas por meio de quatro análises hepta-característica utilizando o programa VCE-6 (KOVAC & GROENEVELD, 2003) sob modelo animal completo. As tendências genéticas e fenotípicas foram obtidas pela regressão dos valores genéticos e fenotípicos respectivamente sobre o ano de nascimento dos animais e os coeficientes da regressão foram estimados por meio do método de quadrados mínimos. Apesar da endogamia média do rebanho ser considerada moderada verificou-se alta porcentagem de indivíduos endogâmicos e reduzido tamanho efetivo populacional. Além disso, o incremento contínuo de endogamia ao longo dos anos demonstra a necessidade de intervenção na seleção dos indivíduos para reprodução, de modo que, problemas futuros possam ser evitados. Os coeficientes de herdabilidade direta foram estimados em 0,24, 0,31, 0,31, 0,21, 0,19, 0,30, 0,41, 0,19 e 0,17 respectivamente para peso ao nascimento (PN), peso aos 120 dias de idade (P120), peso à desmama (PD), peso ao ano (PES12), ganho de peso pós desmama, de 205 aos 550 dias (GP345), ganho de peso na prova à pasto da ABCZ, em 224 dias (GP224), perímetro escrotal (PE), temperamento (TEMP), e repelência (REP) e indicam que a variabilidade genética aditiva existente é suficiente para a obtenção do ganho genético em resposta à seleção desde que sejam realizados ajustes, tanto nas estratégias de seleção adotadas para a obtenção de maiores ganhos nos valores genéticos, como também nos fatores ambientais que possam possibilitar ao máximo, a expressão do potencial genético dos animais. / The present study had as objective to evaluate the population structure, to estimate parameters (coefficients of heritability and correlation) and to determine the phenotypic and genetic trends for growth and adaptability traits in Nellore cattle, Lemgruber line. The database used in the analysis consisted of 39,290 animals in the pedigree and 24,353 animals in the production file. The population parameters were obtained from the software POPREP (Groeneveld et al., 2009) and ENDOG v 3.2 (Gutierrez and GOYACHE, 2005). The descriptive statistics and genetic parameters for each characteristic studied were estimated by four seven-traits analysis using the program VCE-6 (Kovac & GROENEVELD, 2003) which uses the animal model. The phenotypic and genetic trends were obtained by means of phenotypic and genetic values respectively on the year of birth of the animals and the regression coefficients were estimated by the method of least squares. Although the average inbreeding of the herd being considered moderate, there was high percentage of inbred individuals and small effective population size. Furthermore, the continuous increase in inbreeding over the years demonstrates the need for intervention in the selection of individuals for reproduction, so that future problems can be avoided. Direct heritability coefficients were estimated as 0.24, 0.31, 0.31, 0.21, 0.19, 0.30, 0.41, 0.19 and 0.17 respectively for birth weight (PN), weight 120 days of age (P120), weaning weight (PD), weight at 12 months of age (PES12), weight gain after weaning from 205 to 550 days (GP345), weight gain during the pasture test of ABCZ in 224 days (GP224), scrotal circumference (PE), docility (TEMP) and repellency (REP) and indicate that the additive genetic variability is sufficient to obtain genetic gain in response to selection as far as adjustments in selection strategies are adopted to achieve larger gains in breeding values, and environmental factors that may allow the expression of the genetic potential of the animals.

<p> </p><p> </p><p> </p><p>The extinction risk of threatened species has traditionally been assessed by the use of tools of Population Viability Analysis (PVA). Species interactions, however, have seldom been accounted for in PVA:s. The omission of species interactions in risk assessments may further lead to serious mistakes when setting target sizes of populations. Even a slight abundance decrease of a target species may result in changes of the community structure; in the worst case leading to a highly impoverished community. Of critical importance to conservation is therefore the question of how many individuals of a certain population that is needed in order to avoid this kind of consequences. In the current study, a stochastic multispecies model is used to estimate minimum ecological viable populations (MEVP); earlier defined as “the minimum size of a population that can survive before itself or some other species in the community becomes extinct”. The MEVP:s are compared to population sizes given by a single species model where interactions with other species are treated as a constant source incorporated in the species specific growth rate. MEVP:s are found to be larger than the population sizes given by the single species model. The results are trophic level dependent and multispecies approaches are suggested to be of major importance when setting target levels for species at the basal level. Species at higher trophic levels, however, are altogether more prone to extinction than species at the basal level, irrespective of food web size and food web complexity.</p><p> </p>

<p>Sex chromosomes are useful for the study of how factors such as mutation, selection, recombination and effective population size affect diversity and divergence.</p><p>A comparison of gametologous introns in seven different bird species revealed a complete lack of diversity on the female-specific W chromosome. In contrast, Z had at least one segregating site in all examined species. This can be explained by the lower mutation rate and lower effective population size of W but also suggests that selection affects diversity levels on the non-recombining W chromosome.</p><p>In a diverse set of chicken breeds, the Z chromosome showed reduced diversity compared to autosomes and significant heterogeneity in levels of variation. High variance in male reproductive success, leading to a reduced Z chromosome effective population size, can partly explain this observation. In addition, we suggest that selective sweeps frequently act on the Z chromosome and are responsible for a significant part of the observed Z reduction. </p><p>Differences in the mutation rate of Z and W chromosome sequences indicate that the time spent in male germ line is important for the mutation rate, but does not exclude a specifically reduced mutation rate on the Z chromosome. Estimates of mutation rate in autosomal, Z- and W-linked chicken and turkey sequences indicate a slight reduction in the rate on Z. However, due to rate heterogeneity among introns this reduction is not significant and we cannot exclude male biased mutation as the single cause of rate variation between the chromosomal classes.</p><p>Analysis of indel mutation rates in avian and mammalian gametologous introns show frequent occurrence of indels on both W and Y, excluding meiotic recombination as the only source of this type of mutation. The different indel rate patterns in birds (Z>W) and mammals (X=Y) suggest that indels are caused by both replication and recombination.</p>

<p>To shed light on forces that shape the molecular evolution of bird genomes, and in turn avian adaptations, comparative analyses of avian DNA sequences are important. Moreover, contrasting findings in birds to those of other organisms will lend a clearer view on general aspects of molecular evolution. However, few such analyses have been conducted in birds. Progress is presented in this thesis.</p><p>Theories predict a reduction in the mutation rate of the Z chromosome as the harmful effects of recessive mutations are exposed in female birds. We find no evidence for this. Instead, the substitution rates of sex chromosomes and autosomes are largely compatible with expectations from male-biased mutation. This suggests that a majority of mutations arise during DNA replication.</p><p>Substitution rates also vary across chicken autosomes. For instance, microchromosomes accumulate ~20% more substitutions than macrochromosomes. We show that a majority of the autosomal variation in substitution rate can be accounted for by GC content, mainly due to the incidence of mutable CpG-dinucleotides.</p><p>Sequence comparisons also show that the pattern of nucleotide substitution varies in the chicken genome and this reinforces regional differences in base composition. </p><p>The level of selective constraint in at least some avian lineages is higher than in mammalian lineages as indicated by low<i> d</i><i>N</i><i>/d</i><i>S</i><i> </i>– ratios. Larger historical population sizes of birds relative to mammals could explain this observation. Within the avian genome, the<i> d</i><i>N</i><i>/d</i><i>S</i> is lower for genes on micro- than macrochromosomes, potentially owing to a higher incidence of house-keeping genes in the former category.</p><p>Contrasting data on non-synonymous and synonymous substitution for divergence and polymorphism shows that positive selection has contributed more to the evolution of Z-linked than autosomal genes. This is likely explained by the full exposure of beneficial recessive mutations on Z when in female birds.</p>

Sex chromosomes are useful for the study of how factors such as mutation, selection, recombination and effective population size affect diversity and divergence. A comparison of gametologous introns in seven different bird species revealed a complete lack of diversity on the female-specific W chromosome. In contrast, Z had at least one segregating site in all examined species. This can be explained by the lower mutation rate and lower effective population size of W but also suggests that selection affects diversity levels on the non-recombining W chromosome. In a diverse set of chicken breeds, the Z chromosome showed reduced diversity compared to autosomes and significant heterogeneity in levels of variation. High variance in male reproductive success, leading to a reduced Z chromosome effective population size, can partly explain this observation. In addition, we suggest that selective sweeps frequently act on the Z chromosome and are responsible for a significant part of the observed Z reduction. Differences in the mutation rate of Z and W chromosome sequences indicate that the time spent in male germ line is important for the mutation rate, but does not exclude a specifically reduced mutation rate on the Z chromosome. Estimates of mutation rate in autosomal, Z- and W-linked chicken and turkey sequences indicate a slight reduction in the rate on Z. However, due to rate heterogeneity among introns this reduction is not significant and we cannot exclude male biased mutation as the single cause of rate variation between the chromosomal classes. Analysis of indel mutation rates in avian and mammalian gametologous introns show frequent occurrence of indels on both W and Y, excluding meiotic recombination as the only source of this type of mutation. The different indel rate patterns in birds (Z&gt;W) and mammals (X=Y) suggest that indels are caused by both replication and recombination.

To shed light on forces that shape the molecular evolution of bird genomes, and in turn avian adaptations, comparative analyses of avian DNA sequences are important. Moreover, contrasting findings in birds to those of other organisms will lend a clearer view on general aspects of molecular evolution. However, few such analyses have been conducted in birds. Progress is presented in this thesis. Theories predict a reduction in the mutation rate of the Z chromosome as the harmful effects of recessive mutations are exposed in female birds. We find no evidence for this. Instead, the substitution rates of sex chromosomes and autosomes are largely compatible with expectations from male-biased mutation. This suggests that a majority of mutations arise during DNA replication. Substitution rates also vary across chicken autosomes. For instance, microchromosomes accumulate ~20% more substitutions than macrochromosomes. We show that a majority of the autosomal variation in substitution rate can be accounted for by GC content, mainly due to the incidence of mutable CpG-dinucleotides. Sequence comparisons also show that the pattern of nucleotide substitution varies in the chicken genome and this reinforces regional differences in base composition. The level of selective constraint in at least some avian lineages is higher than in mammalian lineages as indicated by low dN/dS – ratios. Larger historical population sizes of birds relative to mammals could explain this observation. Within the avian genome, the dN/dS is lower for genes on micro- than macrochromosomes, potentially owing to a higher incidence of house-keeping genes in the former category. Contrasting data on non-synonymous and synonymous substitution for divergence and polymorphism shows that positive selection has contributed more to the evolution of Z-linked than autosomal genes. This is likely explained by the full exposure of beneficial recessive mutations on Z when in female birds.

The aims of this thesis were to evaluate the effects of agriculture on amphibians in terms of (i) population genetic consequences of agriculture-induced spatial changes of the landscape and (ii) local adaptation and tolerance to frequently used agrochemicals. The study was performed using the common frog Rana temporaria as a model. Abundance, occurrence, genetic diversity and gene flow were negatively affected by agriculture in southern Sweden, but unaffected or even positively affected by agriculture in the central and northern regions, respectively. These test parameters correlated positively with landscape diversity both in the south and in the north. Moreover, the size and occurrence of R. temporaria populations decreased towards the north i.e. the margin of the species’ distribution range. In accordance with theoretical expectations, genetic variability decreased and population substructuring increased as a negative function of (effective) population size. Southern Swedish common frogs are naturally exposed to higher levels of nitrates, and thus have a higher tolerance to high nitrate levels than their northern conspecifics. This suggests local adaptation to naturally varying nitrate levels. Consequently, increased anthropogenic supplementation of nitrate could impact more the northern than the southern Swedish common frog populations. Exposure to the pesticides azoxystrobin, cyanazine and permethrin at ecologically relevant concentrations had small or no effects on R. temporaria tadpoles. The populations with lowest microsatellite variation (fragmented populations) in southern Sweden had considerably lower fitness in terms of survival and growth as compared to those with the highest genetic variability (non-fragmented populations). The results indicate that populations with low levels of neutral genetic variability were phenotypically less differentiated than populations with higher levels of variability. One possible explanation for this is that the degree of population differentiation in low variability populations has been constrained due to lack of suitable genetic variation or inefficiency of selection relative to genetic drift.

In this thesis, molecular genetic methods were used to study a threatened species, the Eurasian otter. Estimates of population size and population dynamics parameters were obtained, the genetic effects of a restocking program was evaluated, and a population viability analysis was conducted to assess which demographic parameters are most important for the future viability of an otter population. Many of the studies were based on noninvasive genetic sampling of faeces. In the genetic evaluation of the restocking program, it was found that the released otters had contributed to subsequent generations. However, the effects were to a large degree limited to the near surroundings of the release areas. Comparison of two census methods, snow-tracking and noninvasive genetic census based on faeces, showed that approximately only half of the otters detected with the genetic census were found with the snow-tracking census. It is recommended to combine these two methods to obtain the most reliable estimates of population size. A short-term study on population dynamics in otters showed that apparent survival was higher in females than in males and that the rate of addition was also high and likely influenced by migration. The population viability analysis incorporated both genetics and demography and revealed that survival to first reproduction was the most crucial demographic parameter affecting the viability of the study population. This result suggests that conservation efforts should be focused on protocols that enhance the survival prospects of young females. Environmental stochasticity was also found to have large effects on the probability of extinction of this population.

The extinction risk of threatened species has traditionally been assessed by the use of tools of Population Viability Analysis (PVA). Species interactions, however, have seldom been accounted for in PVA:s. The omission of species interactions in risk assessments may further lead to serious mistakes when setting target sizes of populations. Even a slight abundance decrease of a target species may result in changes of the community structure; in the worst case leading to a highly impoverished community. Of critical importance to conservation is therefore the question of how many individuals of a certain population that is needed in order to avoid this kind of consequences. In the current study, a stochastic multispecies model is used to estimate minimum ecological viable populations (MEVP); earlier defined as “the minimum size of a population that can survive before itself or some other species in the community becomes extinct”. The MEVP:s are compared to population sizes given by a single species model where interactions with other species are treated as a constant source incorporated in the species specific growth rate. MEVP:s are found to be larger than the population sizes given by the single species model. The results are trophic level dependent and multispecies approaches are suggested to be of major importance when setting target levels for species at the basal level. Species at higher trophic levels, however, are altogether more prone to extinction than species at the basal level, irrespective of food web size and food web complexity.