In 1903 Karl Pearson identified the effects of censorship, or range restriction, on the correlation coefficient. The current report reviews the history and literature examining those effects, corrections for range restriction, and the limitations of previous research. A rationale for further research of the effect of range restriction on logistic regression parameter estimates is presented. / text

This dissertation examines a profit-maximizing project selection and scheduling problem. Assume that a set of potentially profitable projects are available, yet limited available resources may not allow all of them to be pursued. Profit profiles for projects are assumed to be non-increasing functions of project completion times, i.e. profit returns are sensitive to time-to-market. Decision needs to be made on which sub-set of projects should be chosen and how resources should be allocated to these projects such that the total profit is maximized.Formal mathematical models are formulated for various versions of the problem, including such ones incorporating a third team formation aspect. Structure of the problem is examined and insights are gained regarding prioritization of project, specifically. Although prioritization is sub-optimal in general, heuristic solution methods based on prioritization are pursued, since the scheduling sub-problem itself is NP-hard.A decomposition heuristic framework is first proposed to obtain good solutions using minimum computational time. Sets of test instances are generated using project network data from well-known source in the literature. Computational runs reveal that three priority rules achieve significantly better profits than the benchmarking random priority rule.Improving upon the prioritization based decomposition heuristic, an implicit enumeration is proposed. This algorithm does not examine all priority sequences, yet guarantees an optimal priority sequence when the computation is completed. Several fathoming rules are proposed to cut back computational time effectively. Comparison to the profits achieved by the best priority rule and the benchmarking random priority rule shows a significant improvement on profits, yet at a cost of reasonable added computation time.Future research areas include identifying general conditions under which prioritization of projects would lead us to an optimal solution. Developing better upper bounds for the implicit enumeration scheme is also of interest. The team formation aspect has yet to be treated computationally. It would also be of interest to consider how synergy deviation information may be fed back to the earlier stages of project selection and scheduling decision. Trade-off between profit and team synergy may also be considered in the future.

In der vorliegenden Arbeit werden drei Modellrechnungen vorgestellt, die auf dem Selektionsindex basieren und mit denen der Nutzen genomischer Zusatzinformation auf aktuelle Selektionsstrategien für Reitpferde und Schweine ermittelt werden soll. In Kapitel eins werden sowohl die klassischen Instrumente der Tierzucht, der Selektionsindex, die ökonomische Modellierung und die Genfluss-Methode, als auch die genomische Selektion dargestellt. Zudem wird die Software ZPlan+, in welcher diese Instrumente miteinander kombiniert sind, kurz vorgestellt. Im zweiten Kapitel der Arbeit wird das Potenzial der genomischen Selektion für die Reitpferdezucht dargestellt. Der Zugewinn an Genauigkeit eines Zuchtwerts durch Hinzunahme genomischer Information wurde für drei praxisnahe Selektionsschritte betrachtet, nämlich für Pferde ohne Eigenleistung, Pferde mit Eigenleistung und Pferde mit Eigen- und Nachkommenleistungen. Die Modellrechnung zeigte, dass sich durch Hinzunahme genomischer Information besonders die Genauigkeit von Zuchtwerten junger Pferde ohne Eigen- oder Nachkommenleistung beträchtlich steigern ließe. Sobald Nachkommenleistungen unterstellt wurden war der Zugewinn an Genauigkeit durch genomische Information jedoch zu vernachlässigen. Für die praktische Pferdezucht ist die genomische Selektion daher vor allem für die Absicherung vorläufiger Zuchtzulassungen für junge Hengste nach der Körung und für leistungsgeprüfte Hengste ohne Nachkommenleistungen empfehlenswert. Im dritten Kapitel der Arbeit wurde untersucht, wie sich die Art der berücksichtigten Informationsquellen (konventionell oder genomisch) auf die Interaktion der Parameter genetische Korrelation, Heritabilität und ökonomische Gewichtung auswirkte. Als Vergleichsparameter wurde die Standardabweichung des jeweiligen Indexes herangezogen, welche sich direkt proportional zum Zuchtfortschritt verhält. Es wurden drei Indices mit zwei Zuchtzielmerkmalen verglichen, deren Informationsquellen in Anlehnung an die Schweinezucht gewählt wurden. Der erste Index wurde für ein Tier mit Eigenleistung aufgestellt, während im zweiten Index zusätzliche Vollgeschwisterleistungen angenommen wurden. Im dritten Index wurde die Eigenleistung mit genomischer Information für beide Zuchtzielmerkmale kombiniert. Die Genauigkeit der genomischen Zuchtwerte wurde durch Annahme unterschiedlich großer Referenzpopulationen variiert. Je mehr Information im Index berücksichtigt wurde, desto unabhängiger wurde die Standardabweichung des Indexes von den Parametern genetische Korrelation, Heritabilität und ökonomische Gewichtung. Dabei wurden für den Vollgeschwisterindex mit 7 Vollgeschwistern und den genomischen Indexes mit einer Referenzpopulation von 1.000 Tieren vergleichbare Ergebnisse gefunden. Die Anzahl von 1.000 Tieren in der Referenzpopulation kann somit als Mindestmaß für die Schweinezucht angenommen werden. Für geringere Heritabilitäten zeigte sich eine deutliche Überlegenheit des genomischen Indexes über den Vollgeschwisterindex. Im vierten Kapitel der Arbeit wurde ein Schweinezuchtprogramm optimiert. Dazu wurden verschiedene Zuchtziele und Selektionsstrategien gegen Ebergeruch verglichen. Das Zuchtziel war entweder die Selektion gegen den Gehalt der chemischen Leitkomponenten des Ebergeruchs, Androstenon, Skatol und Indol oder die Selektion gegen den, von Testpersonen bestimmten, Human Nose Score. Innerhalb der Selektion gegen die chemischen Komponenten wurden drei verschiedene Informationsquellen miteinander verglichen, nämlich eine Feldprüfung in Form einer Biopsie am lebenden Eber, genomische Selektion sowie die Kombination beider Informationen. Innerhalb der Selektion gegen den Human Nose Score wurden als Informationsquellen eine Stationsprüfung von Voll- und Halbgeschwistern des Selektionskandidaten sowie genomische Selektion miteinander verglichen. Das komplexe Zuchtprogramm wurde deterministisch mit der Software ZPlan+ modelliert. Bei der Selektion gegen die chemischen Komponenten des Ebergeruchs war der Zuchtfortschritt am höchsten, wenn als Informationsquelle die Eigenleistung in Form einer Biopsie der männlichen Selektionskandidaten genutzt wurde. Aufgrund der hohen Erblichkeit der Ebergeruchskomponenten war der Nutzen genomischer Zuchtwerte deutlich geringer und gleichzeitig teurer. Für die Selektion gegen den geringer erblichen Human Nose Score lieferte die genomische Selektion einen höheren Zuchtfortschritt als die Stationsprüfung der Geschwister des Selektions-kandidaten. Auch wenn der Human Nose Score als Zielmerkmal angesehen wurde, erwies sich eine Selektion gegen die chemischen Komponenten als zielführend, da der (korrelierte) naturale Zuchtfortschritt des Human Nose Scores bei Durchführung einer Biopsie deutlich höher war, als bei direkter Selektion gegen den Human Nose Score. Im fünften Kapitel wird das Potential der genomischen Selektion für die Reitpferde- und Schweinezucht diskutiert. Dabei wird besonders auf Möglichkeiten für die Vergrößerung der Referenzpopulation und der Kostenreduktion eingegangen. Für die Reitpferdezucht wer-den zusätzlich Möglichkeiten aufgezeigt, wie der Zuchtfortschritt mit konventionellen Methoden gesteigert werden könnte.

Safety first criterion and mean-shortfall criterion both explore cases of assets allocation with downside risk. In this paper, I compare safety first portfolio selection problem and mean-shortfall portfolio optimization problem, considering risk averse investors in practice. Safety first portfolio selection uses Value at Risk (VaR) as a risk measure, and mean-shortfall portfolio optimization uses expected shortfall as a risk measure, respectively. VaR is estimated by implementing extreme theory using a semi-parametric method. Expected shortfall is estimated by two nonparametric methods: a natural estimation and a kernel-weighted estimation. I use daily data on three international stock indices, ranging from January 1986 to February 2012, to provide empirical evidence in asset allocations and illustrate the performances of safety first and mean-shortfall with their risk measures. Also, the historical data has been divided in two ways. One is truncated at year 1998 and explored the performance during tech boom and financial crisis. the mean-shortfall portfolio optimization with the kernel-weighted method performed better than the safety first criterion, while the safety first criterion was better than the mean-shortfall portfolio optimization with the natural estimation method.

Understanding the functional relationship between characters and components of fitness is a central goal of evolutionary biology. The studies in this thesis examined the ecological causes and evolutionary consequences underlying differences in fitness among individuals of Lobelia cardinalis. Flowering plants experience selection from many sources, which may enhance or oppose selection by pollinators. In the second chapter of this thesis, the role of pollinators and herbivores in shaping selection on floral characters was investigated. Floral traits experienced pollinator-mediated selection and weak selection by weevil larvae and slugs. Because pollinators also forage according to local density of flowers, in the fourth chapter I explored how local density of individual plants and flowers influences fitness of individual plants. Plants at dense sites produced more seeds, consistent with pollinator preference for denser patches. Individual female-phase flowers produced more seeds as the density of surrounding male-phase flowers increased and female-phase flowers decreased. This study highlights how plant phenotype and local density influence pollination and subsequent plant fitness. In L. cardinalis rosette formation (a life-history character) partly shapes the distribution of plants, and may influence plant survival and fitness. In the fifth chapter, I explored how variation in allocation to clonal reproduction among plants (ramets) and genets influenced survival and fitness. Plants that produced more and larger rosettes realized higher survival independent of the phenotype of the parental. Plants that produced one rosette in 2009 produced more seeds in 2010 than plants that produced more than one rosette. This pattern was reversed in the following time period; plants that produced more rosettes in 2010 produced more seeds in 2011. The relative importance of pollinators versus other selective agents in shaping floral traits, as well as the intensity of competition among individual plants and flowers likely depend on the extent to which reproduction is pollen limited. In the third chapter, I explored how pollen limitation affected selection on floral traits via female fitness and found a weak relationship. Although this seems to contradict intuition, several reasons may limit the influence of pollen limitation on selection.

Co-occurring species of flowering plants may compete for pollination, which can cause character displacement by altering natural selection on floral traits. In a gynodioecious species, competition for pollination may also affect the evolution of sexual dimorphism of floral traits by influencing sex-specific selection. I demonstrated that Mimulus ringens did not affect seed set of gynodioecious Lobelia siphilitica, indicating that it is not a competitor for pollination. The presence of M. ringens did not alter selection upon most floral traits of L. siphilitica. I detected sex-specific selection upon five floral traits, supporting the hypothesis that sexual dimorphism evolves in response to sex-specific selection, rather than pleiotropic effects. My results also suggest that the presence of a co-flowering species may provide a context for sex-specific selection, and therefore, influence sexual dimorphism. Thus, my results suggest a link between the fields of study of competition for pollination and the evolution of sexual dimorphism. / NSERC, OGS, Ontario Innovation Trust, Canada Fund for Innovation

Mountain caribou, an arboreal lichen-feeding ecotype of woodland caribou (Rangifer tarandus caribou), have been extirpated from much of their historic range. Mountain caribou are federally listed as Endangered by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) and red-listed by the BC government. Habitat loss and fragmentation of old growth forest is the ultimate cause for population declines. Yet, predation, linked to apparent competition, is the proximate cause for high rates of mortality. One of the most imperiled populations resides in the Purcell Mountains of BC, which was experimentally augmented in 2012 with 19 northern caribou from northern BC. The caribou-predator literature predominantly focuses on the relationship between caribou and wolves (Canis lupus) in northern caribou populations. However, cougars (Puma concolor) have been identified as a major predator of Purcells-South (PS) caribou, yet caribou-cougar interactions remain largely unstudied. I evaluated cougar predation risk in space and time on resident and translocated caribou in the Purcell Mountains. To do so, I determined biologically relevant seasons for resident, donor (i.e. not translocated) and translocated caribou, and cougars. I then used these seasons to investigate seasonal patterns of movement and habitat use between the three groups of caribou and cougars. Next I used resource selection functions (RSFs) to estimate habitat based seasonal variation in predation risk. I used these RSFs to compare the seasonal habitat selection and risk to cougar predation between resident and translocated caribou. Five resident caribou seasons and two cougar seasons were defined. Translocated caribou displayed inconsistent movement behavior with no clear seasonal pattern. Resident caribou remained at high elevations year-round and selected for low risk cougar habitat during the calving season at the home range scale and year-round at the landscape scale. Translocated caribou displayed risky behaviour throughout the study period by traveling to mid to low elevations and habitats selected by cougars. Translocated caribou displayed the same general pattern of elevational movement as their northern conspecifics, spending the majority of their time at lower elevations than resident caribou. Of the 19 translocated caribou, 17 (89%) died during the study, six of which were preyed upon by cougars, two by wolves, and the remainder due to accidents or unknown causes. In summary, translocated caribou did not adopt the predator avoidance or habitat selection strategies of resident caribou. I recommend that future efforts to augment small caribou populations use donor caribou experienced with similar predators and that possess comparable seasonal habitat use to the recipient population. However, because most suitable donor populations are declining, a soft-release of captive-reared mountain caribou might be the best option for mountain caribou recovery efforts. / Graduate / 0366 / 0329

While there has long been interest in the form of sexual selection in males, studies characterizing this selection in females remain sparse. Sexual selection on females is predicted for sex-role reversed Mormon crickets, where males are choosy of mates and nutrient-deprived females compete for matings to gain nutritious nuptial gifts. I used selection analyses to describe the strength and form of sexual selection on female morphology. There was no positive sexual selection on the female body size traits predicted to be associated with male preferences and female competition. Instead, I detected selection for decreasing head width and mandible length. Additionally, I tested the validity of a commonly-used instantaneous measure of mating success (mated vs. unmated) by comparing selection results with those determined using a more detailed fitness measure (cumulative mating rate). The two fitness measures yielded similar patterns of selection, supporting the common sampling method comparing mated and unmated fractions.

While there has long been interest in the form of sexual selection in males, studies characterizing this selection in females remain sparse. Sexual selection on females is predicted for sex-role reversed Mormon crickets, where males are choosy of mates and nutrient-deprived females compete for matings to gain nutritious nuptial gifts. I used selection analyses to describe the strength and form of sexual selection on female morphology. There was no positive sexual selection on the female body size traits predicted to be associated with male preferences and female competition. Instead, I detected selection for decreasing head width and mandible length. Additionally, I tested the validity of a commonly-used instantaneous measure of mating success (mated vs. unmated) by comparing selection results with those determined using a more detailed fitness measure (cumulative mating rate). The two fitness measures yielded similar patterns of selection, supporting the common sampling method comparing mated and unmated fractions.

Directional selection is prevalent in nature yet phenotypes tend to remain relatively constant, suggesting a limit to trait evolution. The genetic basis of evolutionary limits in unmanipulated populations, however, is generally not known. Given widespread pleiotropy, opposing selection on a focal trait may arise from the effects of the underlying alleles on other fitness components, generating net stabilizing selection on trait genetic variance and thus limiting evolution. Here, I look for the signature of stabilizing selection for a suite of cuticular hydrocarbons (CHCs) in Drosophila serrata. Despite strong directional sexual selection on CHCs, genetic variance differed between high and low fitness individuals and was greater among the low fitness males for seven of eight CHCs. Univariate tests of a difference in genetic variance were non-significant but have low power. My results implicate stabilizing selection, arising through pleiotropy, in generating a genetic limit to the evolution of CHCs in this species.