<p>The behavior of individuals in predator-prey interactions plays a fundamental role in population ecology and evolution. Accordingly, the development and assessment of theory that provides a conceptual framework for understanding and predicting behavior in predator-prey systems is of basic importance. This dissertation is focused on developing and assessing theory that provides a quantitative structure for predator-prey interactions.Chapter 1 provides a theoretical treatment of the problem of modeling the behavior of an animal facing two- and three-way tradeoffs among growth, mortality and reproduction, three components of Darwinian fitness. In this situation benefits in terms of one fitness component can be obtained only with a cost in terms of another fitness component. Applying a dynamic optimization method, the theory describes the behavior of phenotypes that adaptively balance growth, mortality and reproduction in a size-structured population. One simple prediction of the theory is that non-growing adults in a stable population should behave so as to minimize the ratio of mortality rate to birth rate. A further extension of the theory using the idea of ideal-free distributions results in a prediction of a positive and linear relationship among the rates of growth, mortality and birth when measured across different habitat patches. These and other results lead to testable predictions and I discuss empirical examples of how subsets of the theory have been and can be assessed. Chapter 2 is an empirical assessment of four alternative models of the functional response of predator to prey (the per capita feeding rate of a predator) using 19 data sets from the literature. The results show that predator-dependent functional responses (i.e., forms that are functions of predator and prey abundance because of predator interference) provide better descriptions of 18 of the 19 data sets than the Holling Type II model, a commonly used prey-dependent functional response (i.e., a model that depends only on prey abundance). Hence some form of predator interference is common in these data. However, no single functional response can best describe all of the data sets. A key result is that the best-fitting predator-dependent model depends on the presence or absence of predator-dependence when prey are very abundant. Accordingly, I suggest use of the Beddington-DeAngelis or Hassell-Varley model when predator feeding rate becomes independent of predator density at high prey density, and use of the Crowley-Martin model when predator feeding rate is decreased by higher predator density even when prey density is high. These results suggest that predator-dependent functional responses should be more widely considered in the literature. Chapter 3 is an empirical assessment of four alternative optimization models of behavior for animals facing a feeding-mortality tradeoff (i.e., increases in feeding rate can be obtained only at a cost of increasing mortality rate by predation). Using lab experiments, I measured the feeding and mortality of stream fish (bluehead chubs) in the presence of varying levels of food availability and green sunfish predators. The models vary in their specifications as to how bluehead chubs might value, in terms of fitness, rates of growth and mortality. Two models that emphasize the importance of growth or mortality alone fit the data poorly, whereas the two models that balance growth and mortality adaptively provide a much better description of the data because bluehead chubs continue to feed, albeit at a reduced level, in the presence of the predator-maintained mortality hazard. A comparison of the two best-fitting models leads to the inference that reproductive value in the bluehead chub should be thought of as a function of body size and age rather than a function of body size alone. A key implication of this result is that risk-taking behavior in these fish can be quantitatively estimated using a parameter that incorporates the benefits of growth and the costs mortality into a single fitness-based metric. The metric, called the marginal rate of substitution of mortality rate for growth rate, provides a link between behavior and life history in the bluehead chub. In summary, this dissertation provides some solutions to the challenge of how to conceptualize and predict animal behavior in predator-prey settings. The emphasis is on developing and assessing alternative hypotheses in a quantitative framework. <P>

<p>Quantitative trait loci (QTL) are those chromosome regions that contribute to variation of quantitative traits. Analysis of QTL is helpful for further study of molecular basis of the quantitative genetic variation. The discovery of highly abundant and dense polymorphic markers (e.g., single nucleotide polynorphisms, or SNPs) covering a whole genome provides an opportunity to localize QTL in a variety of populations. While classical linkage studies have a relatively limited resolution in QTL localization, the association mapping or linkage disequilibrium (LD) mapping approach can offer an alternative way for fine mapping of genes. Currently there are few LD methods available for QTL mapping in natural populations. Development of more efficient methods is still a challenging problem. In this thesis, we first review the LD approach for fine mapping of QTL in Chapter 1. Some basic issues in LD analysis and recent developments in LD methodology are discussed. Particular attention is paid to limitations and potential problems of these methods. This provides the motivation for the research in this thesis.Next, we explore Cockerham's genetic model (Cockerham, 1954) for quantitative traits in Chapter 2. A revised form of the Cockerham model is presented using some coding variables. The relationship between Cockerham model and some specific genetic models for designed experimental populations, such as backcross or F2, is then established. We study extensively the properties of QTL effects and partitions of various genetic variance components for these reduced models under both linkage equilibrium and linkage disequilibrium situations. A general multi-locus-two-allele model is also proposed that may serve as a basis for mapping QTL in natural populations. The main research of the thesis is on development of an exact multipoint likelihood approach to infer QTL in natural populations. In Chapter 3, we first generalize the formulation of the likelihood analysis for a polymorphic marker locus and a trait locus in a general natural population. From this generalization, we derive a closed form solutionof an efficient EM algorithm for the likelihood analysis. This is a major achievement of the research. The importance of this generalization is that it can be readily and systematically extended to multiple markers and multiple QTL.Based on this formulation, a multipoint likelihood analysis with the EM algorithm is developed that can take into account higher-order linkage disequilibria between QTL and markers without making approximation to the likelihood function. This analysis can offer a simultaneous estimation of the linkage disequilibrium structure between a QTL and multiple markers. From this estimation, we find that the linkage disequilibrium between one or a subset of markers and a QTL conditional on other markers can offer as a more accurate measure for fine mapping of QTL. In Chapter 4, we further extend the analysis to multiple QTL and propose a general framework for likelihood analysis of multiple QTL and markers. With this approach, the joint gametic frequencies of QTL and markers (thus various measures of linkage disequilibria between QTL and markers) as well as various genetic effects of QTL (including epistasis) can be estimated simultaneously. This general approach has a lot of potential for a complete analysis of genetic architecture of quantitative traits in natural populations. Although the foundation of this general framework has been laid down, more studies are still needed on a number of issues, such as efficiency and reliability of the optimization algorithm, statistical tests for QTL identification, model selection of complex QTL, and more efficient approaches to analyze large data sets. In the last chapter (Chapter 5), we draw some general conclusions from the research. We discuss the advantages as well as limitations of the approach developed in this thesis and problems for further research.<P>

Thesis (Ph.D.) -- University of Texas at Arlington, 2009.

Woven string kernels are a form of evolvable, directed, acyclic graphs specialized to perform DNA classification. They are introduced in this thesis, given a rigorous theoretical treatment as a mathematical object, and shown to have a number of interesting properties. Two forms of woven string kernels, uniform and non-uniform, are discussed. The non-uniform woven string kernels are repurposed for use as updating rules for cellular automata. The details of their representation and implementation are presented. A chapter of this thesis is devoted to a visualization technique called non-linear projection, an evolvable form of multidimensional scaling that is used in the analysis of experimental results. The woven string kernels are tested on simple and complex synthetic data as well as biological data, using an evolutionary algorithm to find woven string kernels that are acceptable solutions for classification. They perform marginally on the simplest synthetic data - based on GC content - for which they are not entirely appropriate. They exhibit perfect classification on the more complex synthetic data and on the biological data. Woven string kernels have a number of parameters including their height, the number of initial strings from which they are built, and the amount of weaving used to generate the final structure. A parameter study shows that these parameters must be set based on the type of data under analysis. Experimentation with woven string kernels as rules for updating cellular automata show that having a larger population and more available colour states are correlated with an increase in performance as apoptotic one dimensional cellular automata. This thesis concludes with directions for future work related to theory and experimentation, for both uniform and non-uniform woven string kernels.

Suppose two treatments are being compared in a clinical trial with a fixed number of subjects. Half the subjects receive treatment 1, while the other half receive treatment 2. Measurements of some relevant response variable are made on each subject at equally spaced time points. This thesis proposes a procedure for testing at each time point the null hypothesis H₀ of no treatment differences, while controlling the overall probability of type I error. Chapter 1 describes alternative methods for estimating the unknown parameters of models generating short series of measurements. Maximum likelihood estimation, and modified variance based estimation, are considered in 1.1 and 1.2 respectively. A 4 parameter model is described in 1.3. This allows for increasing or decreasing variances of measurements within subjects. In 1.4, maximum likelihood estimation is used in fitting the 4 parameter model to two data sets. In Chapter 2, attention focuses on sequential testing, with data arising from the 4 parameter model. The possibility of using stopping rules based on posterior probabilities, rather than sequentially adjusted P values, is considered briefly in 2.1. Two methods for determining appropriate nominal significance levels for successive tests, when the model for data is assumed known, are described in 2.2. An algorithm for integrating quickly a special case of the multivariate normal density, in high dimensions, presented in 2.3, facilitates implementation of the two methods. In 2.4, test procedures based on the methods of 2.2 are described, in which model parameters are (re-) estimated at each time point, as data accumulate. Test statistics and z-values are adjusted for each test, according to the updated parameter estimates. Simulations show that these procedures give overall significance levels close to those required. One of the two procedures is preferred on the basis of power to detect deviations from H₀.

Thesis (M.S.)--University of Wisconsin--Madison, 1965. / eContent provider-neutral record in process. Description based on print version record. Bibliography: l. 134-141.

This thesis describes research conducted by the author using a high sensitivity, shadow-shield, mobile, whole body monitor and is primarily concerned with its clinical applications. Whole body counting-rates vary when administered radioisotopes are redistributed in the body. Several methods of assessing the magnitude of geometrical variations in counting-rates have been brought together. Point sources in simple water phantoms, a life-like phantom man with simulated organs and human subjects were studied and the relative importance of the sources of variation and the value of the methods assessed. Cobalamin excretion rates were studied in patients with renal and hepatic disease from about one week after injection when, evidently the tracer dose of vitamin has effectively equilibrated with body stores. The daily cobalamin loss was abnormally high in 6 out of 8 patients with renal disease and 3 out of 6 with hepatic disease. In evaluating the treatment of vitamin B12 deficiency, the retention of three parenterally administered vitamin compounds was investigated. From the results, maintenance schedules could be derived empirically for patients with uncomplicated vitamin deficiency and with concomitant renal or hepatic disease. Using a double tracer technique, the oral absorption at different dose levels of coenzyme and other cobalamins was measured. The fraction of the dose absorbed appears to be a function of both the mass and the structure of the cobalamin. The human prostate gland contains a high concentration of zinc and when a carcinomatous change occurs in the gland, there is a distinct fall in the zinc content of the prostatic tissue. A preliminary study of the whole body metabolism and local prostatic uptake of zinc-65 was undertaken to provide the basis of a method for measuring the response of prostatic carcinoma to therapy. The metabolism of iron-59 was studied in an investigation of the anaemia in patients with chronic renal failure and in patients also receiving regular dialysis therapy. It was suggested that the anaemia results from the depression of erythropoiesis rather than iron deficiency. The high iron turnover rate in patients receiving regular dialysis could be explained almost entirely by the blood losses in the Kolff twin-coil artificial kidney. Methods of minimising these blood losses are suggested. The monitor was calibrated using potassium-42 for the measurement of total body potassium. The precision of the calibration and monitoring procedures were examined in detail, a relationship was derived enabling a calibration factor to be calculated from a subject's weight and height. In clinical practice, the measurement of total body potassium may only be meaningful if the 'normal' value is known; this was correlated with the weight, height and age of the subject. To illustrate these techniques, total body potassium was measured in an investigation of the potassium status in patients with chronic, renal failure, rheumatoid arthritis and following ureterosigmoid anastomoses.

No abstract available

Thesis (M.S.)--Brigham Young University. Dept. of Mathematics, 2005. / Includes bibliographical references (p. 56).

Chlamydia trachomatis infections are a common sexually transmitted infection in the United States in which the majority of cases are asymptomatic. Due to this asymptomatic nature, as well as the serious health issues that arise from untreated infections in women, the Centers for Disease Control and Prevention (CDC) recommends a screening policy that annually targets women between the ages of 15 and 25 or older women with risk factors. There is little evidence supporting the efficacy of only screening women and doing so once per year. Through a stochastic epidemiological model, we investigate a variety of screening policies within a college setting and evaluate their impact on infection prevalence. We have developed a MATLAB program using an individual-based modeling approach to evaluate treatment and screening procedures. Using R, we present a statistical analysis of the outcome: under our model conditions, any procedure involving screening will eventually result in eradication of C. trachomatis infections in the population. Screening both men and women, however, seems to eliminate the infection in a population far more rapidly.