No description available.

Recreational hunters play an important role in managing white-tailed deer (Odocoileus virginianus); however, the potential for deer to alter behaviors to avoid hunters has not been addressed within the risk-allocation hypothesis. I evaluated magnitude (i.e., hunter density) and temporal variation (i.e., time of day and initial and prolonged exposure) in human predation risk on movements, resource selection, and observation rates of 37 adult male deer in southern Oklahoma. Deer recognized human predation risk by increasing diel path complexity and use of security cover with greater hunter density. Moreover, deer reduced movement rates and tortuosity while seeking out areas with security cover during prolonged exposure. However, tortuosity and use of security cover remained elevated with greater hunter density. These alterations in behaviors subsequently led to a decrease in observation rates during prolonged exposure. My results clearly support the predation risk-allocation hypothesis by the behavioral responses observed with greater hunter density.

The ‘landscape of fear’ has been proposed as a unifying concept in ecology by linking population distribution patterns through top-down predator-prey mechanisms. The landscape of fear predicts that prey resource selection patterns are influenced by spatially and temporally predictable patterns of predator risk across a landscape. Although the model has been suggested to predict prey space-use patterns across a variety of systems, it remains unclear if individuals exposed to similarly risky environments (i.e., within the home range) will consistently avoid predator risk. I tested the landscape of fear concept using a natural experiment where moose hunting was introduced to a previously hunter-naïve moose population. I quantified hunting risk by developing risk landscape layers derived from harvest data collected over the first three hunting seasons (2011, 2012 and 2013) in Gros Morne National Park, Newfoundland, Canada. Yearly hunter-risk layers were then used as a continuous variable in moose resource selection analysis to understand if moose respond to hunter risk, and if all individuals respond similarly to risk, as predicted by the landscape of fear. I found moose hunters were generally more likely to harvest moose near landscape features that offered easy access and a wide field of view. Moose generally did not avoid hunting risk until the second hunting season, and only during daylight hours. Conversely, at night, moose were generally found to select hunter-risky areas, with the strength of selection progressively increasing each year. I found considerable individual variation in moose response to hunting risk, however, with some individuals failing to alter selection strategies to avoid hunter risk. The motivation to respond to risk may be based on fitness related trade-offs associated with anti-predatory behavior, personality, and/or an individual’s ability to correctly assess risk on the landscape. My research highlights the importance of incorporating individual patterns in resource selection strategies when attempting to address landscape-level processes, such as the landscape of fear concept.

The greater sage-grouse (Centrocercus urophasianus; sage-grouse) has been a species of conservation concern since the early 20th century due to range-wide population declines. To contribute to knowledge of the ecology of sage-grouse populations that inhabit the Box Elder Sage Grouse Management Area (SGMA) in northwestern Utah and quantify their responses to landscape scale habitat manipulations, I monitored vital rates and habitat selection of 45 female sage-grouse from 2014 to 2015. Using telemetry locations of female sage-grouse with known nest and brood fates, I created Generalized Linear Mixed Models to estimate the influence of proximity to pinyon (Pinus spp.) and juniper (Juniperus spp.; conifer) encroachment, and removal projects may have on sagegrouse reproductive fitness in the Box Elder SGMA. The best fit model suggested that for every 1 km a nest was located away from a conifer removal area, probability of nest success was reduced by 9.1% (β = -0.096, P < 0.05). Similarly, for every 1 unit increase in the log-odds of selection for distance to treatment, probability of brood success declined by 52.6% (P = 0.09). The probability of brood success declined by 77.2% (P < 0.05) as selection for conifer canopy cover increased. To evaluate sage-grouse habitat use, I used fecal pellet surveys to estimate relative pellet density in conifer encroachment, removal, and undisturbed sagebrush areas. Sage-grouse pellet densities were estimated at 4.6 pellets/ha (95% CI = 1.2, 10.9), 8.6 pellets/ha (95% CI = 3.8, 15.2), and 50.6 pellets/ha (95% CI = 36.8, 69.6), in conifer encroachment, removal, and undisturbed sagebrush areas respectively. Density estimates did not statistically differ between conifer encroachment and removal areas. To determine if vegetation micro-site characteristics at sage-grouse use sites influenced nest or brood fate, I recorded standard vegetation measurements for all radio-marked sage-grouse nests and a stratified random sample of brood-use sites from 2014- 2015 and compared them to random sites. Micro-site vegetation characteristics measured did not differ for successful and unsuccessful nests. Many characteristics differed between micro-sites used by successful broods and those used by unsuccessful broods. Sites used by successful broods also differed from random sites.

Indices of white-tailed deer (Odocoileus virginianus) abundance in the western Upper Peninsula of Michigan suggested the population declined 40% from the previous 5 year mean following 2 consecutive severe winters in the mid-1990s and has not since increased in population size. I collected estimates and assessed biological and environmental covariates of survival and recruitment of fawns (< 1 year old) and age-specific reproductive and survival rates of adult females (> 1.6 years old) in the southwestern Upper Peninsula of Michigan from 2009–2011. Reproduction did not appear to regulate population growth, as 92 percent of females were pregnant. Annual survival of adult females was 70 percent across years, but poorer annual survival of fawns across years (44 percent) resulted in recruitment being the most influential vital rate to population growth, which increased10 percent from 2009 to 2010, but decreased 13 percent from 2010 to 2011. Variation in population growth emphasized that annual variation in fawn recruitment may have nullified increased growth over time. Most fawn mortalities occurred within 12 weeks of age, emphasizing this period greatly influenced annual survival rates of fawns, and especially population growth. Therefore, I suggest fawns should be considered the priority cohort for deer population management, including mitigation of factors which limit fawn recruitment. Winter severity effects on nutritional condition of adult females primarily influenced survival of adult females and fawns. However, adult female avoidance of interior lowland forests which had greater wolf (Canis lupus) use and commonly aging and over-browsed vegetation ostensibly reduced fawn recruitment through a lack of hiding vegetation and poorer forage. Also, by adult females raising fawns in habitats near roads, the predatory efficacy of coyotes (C. latrans) on adult females and fawns increased. Although predation was the leading cause of deer mortality, bottom-up effects of winter severity on nutritional condition and resource use appeared to be most influential to sustaining a lack of population increase. Hence, I suggest population growth could be improved through habitat management that increases landscape heterogeneity of early successional forests to enhance year-round browse to increase nutritional condition of adult females and hiding cover for fawns.

Wildlife management and conservation frequently rely on understanding mechanisms that influence distribution and abundance of animals. I quantified space use for a population of inland riverine adult male alligators in Mississippi. Results indicated habitat selection is a scale-dependent process and aquatic vegetation, water depth, and water temperature may be important factors influencing alligator foraging and thermoregulation. Apparent habitat suitability and low alligator density did not manifest in an observed body size-based dominance hierarchy. I also analyzed long-term Mississippi alligator spotlight survey data for trends and effects of environmental covariates on counts. Model results indicated alligator counts have increased over time. This response likely reflects benefits accrued from decades of protection and wetland conservation. Distance sampling does not appear to be a feasible monitoring technique for riverine alligator populations. Nevertheless, it is important that survey protocols and monitoring programs account for imperfect detection and model important covariates.

Historically, black bears occurred throughout Mississippi but by 1932, <12 bears remained. Repatriation in neighboring states and conservation efforts in Mississippi have led to the recolonization of at least 2 subspecies (U. a. luteolus and U. a. americanus) of black bears in the state. I compiled available data to provide a synthesis of the history, current status, and management of black bears in Mississippi. Additionally, I used global positioning data collected from radio collared bears to determine the influence of distance to source population, cover type, distance to roads, distance to water, wetland reserve program areas, and human population density on black bear resource selection at various spatial scales. I studied characteristics of space use and resource selection of recolonizing bears in the Mississippi Alluvial Valley (Delta). I assessed the influence of environmental parameters at the female core annual home-range (using 50% kernel density estimator) and male and female seasonal and annual home-ranges (95% kernel density estimator). Distance to source population and distance to roads had significant influence at the core female home-range scale. I found a sex-based difference in annual and seasonal home-ranges. I also found that bears exhibited response to and selection for specific resources with an affinity toward hardwood stands, particularly young-aged hardwoods. My research illustrates the importance of analyzing resource selection at multiple scales to gain a full understanding of parameters that influence the recolonization of a bear population.

A Internet das coisas é constituída de objetos que possuem pequenos sensores e atuadores capazes de interagir com o ambiente. Tais objetos ou coisas estão interconectados entre si e com acesso à Internet por meio de redes com e sem fio. A combinação entre os dispositivos embarcados com sensores e o acesso à Internet possibilita a comunicação dos recursos do mundo físico com o espaço cibernético, desempenhando um papel fundamental na resolução de muitos desafios encontrados na sociedade atual. Porém, a maioria das aplicações existentes são dedicadas a resolver problemas específicos utilizando tais recursos apenas em redes internas, limitando a real capacidade da Internet das Coisas. Diversos trabalhos na literatura propõem a reutilização de tais recursos em forma de serviço por meio de modelos como Dados como Serviço e Sensoriamento como Serviço. Neste contexto, em que potencialmente milhares de recursos podem transferir dados semelhantes de aplicações diferentes, a utilização de técnicas que possam selecionar recursos de forma sensível a contexto torna-se imprescindível. Nesta tese são propostos um conjunto de métodos para melhorar a relação custo-benefício na seleção de recursos em ambientes IoT, auxiliando na tomada de decisão durante a seleção dos recursos que serão ofertados como serviço. Os resultados obtidos por meio de estudos de caso, permitiram a comparação da qualidade da solução e do custo computacional das técnicas aplicadas na seleção de recursos em ambientes IoT, bem como o desenvolvimento de duas novas técnicas para a seleção de recursos, denominadas Elimination Sort e Fast Elimination Sort. / The Internet of Things is composed of objects which have small sensors and actuators capable of interacting with the environment. Such objects or things are interconnected with each other and has access to the Internet through wired and wireless networks. The combination of embedded devices with sensors and access to the Internet become it possible to communicate the resources of the physical world with the cyberspace, playing a key role in solving many challenges found in todays society. However, most existing applications solves a specific problem using its resources just for own purpose, limiting the actual ability of the Internet of Things. Several works propose the reuse of such resources through service models such as Data as Service and Sensing as a Service. In this context, where thousands of resources can transfer similar data from different applications, the use of techniques that can select these features in a context-sensitive way becomes essential. In this thesis, a set of methods to improve the cost-benefit of the process of selection of resources in IoT environments is proposed to support the decision making during resource selection that will be offered as a service. The results obtained through a case study allowed the comparison of the solution quality and the computational cost of the techniques applied for resource selection in IoT environments, as well as the development of two new techniques for the selection of resources called Elimination Sort and Fast Elimination Sort.

Habitat loss and fragmentation are among the greatest threats to biodiversity, and these threats can be exacerbated or alleviated by the presence of interacting species. The effect of habitat loss and fragmentation on predator-prey systems has received extensive theoretical attention, but empirical studies of these systems yield few clear patterns. I examined the influence of prey abundance and spatial distribution on the foraging ecology and spatial ecology of Masticophis flagellum (Coachwhip) using capture-mark-recapture and radio telemetry techniques. I also examined the influence of saurophagous snake abundance on the survival rate of Sceloporus woodi (Florida Scrub Lizard) populations. Masticophis flagellum positively selected lizard and mammal prey, but within these categories it consumed prey species in proportion to their availability. Masticophis flagellum was vagile and constrained its movements within large home ranges. At all spatial scales examined, M. flagellum strongly selected Florida scrub habitat and avoided wetland habitats. The negative effect of saurophagous snake abundance best explained differences in S. woodi survival rates among patches of Florida scrub. Further loss and fragmentation of Florida scrub habitat will likely have a strong negative impact upon M. flagellum. Because it is precinctive to Florida scrub, Sceloporus woodi will also be negatively affected by the loss of this unique habitat. The potential positive effects of reduced predation pressure from M. flagellum that may accompany loss and fragmentation of Florida scrub is likely to be offset by increased predation rates by habitat and dietary generalist predators that incidentally prey upon S. woodi. Despite the sensitivity of these species to loss and fragmentation of Florida scrub, the prognosis is good for both M. flagellum and S. woodi on relatively large protected sites containing xeric habitats managed with prescribed fire.

The extent to which predators regulate prey populations remains a subject of debate. Yet, when predator control is employed as a management strategy, it is often assumed that predators can and do regulate prey populations. From 2011 through 2015, I monitored the demography and space use of coyotes (Canis latrans) and cougars (Puma concolor) on Monroe Mountain in Fishlake National Forest, Utah as part of a larger collaboration investigating the impacts of coyote aerial control on mule deer (Odocoileus hemionus) neonate survival. My primary objective was to assess the impacts of anthropogenic regulation on the respective populations and identify any cascading effects relevant to mule deer management. To meet this objective, I established a monitoring program for both predators by deploying radio-telemetry collars (VHF and GPS) on each, documented predation events, established surveys for small mammals and lagomorphs to monitor primary prey populations during deer parturition (June – August), and collected data on the location and demographic composition of winter-removed coyotes. I analyzed these data primarily in a community-based, animal movement and resource selection framework permitting the integration of data from multiple sources. When evaluating coyote aerial removal as a management strategy, I identified a spatial dependency in the ability to match removals with indices of deer recruitment as Wildlife Services Operations personnel were primarily limited by terrain and tree cover. Thus, matching treatment with deer fawning was highly variable with only a small number of sites where removals were effective. In addition, I found that coyotes selected for sites with the highest densities of lagomorphs while avoiding areas with a high probability of encountering cougars. Coyotes did not select for mule deer fawning sites, although individual coyotes that occupied resource-poor home ranges were more likely to do so. Cougars strongly selected for mule deer high use areas throughout much of the year, only switching to elk (Cervus elaphus) during the cougar harvest season (i.e., winter). Data from cougar kill site investigations match the observed patterns in cougar space use. My results suggest that predator-prey processes are multi-dimensional and dynamic through time, which likely contribute to the lack of resolution regarding the efficacy of predator control and the regulatory potential of predators in general.