Roads cause substantial wildlife mortality, but there is currently limited understanding of the relative magnitude of this mortality source. There are also substantial gaps in knowledge concerning the ecological ramifications of carrion introduced to the environment from vehicle collisions and in particular how vertebrate scavengers may consume carrion resulting from vehicle collisions. Although a variety of factors influence scavenger use of carcasses, the mechanisms influencing competition for this resource between obligate and facultative scavengers have not been thoroughly explored. I conducted a global synthesis of mortality of terrestrial vertebrates documenting 42,755 mortalities of known cause from 120,657 individuals representing 305 vertebrate species. Overall, 28% of mortalities were directly caused by humans and 72% were from natural sources. Vehicle collisions accounted for 4% of mortality overall. Larger birds were more likely than smaller birds to die from vehicle collisions and vehicle mortality of mammals increased over time. There was no difference in proportion of rabbit carcasses scavenged or scavenger arrival time between those placed along roads, power line clearings, and forests. No species arrived at roads quicker than other treatments. Turkey vultures (Cathartes aura) and coyotes (Canis latrans) scavenged equally across treatments, whereas gray foxes (Urocyon cinereoargenteus) scavenged along roads and power lines, but not in forests. Scavenger use of carrion near roads likely relates to factors besides carrion availability, such as traffic avoidance and predation risk. Because some scavengers make substantial use of carrion on roads, this resource could be an important mechanism by which human activities impact wildlife. Scavenging by facultative scavengers did not increase in the absence of competition with vultures. I found no difference in scavenger presence between control carcasses and those from which vultures were excluded. Facultative scavengers did not functionally replace vultures during summer in this study. These results suggest that under the conditions of this study, facultative scavengers would not compensate for loss of vultures. Carcasses would persist longer in the environment and consumption of carrion would likely shift from vertebrates to decomposers. Such changes could have substantial implications for disease transmission, nutrient cycling, and ecosystem functioning.

In the New Zealand context, considerable academic and government attention has been given to the socioeconomic and ethnic disparities in health, and how they have evolved over time. Despite evidence of clear regional health patterning within New Zealand, there has been very little research monitoring how the geographical trends in health have evolved over time. The period 1980 to 2001 is very important in New Zealand's contemporary history, as it was a time of rapid social and economic change. For this reason, researchers are motivated to examine the extent to which health differentials evolved during the same period. The reduction of health inequalities are at the top of the Government's health agenda, it is therefore important not only to monitor the success, or otherwise, of a reduction in social inequalities, but also, geographic inequalities. This thesis examines the extent of geographic inequalities in mortality in contemporary New Zealand, and whether or not mortality has become geographically polarised between 1981 and 2000. This thesis builds on research carried out in New Zealand, and seeks to delve deeper into the specifics of the geographic variation of mortality. Importantly, it fills several knowledge gaps during this period, which include: the geographic inequality of cause-specific mortality, the difference in regional cause-specific mortality between males and females, and the inequalities of mortality at a finer geographic resolution. A significant debate revolves around the relative contribution of compositional and contextual explanations for the geographic variation of health outcomes. The research undertaken in this thesis examines the contribution of population change and deprivation to the geographic inequalities of mortality. Numerous key findings were identified in this research, four of which are as follows: In 2000, significant geographic inequalities in cause-specific mortality existed within New Zealand; between 1981 and 2000 the geographic mortality gap remained relatively stable; and when the geographic areas are sorted by deprivation, the results indicate that there has been a widening of the mortality gap. Analysis of the relationship between population change and mortality provide cautious support for the finding that mortality rates are higher in areas that have experienced population decline and conversely, that mortality rates are lower in areas where there has been a growth in population. The high and stable levels of geographic inequality should be of great concern to policy makers as the results of this research indicate that policies addressing health inequalities in New Zealand are not sufficiently potent.

In the New Zealand context, considerable academic and government attention has been given to the socioeconomic and ethnic disparities in health, and how they have evolved over time. Despite evidence of clear regional health patterning within New Zealand, there has been very little research monitoring how the geographical trends in health have evolved over time. The period 1980 to 2001 is very important in New Zealand's contemporary history, as it was a time of rapid social and economic change. For this reason, researchers are motivated to examine the extent to which health differentials evolved during the same period. The reduction of health inequalities are at the top of the Government's health agenda, it is therefore important not only to monitor the success, or otherwise, of a reduction in social inequalities, but also, geographic inequalities. This thesis examines the extent of geographic inequalities in mortality in contemporary New Zealand, and whether or not mortality has become geographically polarised between 1981 and 2000. This thesis builds on research carried out in New Zealand, and seeks to delve deeper into the specifics of the geographic variation of mortality. Importantly, it fills several knowledge gaps during this period, which include: the geographic inequality of cause-specific mortality, the difference in regional cause-specific mortality between males and females, and the inequalities of mortality at a finer geographic resolution. A significant debate revolves around the relative contribution of compositional and contextual explanations for the geographic variation of health outcomes. The research undertaken in this thesis examines the contribution of population change and deprivation to the geographic inequalities of mortality. Numerous key findings were identified in this research, four of which are as follows: In 2000, significant geographic inequalities in cause-specific mortality existed within New Zealand; between 1981 and 2000 the geographic mortality gap remained relatively stable; and when the geographic areas are sorted by deprivation, the results indicate that there has been a widening of the mortality gap. Analysis of the relationship between population change and mortality provide cautious support for the finding that mortality rates are higher in areas that have experienced population decline and conversely, that mortality rates are lower in areas where there has been a growth in population. The high and stable levels of geographic inequality should be of great concern to policy makers as the results of this research indicate that policies addressing health inequalities in New Zealand are not sufficiently potent.

Cause-specific mortality (CSM), among other global health estimates, has garnered prominence in the contemporary public health field. CSM has been associated with several factors, however, research comparing CSM for prefracking versus postfracking periods is sparse. Hydraulic fracturing or fracking is a technique of extracting oil and gas from deep underground. The purpose of this study was to evaluate the difference among mean CSM scores from 1975 through 2015 in the available cities and counties of residence in Colorado and to determine the impact of gender, marital status, county of residence, and city of residence on CSM scores (prefracking period 1975-1977 versus postfracking period 1999-2015) among adults aged 45-70 years. In this retrospective quantitative study, the socioecological model of health was used to analyze 73,251 cases obtained from the Colorado Department of Public Health and Environment. One-way analysis of variance and multiple regression were used to analyze data. Results showed that Denver County had a higher mean CSM score compared to other counties in Colorado. Regression results revealed a significant but weak association between CSM scores and gender, marital status, city of residence, and county of residence. If gender, marital status, and county of residence can be significant predictors of CSM, this information could have social change implications by influencing decisions regarding CSM and fracking.

Maintaining desired numbers of wildlife species requires an understanding of species-specific population dynamics. For ungulate species such as the white-tailed deer (Odocoileus virginianus), understanding the neonatal survival rate of a population and factors that influence that survival rate, may be two of the most important factors to successful deer management. We examined neonatal survival in an eastern Kentucky population of deer living in relatively low densities (/ km2), with adequate habitat and supposedly poor population growth. Neonates (102) were captured in the summer birth periods of 2014 - 2016 and radio-monitored until the beginning of the fall archery deer season. We found moderate-to-low survival estimates to four months of 43% (95% CI: 29 – 57%) that are consistent with many areas in the Midwest and southeastern United States. Predation, including suspected predation events, from bobcats (Lynx rufus) and coyotes (Canis latrans) accounted for 80% of all neonate mortalities. A thorough examination of the survival and mortality in the neonate component of this population is discussed herein.

L’analyse des disparités géographiques de mortalité en population générale est essentielle pour la planification sanitaire. Elle a pour objectif d’identifier les facteurs impliqués dans la genèse de ces différences. Ceci contribue à orienter les décisions portant sur les populations à cibler et l’allocation de ressources. L'objectif de ce travail est de participer à expliquer les variations géographiques de la mortalité par cause de décès en population générale.En première partie, une estimation des disparités géographiques de mortalité par cause calculée depuis la base nationale exhaustive des causes médicales de décès a été réalisée à l'aide d'une mesure synthétique. Sur la base de ces constatations, une analyse écologique de l’évolution de la distribution géographique de la mortalité a été proposée. Faisant l’hypothèse que les changements démographiques sont un marqueur des changements socio-sanitaires, une typologie des changements démographiques ainsi qu’un modèle de Poisson ont aidé à investiguer cette association avec l'évolution de la mortalité. De manière générale les disparités géographiques de mortalité ont augmenté sensiblement à partir de 1990. La contribution de l’échelle régionale aux disparités spatiales qui était très élevée en 1975, va évoluer de manière différenciée selon le sexe et la cause de décès. L'évolution de la mortalité était par ailleurs associée négativement à l'accroissement de la population. Modulé par le désavantage social et le degré d’urbanisation, ce lien était plus fort pour les maladies alcolo-tabagiques et les morts violentes.La deuxième partie était consacrée à l’investigation des facteurs de risques individuels et contextuels à l’aide d’un modèle multiniveaux. Dans un premier temps, la problématique d’une telle analyse a été élargie à la fiabilité de l’analyse multiniveaux des évènements rares de santé. Evaluée à l’aide de simulations, les recommandations sur la faisabilité d’une analyse en population générale ont été appliquées dans un second temps sur l’Echantillon Démographique Permanent (EDP). Nous avons ainsi tenté d'expliquer les disparités géographiques de mortalité par la composition socio-démographique, l’environnement physique étant caractérisé par l’accessibilité géographique aux soins et l’environnement social étant mesuré par un score de désavantage social. L'étude de faisabilité de l'analyse multiniveaux a mis en évidence la très forte sensibilité de la fiabilité des estimations au choix de l'échelle géographique et à l'incidence de l'évènement de santé étudié. Il existe une association contextuelle en population générale avec le désavantage social du lieu de résidence. Cette association est systématique et régulière pour toutes les causes de décès investiguées. Seules les disparités géographiques de mortalité par maladies circulatoires étaient en grande partie expliquées par les facteurs individuels et contextuels. L'accessibilité géographique aux soins n’était pas retrouvée associée à la mortalité.Ce travail met l'accent sur la multitude des facteurs en jeu et la nécessaire complémentarité des approches pour appréhender les disparités géographiques de mortalité. La rareté de l’évènement étudié dans l'EDP et l’absence d’information socio-économique individuelle fiable dans la base exhaustive des causes de décès sont deux contraintes majeures de ce type d'analyse qui nécessitent le déploiement de méthodologies appropriées. Néanmoins, les perspectives d’enrichissement des données par des bases administratives et médico-administratives laissent présager des retombées importantes. Malgré l'incertitude résiduelle non négligeable sur certain résultats, l'étude met en évidence avec des éléments plus précis l'importance simultanée de facteurs individuels et contextuels en population générale. / The analysis of geographical disparities in mortality in the general population is essential for health planning. The aim is to identify the factors involved in the genesis of these differences. This helps to guide decisions about the populations to be targeted and the allocation of resources. The objective of this work was to help explain geographic variations in mortality by cause of death in the general population.In the first part, an estimate is made of the geographical disparities in mortality by cause calculated from the comprehensive national database of medical causes of death using a synthetic measure. On the basis of these findings, an ecological analysis of the evolution of the geographical distribution of mortality is proposed. Assuming that demographic changes are a marker of health changes in society, a typology of demographic changes and a Poisson model were used to investigate this association with mortality trends.In general, geographical disparities in mortality have increased markedly since 1990. The contribution of the regional scale to spatial disparities, which was very high in 1975, has evolved differently according to sex and cause of death. The evolution of mortality is also negatively associated with the increase in population. Depending on the degree of social disadvantage and degree of urbanization, this link is stronger for alcohol-tobacco and violent deaths.The second part is devoted to investigating individual and contextual risk factors using a multilevel model. Initially, the problem of such an analysis was extended to the reliability of multilevel analysis of rare health events. Evaluated by using simulations, the recommendations on the feasibility of a general population analysis were then applied to the Permanent Demographic Sample (PDS). We thus attempted to explain geographical disparities in mortality by socio-demographic composition, the physical environment being characterized by geographical accessibility to care and the social environment being measured by a social disadvantage score.The multilevel feasibility analysis revealed the very high sensitivity of the reliability of the estimates to the geographical scale and the impact of health events studied. In the general population, there was a contextual association with the social disadvantage of the place of residence. This association was systematic and regular for all causes of death investigated. Only geographical disparities in mortality from circulatory diseases were largely explained by individual and contextual factors. Geographic access to care was not found to be associated with mortality.This work emphasizes the multitude of factors involved and the need to combine approaches when dealing with geographical disparities in mortality. The scarcity of events studied in the PDS and the lack of reliable individual socio-economic information in the exhaustive database of causes of death are two major constraints of this type of analysis, in which the appropriate methodologies should be used. Nevertheless, the prospect of obtaining richer data thanks to the use of administrative and medico-administrative bases suggests that significant benefits may be obtained from this approach. Despite the considerable residual uncertainty on certain results, the study clearly highlights the simultaneous importance of individual and contextual factors in the general population.

Across the western United States, most populations of Rocky Mountain elk (Cervus canadensis) are managed through hunter harvest (Sandrey 1983). In hunted populations, human harvest is the leading source of mortality (Ballard 2000). With the exception of elk in national parks, populations are primarily managed through hunter harvest. Other sources of mortality include disease, automobile collisions, and predation from mountain lions, coyotes, wolves, and black bears (Childress 2003, Hornocker 1970). As a species of management concern and high economic value (Pickton 2008), hunted populations of elk are carefully managed to target population sizes to guarantee a sustainable supply and ensure available harvest in the future. Estimates of survival, an understanding of cause-specific mortality, and knowledge of high-risk harvest areas are essential to effective management strategies (Stussy 1994). In the first chapter, I examined habitat use of elk during the hunting season and determined habitat characteristics that best predicted vulnerability to harvest. In the second chapter, I calculated annual survival and determined cause-specific mortality of elk.

Background: While death from melanoma of the skin has been gradually decreasing over the past few decades, melanoma continues to be the leading causes of death among skin cancers. Less is known about specific causes of mortality among patients with melanoma and how or whether trends in cause of death among patients diagnosed with melanoma have changed in recent years. Objective: To examine temporal trends in the cause-specific mortality among adult patients diagnosed with melanoma in the US between 2000-2013. Methods: US patients ≥ 45 years when diagnosed with melanoma were identified using data from the Surveillance, Epidemiology, and End Results Program, 18 Registries (SEER-18). Joinpoint regression analysis was used to examine the trends in cause-specific mortality among patients who were diagnosed with melanoma and died from either melanoma or other causes of death. Trends were also examined separately by age, sex, and geographic region. Results: A total of 52,675 patients diagnosed with melanoma who died from either melanoma or other cause of death (median age 74 years, 67% male) were included in the analysis. Overall, 31% of deaths were due to melanoma specifically, whereas 69% died from various other causes. A marked decline in melanoma-specific mortality was observed overall and across strata by age, sex, and region in the US beginning around 2013-2014. Among all causes of death, 55% were due to melanoma within 1 year after diagnosis and declined to 25% over the course of 6 years. A marked decline of at least 2.5% in mortality per year from other causes was observed among females, males, those 65 – 74 years or 75 years and older, and those living in northeastern, midwestern, western, and southern regions of US who were diagnosed with melanoma. Conclusions: Changes in cause-specific mortality rate among patients with melanoma were observed overall and across different subgroups. Our findings show that, among those diagnosed with melanoma, the risk of melanoma-specific death is decreasing within the last two decades, and that the deaths among those with melanoma are more likely to be from other causes such as heart disease, lung cancer, and other conditions. Future studies are needed to assess the trends in melanoma mortality as treatments and diagnostic methods continue to advance.

The eastern wild turkey (Meleagris gallapovo silvestris) was nearly extirpated from the Mississippi Alluvial Valley (MAV) from overharvesting and habitat loss in the early 1900s. Habitat restoration with hardwood regeneration has likely improved habitat suitability for turkeys in the MAV. I studied establishment and movement of translocated turkeys to assess feasibility of turkey restoration in the MAV. Turkeys were translocated to 2 study sites in the MAV during winters of 2009 and 2010. I monitored movement, survival, and resource selection of translocated turkeys using radio telemetry (February 2009–April 2011). Movement increased with increasing resource dispersion. Mortalities did not differ among 4 causes. Biweekly survival of translocated turkeys was related inversely to spring precipitation, but related positively to precipitation during other seasons. Mature and regenerating hardwood forests served as nesting and brooding habitat; therefore, increasing hardwood forests will likely improve success of future turkey restoration in the MAV.

<p>The principle objective of this investigation was to develop an understanding of the physiological response and ecological aspects of winter torpor of small carnivores, specifically striped skunks (<i>Mephitis mephitis</i>) in the northern environment. An experiment was undertaken to investigate the physiological response of skunks to solitary and communal over-winter strategies. Solitary skunks were able to undergo daily torpor to conserve energy to survive the winter, whereas communal skunks were able to use social huddling to reduce energy expenditure and rarely entered torpor. Due to seasonal changes in life requirements, den selection criteria change throughout the year. I examined the landscape metrics and habitat characteristics surrounding dens to evaluate the hierarchical selection and use of dens during winter (i.e., for torpor) and summer (i.e., for parturition). Den structures commonly used for winter dens were buildings, whereas den structures used for maternity dens were rockpiles and underground burrows. Habitat surrounding den sites influenced den use; animals chose den sites closer to roads, water sources, habitat edges, and crop fields. Seasonal movements of skunks from winter dens to the following summer home ranges were examined with respect to winter grouping (i.e., solitary or communal) and winter den structure (i.e., underground burrow or building) to investigate factors that influence spatial distribution of skunks. Male and female skunks moved similar distances (~1.5 km) from winter dens to the center of home ranges established in summer, irrespective of winter grouping status and winter den structure use. Due to limited movement in spring, skunks from communal winter dens had higher spatial overlap of summer home ranges than did skunks from solitary dens, producing spatial aggregation of skunk activities surrounding winter communal den sites. Consequently, winter communal dens represent localized hotspots in the landscape. The effects of extrinsic variables (i.e., environment and diseases) and intrinsic characteristics (i.e., age, gender, and body condition) on cumulative winter survival rate were examined. Winter survival rates were driven mostly by winter condition (i.e., low ambient temperature and snow depth), rabies, and body condition. These results suggest that winter severity probably poses a limit on the northern distribution of the species.</p>