Search

Despite a long-standing interest in understanding inbreeding, infidelity and cooperation in animal populations we still do not have a full picture of the interplay between these factors, especially in wild populations. However, the development of paternity assignment methods and statistical tools provides us with an opportunity to gain new levels of insight. In this thesis I combine behavioural data I collected during fieldwork with long-term records from a study of a superb fairy-wren (Malurus cyaneus) population based in the Australian National Botanic Gardens. I investigate mating strategies in general, and inbreeding avoidance in particular, as well as assess the role of social environment in context-dependent mate choice in this cooperative breeder. First, in Chapter 2, I reconstruct a genetically-informed pedigree of the superb fairy-wren population using 26 years of data, in order to quantify the rates of inbreeding (only 6% of individuals were inbred) and extra-pair reproduction (61% of offspring had extra-pair paternity). I then test whether these high rates of infidelity might have evolved as a plausible mechanism for inbreeding avoidance in this population. I found evidence of inbreeding depression in nestling mass, but not in fledgling survival. Kinship between social partners did not predict infidelity, except in the case of mother-son social pairings, for which there was 100% infidelity. Nevertheless, extra-pair offspring were less likely to be inbred than within-pair offspring. Moreover, social environment - the number of helpers in a group - did not affect inbreeding or inbreeding depression. I concluded that, overall, inbreeding avoidance is unlikely to explain the extreme levels of infidelity in this study system. In Chapter 3, I explore the effects of the social environment, and in particular different levels of relatedness of helpers in a group, on the mating patterns of the dominant female. My analysis demonstrated that whilst the presence of helper-sons did not affect a female's infidelity to her social partner, her rates of infidelity increased in the presence of unrelated helpers. The presence of unrelated helpers in a group was associated with increased rates of both extra-pair paternity and, unexpectedly, of extra-group paternity. Across a total of 1936 broods over 26 years, broods produced in groups assisted by at least one unrelated helper contained 67% extra-group offspring, compared to 58% in groups with only helper-sons. These findings suggest not only that the 'constrained female hypothesis' cannot explain the high levels of infidelity observed in this species, but also that the social environment can affect conspecific interactions in complex ways. I then investigate mate choice and inbreeding avoidance from the perspective of the males in Chapter 4. Over two field seasons I collected behavioural data on male visits to female territories and data on courtship displays. I found that sons rarely visited (< 2% of visits) and never displayed to their mothers. My results thus indicate that mother-son inbreeding avoidance is not driven entirely by females: males may face opportunity costs too and thus opt to exercise a level of mate choice through exercising strategic courtship. Finally, in Chapter 5, I explore the factors driving the apparent selection on body size through juvenile survival identified in Chapter 2, by conducting a quantitative genetics analysis on nestling mass and fledging survival using mixed effects animal models. This approach allowed me to partition the variance in the two traits, and the covariance between them, into different sources, including additive genetic variance and covariance. I found that roughly two-thirds of the overall phenotypic variance in mass, and a similar proportion of the overall phenotypic variance in survival, were explained by the effects of nest, hatch date, cohort and additive genetic effect. Nest explained the largest proportion of variance in each of the traits: 43% in mass and 58% in survival. I demonstrated, through estimation of genetic covariances, that the association between mass and survival observed at the phenotypic level had no genetic basis and was instead caused by temporal factors (hatch date and cohort): the result indicates that there is no causal effect of mass on survival, and hence no potential for an adaptive response to selection despite heritability of both mass and survival. I conclude with a discussion of the overall insights provided into the mating system and selection pressures in this wild cooperatively breeding songbird. I draw attention to the limitations of my analyses and suggest avenues for future research.

The evolution of extra-group paternity (EGP) is a contentious issue in evolutionary biology. This thesis examines the factors and adaptive benefits driving EGP in a high-density, group-living population of European badgers (Meles meles). To improve power to assign parentage, I isolated and characterised 21 new polymorphic microsatellite markers. I genotyped 83% of 1410 badger trapped 1987‒2010 using 35 autosomal microsatellite markers. Maternity and paternity were assigned at 80% confidence ca. 82% of individuals. 48% of paternities were extra-group, where 85% were attributable to neighbouring-group males and EGP was detected in 47% of litters; thus badger social group do not correspond with a breeding unit. I tested whether indirect genetic benefits explain these high EGP rates. (1) ‘Good-gene-as-heterozygosity Hypothesis’: Paternal heterozygosity, but not maternal or an individual’s own heterozygosity, associated positively with first-year survival probability. Under benign environmental conditions, cubs fathered by more heterozygous males had a higher first year survival probability. Despite this correlation, the EGP rate per litter correlated with neither average nor maximum within-group heterozygosity of candidate fathers. (2) Fitness benefit Hypothesis: Extra-group offspring (EGO) had lower first-year survival probability and lived 1.3 years less than within-group offspring (WGO). Female WGO produced more litters and offspring over their lifetime than female EGO, whereas male EGO produced more offspring than male WGO. (3) Inbreeding avoidance hypothesis: The EGP rate within a litter increased with greater average pair-wise relatedness between mothers and within-group candidate fathers. No inbreeding depression on first-year survival probability was detected, but small sample sizes limited statistical power. Socio-ecologically, at the litter level, EGP correlated negatively with the number of within-group candidate fathers, and positively with neighbouring-group candidate fathers. In conclusion, EGP in badgers may reduce inbreeding and be maintained in the population through a sex-specific antagonistic selection and indirect genetic benefits may occur when the total fitness benefits of producing extra-group sons outweigh the costs of producing extra-group daughters. These indirect genetic benefits only partially explain the evolution of promiscuity in European badgers, highlighting that evolutionary factors underlying promiscuity remain unclear.

Inbreeding occurs when relatives mate and have offspring. Inbreeding depression is hypothesized to have influenced the evolution of mating systems and behavioural mechanisms of inbreeding avoidance in the animal kingdom. Inbreeding in the wild is difficult to measure, as in order to build a pedigree allowing us to identify matings between relatives, the identity of as many as possible members of a population needs to be known. For a long time, the main source of knowledge about inbreeding depression was based on laboratory and agricultural studies, which did not reflect the array of environmental pressures wild populations have to cope with. In consequence, the deleterious consequences of inbreeding have often been underestimated. This is problematic because accurate estimates of the effect size of inbreeding depression are needed to study the strength of selection on inbreeding avoidance mechanisms, and are also of importance to conservation genetics. The aim of this thesis was to use pedigree data to infer the occurrence and effects of inbreeding using over forty years of breeding events of the great tit Parus major from Wytham Woods, Oxfordshire. The effects of inbreeding on fitness were investigated across a life-history continuum, and across environments. I found that close inbreeding (f=0.25) resulted in pronounced inbreeding depression, which acted independently on hatching success, fledging success, and recruitment success, and reduced the number of fledged grand-offspring by 55%. My results therefore suggest that estimates of fitness costs of inbreeding must focus on the entire life cycle. I also show that the variation in the strength of inbreeding depression varies across environments, particularly so the more the environmental variable considered is linked to fitness. These results emphasise the need of using relevant environmental contrasts when investigating inbreeding by environment interactions. I further asked whether individuals involved in matings with relatives differed relative to individuals mating with unrelated partners. I did not find any evidence for clear predictors of inbreeding, and I show that inbreeding depression in our population is entirely independent of any tendency for low quality parental genotypes, or phenotypes, to inbreed. Neither did I find any evidence for active inbreeding avoidance: great tits did not mate less often with kin than expected based on several scenarios of random mating, nor did I find increased rates of extra-pair paternity among birds breeding with relatives. In fact, I observed quite the contrary, as birds mating with kin exhibited a higher than average rate of close inbreeding relative to all scenarios of random mating investigated, showed lower rates of extra-pair paternity and divorce than birds mated to unrelated partners. I hypothesise that cases of occasional inbreeding in this population may result from mis-imprinting or a related process whereby some birds develop particularly strong bonds that are at odds with all predictions of avoiding inbreeding. Finally, I asked to what extent natal dispersal, a behaviour that is often hypothesized to play an important role in avoiding inbreeding, indeed reduces the likelihood of inbreeding. I found that male and female individuals breeding with a relative dispersed over several-fold shorter distances than those outbreeding. This led to a 3.4 fold increase (2.3-5, 95% CI) in the likelihood of close inbreeding relative to the population average when individuals dispersed less than 200m. This thesis demonstrates that inbreeding has deleterious effects on a wild population of birds, occurring throughout an individual’s life, and is of varying strength across environments. My findings strongly support the theory that natal dispersal should be considered as a mechanism of prime importance for inbreeding avoidance.

Humans are altering natural systems around the globe in myriad ways. For plant species, such anthropogenic changes have resulted in increasingly fragmented populations, desynchronized interactions with mutualists, and shifted geographic ranges, among other effects. However, despite numerous examples of human impacts on plant populations, the consequences of these changes on plant reproduction remain poorly understood. In my thesis, I investigate the impacts of two forms of anthropogenic change–habitat disturbance and climate warming–on plant reproduction and fitness. I take two distinct approaches to address questions posed at local and regional scales. In Chapter I, I ask how inbreeding depression varies across the life cycle of the critically imperiled California endemic species, Collinsia corymbosa (Plantaginaceae). I show that, consistent with other primarily outcrossing species, inbreeding depression in C. corymbosa is most pronounced late in life history, specifically during the female reproductive phase of the life cycle. Inbred plants demonstrated significantly lower rates of autonomous autogamy (δ [delta] = 0.448) and flower production (δ [delta] = 0.225), limiting the ability of this species to set seed in the absence of pollinators. In Chapter II, I ask whether flowering and fruiting dates have advanced for 14 spring-flowering plant species in the Blue Ridge and Ridge & Valley ecoregions of eastern Tennessee over the past century. Additionally, I investigate how phenological sensitivity to spring temperature varies between ecoregions. Utilizing phenological observations sourced from 1000+ digitized herbarium specimens, I show that the sensitivity of flowering phenology to spring temperature at the community level varies between the Ridge & Valley and Blue Ridge (2.7 and 1.3 days earlier per degree Celsius warming, respectively). Further, I show that, while the flowering phenology of the majority of species investigated is sensitive to spring temperature in both ecoregions, flowering and fruiting dates have not significantly advanced in recent decades. Overall, I found variation in plant reproductive responses to anthropogenic change at the maternal family, population, species, community, and regional levels. Together, my research demonstrates that assessing reproduction and fitness at these multiple scales provides nuanced insights into the adaptive capacity and ultimate persistence of species in the Anthropocene.

The wheat crop in the United States, and indeed globally, has shown high levels of genetic uniformity. A lack in genetic variation is causing dramatic fluctuations in total wheat crop yields in response to environmental conditions and has created serious production risks. This study was undertaken as a step to address this issue with an aim of develop a new high-performing line incorporating wild wheat germplasm for the Colorado State University breeding program. The purpose of using wild germplasm is to provide a source of genetic variation not currently present in the US wheat crop. This study grew 72 different crosses with eight controls and included two reps each for both an irrigated and unirrigated treatment. Traits that were assessed included grain yield, harvest index (HI), heading date, plant height, lodging and other physical features that were assessed visually. The results showed that one first generation cross (HRS2015-378) clearly stood out above all other lines, including the commercial cultivar controls. HRS2015-378 was ranked first in both treatments for HI along with third and fifth for grain yield in the dry and wet treatments respectively. The values for HI of this line were 0.34 and 0.52 with grain weights from 1m biomass strips of 118 and 150g for the dry and wet treatments, respectively. This cross along with three others, two of which were also first generation crosses and one a second generation cross, were selected for the Colorado State University breeding program for their consistently high performance compared to the commercial controls in all traits measured across both treatments. In the coming years these crosses will undergo more selective and intense breeding across multiple locations with the expectation that at least one will eventually be released as a new commercial variety.

The goals of captive breeding programs for endangered species include preserving genetic diversity and avoiding inbreeding. Typically this is accomplished by minimizing population mean kinship; however, this approach becomes less effective when errors in the pedigree exist and may result in inbreeding depression, or reduced survival. Here, both pedigree- and DNA-based methods were used to assess inbreeding depression in the critically endangered Attwater’s prairie-chicken (Tympanuchus cupido attwateri). Less variation in the pedigree-based inbreeding coefficients and parental relatedness values were observed compared to DNA-based measures suggesting that errors exist in the pedigree. Further, chicks identified with high parental DNA-based relatedness exhibited decreased survival at both 14- and 50-days post-hatch. A similar pattern was observed in later life stages (> 50 days post-hatch) with birds released to the wild; however, the pattern varied depending on the time post-release. While DNA-based inbreeding coefficient was positively correlated with mortality to one month post-release, an opposite pattern was observed at nine months suggesting purging of deleterious alleles. I also investigated whether immunocompetence, or the ability to produce a normal immune response, was correlated with survival; however, no significant correlation was observed suggesting that inbreeding was a more important factor influencing survival. Pairing individuals for breeding by minimizing DNA-based parental relatedness values resulted in a significant increase in chick survival. This study highlights the importance of using DNA-based methods to avoid inbreeding depression when errors exist in the pedigree.

Although the process by which fertilization occurs in bryozoans is well described, the ability to self-fertilize and the subsequent ecological consequences are poorly understood. Culturing experiments were conducted examining the effects of selfing on offspring survival and reproduction in the simultaneous hermaphrodite Bugula stolonifera collected from Eel Pond, Woods Hole, MA. Results from these experiments document significant decreases in survival and fecundity of selfed offspring, compared to outcrossed controls, suggesting that these animals are not routinely self-fertilizing in Eel Pond. How these arborescent colonies minimize selfing remains unclear, but it is hypothesized that conspecific aggregations could serve to minimize the chances that a colony utilizes its own sperm for fertilization. The genetic composition of these aggregations was investigated using a newly developed microsatellite library. As larvae routinely metamorphose on conspecific colonies, the possibility that larvae select or avoid their maternal colony was also investigated. Analyses of genetic structure document homogeneity throughout these aggregations on extremely small spatial scales, suggesting high amounts of larval dispersal within aggregations. When combined with results from parentage-exclusion and kinship analyses, these results indicate that a colony's nearest neighbors are not composed of siblings, potentially minimizing inbreeding. Molecular analyses were then used to determine if the high larval dispersal within aggregations resulted in high mixing between aggregations. Sites within Eel Pond separated by 100-300 m were routinely sampled from 2009 to 2011, and analyses were conducted to investigate potential inter- and intra-annual genotypic differentiation within and between aggregations. Results document that although low levels of mixing could result in increased homogeneity between some aggregations, barriers to genetic exchange prevent mixing between most sites. Further, inter-annual comparisons within sites document that significant differentiation can occur between reproductive seasons. Hence, any potential homogeneity achieved between sites during one reproductive season will likely be lost by the beginning of the next reproductive season. Additionally, while sampling in Eel Pond in 2010, I document the first occurrence from the western Atlantic Ocean of another aggregating arborescent bryozoan, Tricellaria inopinata. The growth and reproductive biology of these animals was monitored throughout 2011; results suggest that this introduction is likely to persist.

Self-incompatibility is thought to have played a profound role in the evolution of the angiosperms. However, there is little evidence of self-incompatibility systems in early diverging lineages of flowering plants. Illicium parviflorum, one such early-divergent angiosperm, is an evergreen perennial species endemic to central Florida, particularly within the Ocala National Forest. Although locally abundant, I. parviflorum is currently listed as endangered at the state level due to being under constant threat of habitat disturbance and over-harvesting. Notably, this species had been described as self-incompatible due to its low seed-set. However, low seed set may also be a result of strong, early inbreeding depression. Using cross-pollinations, histology, and molecular analysis, I provide conclusive evidence that I. parviflorum possesses the ability to self-fertilize, while finding no evidence of a self-incompatibility system. Furthermore, cross-pollinations of individuals within and between populations revealed heterosis, while seeds collected from self-pollinations were smaller than those collected from out-crosses, suggesting that inbreeding may be reducing fitness within populations. An analysis used to estimate parental genotypes of individuals in a population using AFLP markers identified two out of 23 plants to be the result of natural self-pollination, while the mean (+ s.e.) pollen: ovule ratio of I. parviflorum was found to be 511 + 86, a ratio consistent with a species that relies primarily, but not exclusively, on outcrossing. Pollen/ovule ratios of I. parviflorum and other small flowered Illicium are lower than their larger flowered, derived relatives, suggesting that the ancestral floral type to the Illicium lineage was self-compatible. These results support the hypothesis that early angiosperm species had the ability to self-fertilize and that self-incompatibility systems did not arise until after the origin of the bisexual flower.

Many evolutionary and ecological factors can have strong and significant effects in small isolated populations when compared to larger, more connected populations. For example, isolated populations of endangered species are often at greater risk for extinction because of the potential loss of genetic diversity and the negative effects of inbreeding and inbreeding depression. Reduced levels of gene flow among isolated populations may also lead to differentiation in important ecological factors such as diet. The eastern massasauga (Sistrurus c. catenatus) is an endangered snake that is found in isolated populations throughout eastern North America. In this dissertation, I address both genetic and ecological consequences of these snakes existing in small isolated populations. In Chapter 2, I used data from 19 microsatellite loci to assess levels of genetic variation and to compare rates of historical and contemporary gene flow among populations of the these snakes. Overall, massasauga populations display high levels of genetic differentiation and differ by an order of magnitude in genetic effective population size. Historical and contemporary migration rates are low and similar in magnitude. There is little evidence that populations have experienced a sharp decline in population size and a comparative modeling approach favors a model of long-term drift-migration equilibrium.In Chapter 3, levels of inbreeding were accessed in these populations and I tested for the presence of a heterozygosity-fitness correlation between individual multilocus heterozygosity (MLH) and an estimate of fitness, body condition This analysis revealed significant variation in mean MLH and mean body condition among populations but no correlation between them. This result was true even after controlling for non-genetic factors such as sex, season of capture and year of capture. Tests for inbreeding were also negative and suggest snakes are mating randomly within each population. Finally, in Chapter 4, I used stable isotope analysis of carbon and nitrogen to study the patterns of variation in diet over time and space across the range of this species. I employed a set of novel techniques in isotope biology to assess the relative importance of among-population or –individual variation and then calculated Manly alpha selectivity indices to determine if prey preference exists for mammalian prey species. Stable isotope analysis revealed diet is variable among populations with shrews and voles being among the most important prey items. A preference for shrews was observed in two of three populations and an avoidance of voles was observed in all three populations. Overall, my results indicate populations have been isolated for thousands of years and that recent habitat fragmentation caused by humans has had little effect on the patterns of genetic variation in this species. Although populations have been isolated over evolutionary timescales they appear to have retained a large enough size to avoid inbreeding and the negative effects of inbreeding depression. The lack of any correlation between genetics and fitness implies genetics may be less important for the overall fitness of these snakes when compared to ecological factors. Stable isotopes appear to be a viable method for studying diet in snakes and this analysis represents an important first step in understanding if variation in body condition is a result of variation in diet.

In these studies, we determined the quality of Trichogramma parasitoids reared in several Mexican insectaries, evaluated the clonal variability found in Trichogramma pretiosum populations where the infection has gone to fixation, and determined the potential for genetic changes taking place during mass rearing of Trichogramma spp. The results showed that there is substantial potential for adaptation to mass rearing conditions when mass reared populations are initiated with genetically variable populations. In addition, we determined that inbreeding depression is possible even in species such as Trichogramma pretiosum where their sex determination and normal sibmating mating system would predict a minimal effect of inbreeding. With respect to the quality of mass reared Trichogramma, we found discrepancy between the reported and the reared specie, presence of unnoticed species replacement, and we found that the reared species barely met the minimum standards suggested by the international organization for biological control/ European community. From the PI-Wolbachia infected population, we found a surprisingly large amount of variability among the different clonal lines when life history measurements were made. In this dissertation, we also outlined several guidelines for insectaries regarding how to maintain high "field quality" of the natural enemies they produce. Finally, our findings of the high diversity of clones in entirely asexual populations may result in a more rational future use of asexual wasps in biological control.

This study focused on soybean heterosis in regard to yield, seed size, and height and consisted of three objectives: (1) to study three new hybrids and retest a fourth hybrid for heterosis, (2) to study the correlation between heterosis in the F2 generation and inbred line variance, (3) to find SSR markers that help to explain heterosis in a heterotic hybrid. Objective one was evaluated in 2007 and 2008. The four hybrids were tested in yield trials that consisted of F1 - F4 generations and parents. Results showed some heterosis, but there was a lack of consistency across years due to a genotype by year effect. The second objective was evaluated in 2008. Yield trials examining six populations were tested over three locations. The hybrid seed were developed in 2007 and F1 plants were grown to generate F2 bulks, the inbred lines were developed by single seed descent and tested as F6:8 lines. A correlation between heterosis and transegregate lines was found for yield and seed size. For the third objective to identify SSR markers for heterosis, the cross of Hutcheson x Holladay was used to develop F2:3 and F2:4 lines. A DNA sample was taken from 360 individual F2 plants and screened with 40 polymorphic markers that represented each arm of the 20 linkage groups. Yield, seed size and height were taken on the F2:3 and F2:4 lines and used to evaluate the marker data. No consistent marker for additive and dominance variance was found between the generations.

New Zealand ecologists and wildlife managers have assumed traditionally that threatened insular endemic bird species are less susceptible to the predicted genetic consequences of declines and inbreeding. Conservation has relied on treating deterministic extinction pressures to trigger population recoveries, with few measures taken to minimise stochastic genetic threats to long-term viability. This study tested that assumption empirically by exam-ining extinction vulnerability in two small, critically threatened Chatham Island black robin populations protected from external threats on habitat-restricted islands but not man-aged purposefully for genetic health. Analysis of the factors eliminating the species’s original island populations indicated loss through habitat change and mammalian predation within perhaps no more than two to three decades of human contact. Loss was undoubtedly hastened by intrinsic extinction-proneness (low reproductive output, forest-interior preference, sensitivity to habitat quality, intolerance of open spaces), greater in black robins than in other Petroica species in New Zealand and Australia. Low productivity in ancestral survivors prevented spontaneous recovery after translocation to better habitat, requiring intensively assisted breeding to raise population sizes quickly. Strong growth in response to management did not suggest a moribund species but post-management growth rates were significantly slower. Analysis of demographic data from 1980-81 to 2001-02 showed breeding effort to be comparable in all populations but breeding failures of varying severity at different life-history stages affected natural recovery in each. The worst affected suf-fered from briefer adult life-spans and more severe failure late in the breeding cycle. Generally colder sub-canopy air temperatures suggested adverse environmental pressure on this population. Extant populations are highly inbred and severely depleted genetically fol-lowing successive bottlenecks (prolonged historically) and recovery from a single-pair event between 1979-80 and 1982-83. At current small sizes, accumulating levels of relat-edness are anomalously high. Contrary to perceptions that black robin recovery indicates normal fitness, viability was shown to be threatened by inbreeding depression expressed cumulatively in the breeding cycle as low juvenile production and survival, thus aggravating a principal vulnerability. Declines did not result during the study period and effects were limited to reproductive fitness costs in black robin fathers only. There was no evi-dence of harmful effects on survival probability or from maternal inbreeding but an emerging kinship effect on yearling production is possible. The restricted costs in such highly inbred populations (relative to less inbred congeners) suggest historical purging of genetic load. Work is needed to clarify this and the extent to which inbreeding could reduce re-sponses to new extinction pressures. Recent population declines may elevate levels of in-breeding further. Genetic management is strongly recommended, including habitat expan-sion to maximise population sizes and prevent further cross-breeding with a sympatric congener. Management and monitoring recommendations apply valuable lessons derived from a critique of past practices. In summary, findings give qualified support to the hy-pothesis of reduced susceptibility to inbreeding depression in threatened insular endemic birds. Findings do not support the assumption implicit in black robin management historically and in New Zealand wildlife management practice generally that stochastic genetic threats are unimportant risks to viability.

This is the first study to test the possibility that autonomous selfing in carnivorous plants functions as a mechanism to reduce the pollinator-prey conflict by reducing pollen-limitation. Carnivorous plants can rely on insects both for pollen supply and nutrition and past studies suggest that pollinator-prey conflict will occur when flowers and traps compete for this shared resource, resulting in either a pollen or prey-limitation. The extent of either limitation will initially be governed by insect availability in a given environment. Conflict is most likely to occur in resource and insect-limited environments. When pollinators are trapped as prey (pollinator-prey overlap), pollen-limitations can be exacerbated. I additionally propose that conflict can occur when pollinating prey visit flowers instead of traps exacerbating prey-limitations. Trade-offs in the selection on traits that attract insects as pollinators and prey may occur when plant fitness is limited by both pollen and prey. To reduce pollen-limitation, traits that reduce pollinator dependency should be selected and, in the event of pollinator-prey overlap, the selection of trap-flower separation may provide protection of pollinators. I quantify pollen-limitation and pollinator-prey overlap for two New Zealand Drosera species, Drosera arcturi and D. spatulata to explore the influence of various trap morphologies on prey types captured and pollinator-prey overlap. Only D. arcturi caught insects from pollinating families (96% of pollinators were from families also caught as prey) although both Drosera species were outcross pollen-limited. I examined the role self-fertilisation to reduce pollen-limitation in these and three other Drosera species (D. arcturi, D. spatulata and D. stenopetala in alpine habitats and D. auriculata and D. peltata in lowland habitats) including the costs of inbreeding depression in the three alpine Drosera species. I found that all Drosera species examined were fully self-compatible and that selfing reduced pollinator-prey conflict (mediated by pollen-limitation) by reducing pollinator dependency. The three alpine Drosera species autonomously selfed when outcrossing was unavailable providing reproductive assurance when pollinators are scarce. Furthermore, measures of cumulative lifetime inbreeding depression found no inbreeding depression for any species. To relieve outcross pollen-limitations, traits that increase floral attraction may also be selected. I therefore test whether the pollinator attraction trait of flower height and the pollinator protection trait of trap-flower separation increase outcrossing success for two Drosera species. Within species patterns were not detected, but the species that exhibited taller flowers further from its traps, D. spatulata, experienced greater outcrossing success than D. arcturi with shorter flowers closer to its traps. Since pollinator dependency is low for both species, I argue that the increased overlap and decreased separation in D. arcturi indicates the importance of capturing pollinators as prey, whereas the lack of overlap and increased trap-flower separation in D. spatulata suggests that there are some benefits to outcrossing that I did not detect. To reduce prey limitations, traits that increase trap attraction/effectiveness should be selected. It is likely, however, that both pollen and prey limitations can occur simultaneously, in which case traits that relieve both limitations may be selected (e.g. floral attraction of pollinating prey). I test whether resource allocation to flowers alleviates both pollen and prey limitations and found that flowering indeed increases per plant prey capture for Drosera arcturi. Though selfing reduces pollen-limitations, continuing to invest in floral attraction may also benefit prey capture, relieving pollinator-prey conflict mediated by both pollen and prey-limitations. This study is the first to explore the role of floral traits in relieving prey-limitations in addition to demonstrating that selfing in carnivorous plants not only reduces pollinator dependency, but also influences the consequences of pollinator-prey overlap. This highlights the necessity of examining breeding systems in all future studies of pollinator-prey conflict.

Mayapple (Podophyllum peltatum L.) is a common clonal understory herb in temperate eastern North America. Its fecundity is pollen-limited because its flowers are nectarless, and native pollinators do not collect its pollen. I conducted field studies in southeastern Michigan to determine mayapple’s compatibility system and whether neighboring plants facilitated its pollination. I tested for facilitation by correlating the degree of pollen limitation with the abundances of neighbors and measuring whether the removal of neighboring flowers increased pollen limitation of fecundity. Mayapple populations in four sites were self-incompatible (SI), but all clones in one site were self-compatible (SC). This difference could reflect genetic differences or possibly differences in inbreeding depression due to resources. The site with SC had the highest light availability and outcross fruit set, suggesting that abortion of inbred ovules may be lower under high resources, resulting in expression of SC. Visitation to mayapple flowers was consistently low (0.03-0.06 visits/flower/hour), and fruit set was pollen-limited (pollen supplementation increased fruit set 3 – 18-fold) in all three years of this study. Based on regressions of pollen limitation violets facilitated fruit set in 2005 and 2007 but reduced seed set in 2005. Garlic mustard and spring beauty reduced fruit set in 2005. Except for violets in 2007, co-flowering species did not affect pollen limitation in 2006 or 2007. Floral removal did not change pollination success in 2006, confirming that neighbors neither facilitated nor competed with mayapple for pollination in that year. Neighboring plants could also reduce mayapple fecundity through interspecific pollen transfer (IPT). The addition of Phlox divaricata pollen did depress fruit set, but Geranium maculatum pollen did not. However, foreign pollen was rare on mayapple stigmas suggesting that IPT is unlikely to be important in the field. Mayapple could also facilitate or compete with neighboring plants for pollination. However, correlations showed no effect of mayapple on the pollination success of wild geranium, and hand-pollination with mayapple pollen did not significantly depress fruit or seed set. In a review of studies on pollination facilitation, I propose that future studies employ similar methods and measure effect sizes for comparisons and meta-analyses.

Inbreeding can have profound negative effects on individuals by reducing fertility and viability. In populations, inbreeding depression can reduce growth rates and increases extinction risk. The aims of this thesis are to investigate inbreeding depression in male guppies (Poecilia reticulata) and to study the evolution of mechanisms for inbreeding avoidance in females, using guppies from a feral population in Queensland, Australia. Male guppies are highly polymorphic in their sexual ornamentation, indeed they show one of the most extreme polymorphisms observed in nature. Female guppies exhibit complex mate choice based on preferences for ornamentation, as well as social context. I aim is to examine how these factors of inbreeding avoidance alter sexual selection. In male guppies I found strong inbreeding depression in male sperm numbers, which is amplified under semi-natural compared to laboratory conditions (Chapter 2). Moreover, inbreeding depression results in low fertility under sperm competition: an experiment using artificial insemination techniques reveals that highly inbred males are heavily disadvantaged in gaining paternity (Chapter 3). On population level, inbreeding depression is manifest in reduced growth rates, predominantly in the early stages of inbreeding (Chapter 4). Population growth at inbreeding coefficients f=0.375-0.59 did not seem to lead to inbreeding depression, whereas lower levels of inbreeding reduced population growth. Although the growth rates in inbred populations appear normal, severe inbreeding depression is uncovered after outbred immigrants are added. Specifically, male immigrants are most efficient in short-term genetic rescue, probably due to insemination of large numbers of females whereas females are limited in the number of eggs they can produce (Chapter 4). Male ornamental traits show significant inbreeding depression in semi-natural conditions only (Chapters 2 & 3). Inbreeding avoidance mechanisms seem to have evolved in females: they prefer courtship displays of non-inbred males (Chapter 2), unfamiliar males (Chapter 5) and males with rare patterns (Chapter 6). This preference might increase the mating success of immigrants, and may have evolved to facilitate the avoidance of inbreeding. Together with context-independent preferences for ornament combinations (Chapter 6), it also offers an explanation for the maintenance of polymorphism in this species.

Biological invasions are now the second leading cause of loss of biodiversity. Recently, hybridization has been hypothesized as a mechanism for invasion success. The combination of individuals from different gene pools may create novel genotypes having increased invisibility. The goal of my research was to examine the role that intraspecific hybridization may play in invasion success by using the agricultural weed Silene latifolia as a model. I used a common garden experiment to examine whether crossing parents of different ancestry results in offspring with differing quality. Plants were pollinated in three treatments reflecting parent plants within the same population (P), between-populations, within-continent (R), and between continent (C). Plants which had been produced by mating plants located intermediate distances apart (R) germinated faster, and had a greater probability of germinating and surviving. This thesis contributes to our knowledge of the potential role hybridization may play in a successful invasion.

Dispersal is a critical life-history component; it determines gene flow and has profound effects on population structure, demography, social systems, and population viability. To add to our knowledge of dispersal and, in particular, our understanding of the relationship between dispersal and inbreeding, I studied three aspects of the biology of the red-cockaded woodpecker: dispersal of breeding females; the costs, benefits, and frequency of inbreeding; and the effect of inbreeding on natal dispersal. Dispersal of breeding female red-cockaded woodpeckers is strongly associated with inbreeding avoidance and mate choice, weakly associated with site choice, and not found to be associated with social constraints. Estimates of mortality for non-dispersing and dispersing breeding females were 24 and 59 percent per year, respectively-rare evidence of the cost of breeding dispersal. Significant costs of close inbreeding were found. Closely related pairs (kinship coefficient greater than 0.1) had lower hatching success as well as lower survival and recruitment of fledglings than unrelated pairs. Moderately related pairs (kinship coefficient between 0 and 0.1) and moderately inbred individuals had increased hatching success, but did not produce more young. Despite documented costs of close inbreeding and a predictable spatial distribution of closely related males near the natal territory, female fledglings disperse a median of only two territories and a modal distance of one territory. Natal dispersal of females is affected by closely related males on the natal site but unaffected by closely related males or moderately related males that are off the natal site.