Whale-watching provides economic opportunities worldwide and particularly proliferates in developing countries, such as the Philippines. The sustainability of whale-watching is increasingly debated as these activities also negatively impact cetaceans through changes in behavior, communication, habitat use, morbidity, mortality, and life-history parameters. This study evaluated the total annual cost, revenue, and profit of whale-watching operators in Bais, Philippines, and predicted the changes in the population for spinner dolphin Stenella longirostris with varying levels of whale-watching effort. Total revenue was 3,805,077 PHP ($92,478 USD) while total cost was 5,649,094 PHP ($137,294 USD) with a discount rate of ten percent. The total annual profit of whale-watching in Bais was – 1,844,017 PHP (– $44,817 USD). On average, each operator in Bais lost 160,350 PHP ($3,897 USD) per year from whale-watching. Through time, the spinner dolphin population decreased as it was exposed to more vessels, causing effort to increase, and thus decreased profit for operators. Under current whale-watching effort, the spinner dolphin population was predicted to decrease by 94 percent in 25 years. If Bais reduced effort in their operations to only three vessels whale-watching per day, the spinner dolphin population increased to 80 percent of its initial population size. This was the first study to predict the spinner dolphin population and estimate the total annual profit from whale-watching in Bais, Philippines. It provided data to locals for efficient, profitable, and sustainable decisions in whale-watching operations.

<p>  </p> <p>Blanding’s Turtle populations face direct threats to their survival. To help protect populations, habitats that can best support Blanding’s Turtle populations need to be identified across their range. Blanding’s Turtles have been a difficult to detect species and may be present at a site even if not detected during targeted surveys. Additionally, Blanding’s Turtles may be present at a site but may have little to no recruitment so additional measures of site suitability beyond species presence are needed to determine more suitable or higher quality habitats. In my research, I attempt to determine suitability of sites for Blanding’s Turtles across Michigan and Ohio using data collected from rapid assessment protocols fit into single season occupancy models with wetland and upland landcover types as co-variates of occupancy. To further determine the suitability of sites based on these data, I created single season occupancy models for juvenile Blanding’s Turtles and used N-mixture abundance modelling to determine relative abundance of Blanding’s Turtles at a site using the same landcovers as covariates of occupancy and abundance. Both modelling frameworks also allowed me to include detection covariates that could increase Blanding’s Turtle detection in future surveys. </p> <p>Detection was largely influenced by Julian date with the highest probability of detection occurring from mid-May through late June. Length of trapping surveys was also found to influence Blanding’s Turtle detection with a substantial decrease in daily trap capture rates by the fourth trap night of a survey. Michigan occupancy and abundance models found that the most suitable sites in Michigan would have high percentages of high-quality upland forest and woody wetland landcovers, with the percentage of open water supporting the occupancy of turtles but having no discernable effect on abundance. Total upland forest also significantly increased the probability of juvenile occupancy in Michigan. In Michigan, I also observed that survey method can greatly influence the estimates of occupancy and abundance, and I determined that visual surveys cannot accurately determine these estimates. The heavily disturbed nature of Ohio’s landscape took away from the predictive power of landcovers used in my research for Blanding’s Turtle occupancy and abundance. The vast difference between occupied habitats in Michigan and Ohio also takes away from the predictive power of the regional level model and relative abundance of Blanding’s Turtle populations cannot be accurately determined at this scale using the spatial covariates I included. However, total undisturbed forest and total wetland proved to be positive covariates of Blanding’s Turtle abundance and occupancy for adult and juvenile turtles across both states, but the habitats used in each state vary greatly so future conservation decisions should be made on the state level as largest spatial scale. Using my models for Michigan suitable sites can be determined within the state and compare relative abundance between sites to determine healthier populations. For future analysis in Ohio, different, smaller scales spatial covariates should be used to explain differences in occupancy and abundance between sites.</p>

Snapping turtles (Chelydra serpentina) are being harvested in unprecedented numbers in the United States (US) to meet the needs of international markets. Over three million live snapping turtles from farm and wild caught stock were exported from the US to Asia in 2012-14 alone. In the Commonwealth of Virginia, records indicate that 29,860 snapping turtles were commercially harvested between 2000 and 2015. Size limits are often used to regulate harvest pressure in snapping turtles and other game species. I analyzed the historic harvest of eleven US states to test the efficacy of minimum-size limit regulations at reducing commercial harvest pressure. Further, I conducted a four-year mark-recapture study on three Virginia waterways that have each experienced a different level of historic commercial harvest. As part of the larger mark/recapture project, I conducted radio telemetry on 23 turtles to examine seasonal, body size, and sex-specific effects on home range size of snapping turtles in a lotic system. I incorporated survival and growth rates from this study, demographic rates from the literature, and state-collected harvest rates into a hybrid age/stage population matrix model to estimate the population growth rate at three harvest levels (0%, 21%, 58%) that were estimated based on annual commercial landing reports on file with the Virginia Department of Game and Inland Fisheries. I used the model to test population viability under multiple size limit regulations, and used sensitivity analyses to identify adult stages most critical to the overall population growth rate. Based on model estimates, size-limits were effective at reducing harvest by 30-87% in years with high harvest pressure. However, most size limit regulations result in the removal of larger breeding adults, which has been shown to be detrimental to long term population viability. Based on radio-telemetry data, I found evidence that snapping turtles utilize lotic and lentic habitats differently, which can have implications for management of this iconic species. Matrix population modelling predicted that population densities at the moderate and high harvest site were reduced by 47% and 62%, respectively, when compared to the no harvest site. Model results indicate that, while an increase to the minimum-size limit in 2012 protected a larger portion of the population, that the commercial harvest of snapping turtles in the Commonwealth of Virginia is not sustainable under current state regulations. Our analysis suggests that minimum-size limits of 35.6 cm curved carapace length or greater will maintain viable populations by protecting a larger portion of reproducing snapping turtles within a population.

Nowadays, in sub-Saharan Africa (SSA), about 40% of the population is urban and this region is expected to face the highest growth rates during the next decades. By 2100, the three most populated cities in the world will be located in SSA. As a consequence of the extremely fast transformations experienced during the last decades, SSA cities are facing social and environmental issues combined with a lack of financial means and capacity in urban planning and management. The poorest often constitute a large part of the urban population that is extremely vulnerable to health and disaster risks.In SSA cities, up-to-date and spatially detailed geographic information is often missing. This lack of information is an important issue for many scientific studies focusing on different urban issues and there is a real need to improve the availability of geoinformation for these cities in order to support urban planning, urban management, environment monitoring, epidemiology or risk assessment, etc. The work presented in this thesis aims to develop different frameworks for the production of geoinformation. For this purpose, advantage is taken of Very-High Resolution Remote Sensing imagery (0.5 meters) and open-source software. These frameworks implement cutting-edge methods and can handle a large amount of data in a semi-automated fashion to produce maps covering very large areas of interest. In the spirit of open science, the processing chains are entirely based on open-source software and are released publicly in open-access for any interested researchers, in order to make the methods developed completely transparent and in order to contribute to the creation of a pool of common tools and scientific knowledge. These frameworks are used to produce very detailed land-cover and land-use maps that provide essential information such as the built-up density, or the fact that a neighborhood is residential or not. This detailed geoinformation is then used as indicators of presence of populated places to improve existing population models at the intra-urban level. / Option Géographie du Doctorat en Sciences / info:eu-repo/semantics/nonPublished

This thesis is presented as a collection of five papers and one report, each of which has been either published after peer review or submitted for publication. It covers a broad range of applied statistical methods, from deterministic modelling to integrated Bayesian modelling using MCMC, via bootstrapping and stochastic simulation. It also covers a broad range of subjects, from analysis of recreational fishing diaries, to genetic mark recapture for wombats. However, it focuses on practical applications of statistics to the management of wild populations. The first chapter (Hoyle and Jellyman 2002, published in Marine and Freshwater Research) applies a simple deterministic yield per recruit model to a fishery management problem: possible overexploitation of the New Zealand longfin eel. The chapter has significant implications for longfin eel fishery management. The second chapter (Hoyle and Cameron 2003, published in Fisheries Management and Ecology) focuses on uncertainty in the classical paradigm, by investigating the best way to estimate bootstrap confidence limits on recreational harvest and catch rate using catch diary data. The third chapter (Hoyle et al., in press with Molecular Ecology Notes) takes a different path by looking at genetic mark-recapture in a fisheries management context. Genetic mark-recapture was developed for wildlife abundance estimation but has not previously been applied to fish harvest rate estimation. The fourth chapter (Hoyle and Banks, submitted) addresses genetic mark-recapture, but in the wildlife context for estimates of abundance rather than harvest rate. Our approach uses individual-based modeling and Bayesian analysis to investigate the effect of shadows on abundance estimates and confidence intervals, and to provide guidelines for developing sets of loci for populations of different sizes and levels of relatedness. The fifth chapter (Hoyle and Maunder 2004, Animal Biodiversity and Conservation) applies integrated analysis techniques developed in fisheries to the modeling of protected species population dynamics - specifically the north-eastern spotted dolphin, Stenella attenuata. It combines data from a number of different sources in a single statistical model, and estimates parameters using both maximum likelihood and Bayesian MCMC. The sixth chapter (Hoyle 2002, peer reviewed and published as Queensland Department of Primary Industries Information Series) results directly from a pressing management issue: developing new management procedures for the Queensland east coast Spanish mackerel fishery. It uses an existing stock assessment as a starting point for an integrated Bayesian management strategy evaluation. Possibilities for further research have been identified within the subject areas of each chapter, both within the chapters and in the final discussion chapter.

No description available.

To underpin and improve advice given to government and other interested parties on the state of Britain’s common songbird populations, new models for analysing ecological data are developed in this thesis. These models use data from the British Trust for Ornithology’s Constant Effort Sites (CES) scheme, an annual bird-ringing programme in which catch effort is standardised. Data from the CES scheme are routinely used to index abundance and productivity, and to a lesser extent estimate adult survival rates. However, two features of the CES data that complicate analysis were previously inadequately addressed, namely the presence in the catch of “transient” birds not associated with the local population, and the sporadic failure in the constancy of effort assumption arising from the absence of within-year catch data. The current methodology is extended, with efficient Bayesian models developed for each of these demographic parameters that account for both of these data nuances, and from which reliable and usefully precise estimates are obtained. Of increasing interest is the relationship between abundance and the underlying vital rates, an understanding of which facilitates effective conservation. CES data are particularly amenable to an integrated approach to population modelling, providing a combination of demographic information from a single source. Such an integrated approach is developed here, employing Bayesian methodology and a simple population model to unite abundance, productivity and survival within a consistent framework. Independent data from ring-recoveries provide additional information on adult and juvenile survival rates. Specific advantages of this new integrated approach are identified, among which is the ability to determine juvenile survival accurately, disentangle the probabilities of survival and permanent emigration, and to obtain estimates of total seasonal productivity. The methodologies developed in this thesis are applied to CES data from Sedge Warbler, Acrocephalus schoenobaenus, and Reed Warbler, A. scirpaceus.