Browsing by Author "Cavell Brownie, Committee Member"

Now showing 1 - 14 of 14

The Association of Weed Species Richness and Abundance with Field Margin Type in Crop Fields
(2004-05-13) Jelinek, Susan T; Cavell Brownie, Committee Member; Michael G. Burton, Committee Member; Nancy G. Creamer, Committee Co-Chair; J. Paul Mueller, Committee Co-Chair
Natural vegetation on farms such as field margins, successional fallow fields, ditch systems, and neighboring forests provide increased biodiversity, structural diversity, habitat for wildlife and beneficial insects, and can act as protective buffers against agrochemical drift. Nevertheless, farmers frequently view these areas as potential sources of weeds, insect pests, and diseases. Objectives of this study were to examine weed species richness and abundance in cropland bordered by managed versus unmanaged field margins to determine if differences in weed infestation exist. Weed abundance and richness were measured in crop fields along permanent transects that extended from the field edge to the center of the crop fields. Presence/absence data for all plant species in the field margin were also recorded. Transect data from fields with margins of natural vegetation were compared to transect data from fields with managed margins using analysis of variance. There were no differences between log total abundance of weeds in crop edges adjacent to managed and unmanaged field margins (P=0.44). For both margin types, more weeds were found near the field edge than in the center of the field (1.37±0.08 to 0.52±0.07 and 1.39±0.07 to 0.41±0.06, for managed and unmanaged field margins respectively). Species richness was slightly higher along crop edges of managed field margins (7.35±0.32) than crop edges along unmanaged field margins (6.55±0.31). Across all sampling dates, a total of 105 plant species were identified in the field margins. Of these species, 42 (40% of all species) were found somewhere in a field when all sampling dates were pooled. Managed field margins had lower species richness than unmanaged field margins - less than half the mean number of species (5.8±0.28 versus 14.7±0.62 species, respectively). No association was found between plant species occurring in the field margin and in the crop field by generating 2 x 2 contingency tables via PROC FREQ and testing the association with Fisher's exact two-sided test. Using logistic regression via PROC GENMOD, margin type and weed presence in the field margin were not effective predictors of weed occurrence in the crop field. KEYWORDS: field margin, weed populations, crop edges, farm natural areas
Avian Point Count Surveys: Estimating Components of the Detection Process
(2004-05-17) Alldredge, Mathew Wade; Cavell Brownie, Committee Member; Theodore R. Simons, Committee Co-Chair; Kenneth H. Pollock, Committee Co-Chair; James D. Nichols, Committee Member; James F. Gilliam, Committee Member
Point count surveys of birds are commonly used to provide indices of abundance or, in some cases, estimates of true abundance. The most common use of point counts is to provide an index of population abundance or relative abundance. To make spatial or temporal comparisons valid using this type of count requires the very restrictive assumption of equal detection probability for the comparisons being made. We developed a multiple-independent observer approach to estimating abundance for point count surveys as a modification of the primary-secondary observer approach. This approach uses standard capture-recapture models, including models of inherent individual heterogeneity in detection probabilities and models using individual covariates to account for observable heterogeneity in detection probabilities. Two-observer models provided negatively biased estimates because they do not account for individual heterogeneity in detection probabilities. Models accounting for individual heterogeneity are always selected as the most parsimonious models for this data type. We also developed a time of detection approach for estimating avian abundance when birds are detected aurally, which is a modification of the time of removal approach. This approach requires collecting detection histories of individual birds in consecutive time intervals and modeling the detection process using a capture-recapture framework. This approach incorporates both the probability a bird is available for detection and the probability of detection given availability. Analyses presented demonstrate the importance of models accounting for individual heterogeneity in detection probabilities. We recommend time of detection point count surveys be designed with four or more equal intervals. We also present a multiple species modeling strategy since many point count surveys collect data on multiple species and present the approach for distance sampling, multiple observer, and time of detection approaches. The purpose of using a multiple species modeling approach is to obtain more parsimonious models by exploiting similarities in the detection process among species. We present a method for defining species groups which leads to an a priori set of species groups and associated candidate models. Multiple species models worked well and in many cases gave more parsimonious models than a species specific modeling approach, especially for the multiple-observer and time of detection approaches. Parameter estimates for multiple species models are more precise than single species models. We recommend this approach for all situations where data on multiple species is collected. Finally, we present a method for estimating the availability probability of birds during a point count based on singing rate or detailed singing time data. This approach requires data collected in conjunction with point count surveys that describe the singing rates or singing time distribution of the bird population of interest. The singing rate approach requires the assumption that an individual bird sings following a random process but rates may vary between birds. We modeled this using a finite-mixture Poisson model. The singing time approach is a nonparametric approach and does not require this restrictive assumption. Analysis of Ovenbird singing rate data demonstrates the importance of accounting for availability bias when estimating abundance, especially as count lengths get short. We recommend this approach when 'snapshot' type counts are necessary. Analyses presented throughout this thesis demonstrate the importance of accurately modeling the detection process to estimate abundance. The importance of accounting for individual heterogeneity in detection probabilities was evident in every chapter. Using a point count method that accounts for individual heterogeneity is crucial to estimating abundance effectively and making valid spatial, temporal and species comparisons.
Distribution of Tomato Spotted Wilt Virus (TSWV) in Relation to Wild Weedy Hosts and Susceptible Crops Over a Large Agricultural Landscape
(2005-04-25) Stout, Robyn Deanne; H. Michael Linker, Committee Co-Chair; Michael G. Burton, Committee Co-Chair; Cavell Brownie, Committee Member
Tomato spotted wilt virus (TSWV) is an important agronomic and horticultural pest in NC and other parts of the world. Many have speculated that management of weedy species that host the virus may aid in decreasing TSWV occurrence on a farm. In May of 2003 and 2004, weedy species were sampled around six fields, two fields per replication, at the Center for Environmental Farming Systems (CEFS) in Goldsboro, NC, using a 15 m x 15 m grid to map the overwintering pattern of TSWV on the farm. Weeds were tested for TSWV infection and weed density ratings were assigned in each grid cell. Indicator plants (tomato and peanut) were used to map TSWV occurrence within the fields. Thrips movement was monitored with sticky traps to reveal spatial patterns of movement across the farm and, along with indictor plants, were evaluated biweekly through the summers of 2003 and 2004 to monitor temporal movement/occurrence. Plots were located at field edges and in field centers to detect the differences between trivial movement of infected thrips and movement up to 60 m from bordering weedy hosts. TSWV occurrence was equally likely to be found in the center versus the edge plots, revealing that complete eradication of host species up to 60 m from a field may not aid in decreasing TSWV in a field. TSWV occurrence in and around the three replications that were spread across the farm revealed an effect of replication in 2003 when more virus was found in the southern-most replication. Ranunculus sardous (hairy buttercup) tested positive most often compared to all other weed species tested and was also found in the highest density surrounding the southern-most replication.
Effects of Litter Production, Biochemistry and Plant Community Composition on Carbon and Nutrient Cycling under Elevated Carbon Dioxide and Tropospheric Ozone.
(2008-10-17) Liu, Lingli; Christian Giardina, Committee Member; John King, Committee Chair; Cavell Brownie, Committee Member; Lee Allen, Committee Co-Chair; Fitzgerald Booker, Committee Member
Elevated CO2 and O3 have the potential to alter the productivity, biochemistry and species composition of leaf litter, which will affect litter decomposition, thereby controlling nutrient release rates and soil carbon formation. To assess those effects, leaf litter was collected from aspen (Populus tremuloides Michx) and birch (Betula papyrifera Marsh) communities in 2003 at Aspen Free-Air Carbon Dioxide Enrichment experiment in Rhinelander, WI. A 935 day in situ litter decomposition study was conducted. The results suggested that small changes in litter chemistry under elevated CO2 and O3 will occur, and combined with changes in litter biomass production could significantly alter the inputs of soluble sugars, condensed tannins, soluble phenolics, cellulose and lignin to forest soils. Elevated CO2 significantly increased the fluxes to soil of all nutrients (N, P, K, S, Mg, Ca, Cu, Mn, and Zn) and elevated O3 had the opposite effect. Atmospheric changes had little effect on nutrient release rates, except for decreasing Ca and B release under elevated CO2 and decreasing N and Ca release under elevated O3. Elevated CO2 significantly reduced litter mass loss (-10 %) in the first year, but increased litter mass loss (+46 %) in the second year. Elevated O3 reduced litter mass loss (-13 %) in the first year, and had no effect on mass loss in the second year. The mean residence time of birch/aspen litter (3.1 years) was significant lower than that of pure aspen (4.8 years). To examine how changes in litter biochemistry and production under elevated CO2 influence microbial activity and soil C formation, a 230-day microcosm incubation was conducted with five mass addition levels. The results indicate that small decreases in litter [N] under elevated CO2 had minor impacts on microbial C, microbial N and dissolved organic C. Increasing mass addition resulted in higher total C and new C accumulating in whole soil and mineral soil fractions, associated with higher cumulative C loss by respiration and greater breakdown of old C. Higher mass addition led to more total N retained in whole and mineral soil, but also greater C sequestration per unit N.
Enhancing Genetic Gain in Maize with Tropical Germplasm, QTL Mapping, and Spatial Methodologies
(2008-05-16) Jines, Michael P.; Major Goodman, Committee Chair; James B. Holland, Committee Member; Cavell Brownie, Committee Member; Paul Murphy, Committee Member
Advance-cycle breeding is restricting the germplasm base for U.S. maize (Zea Mays L.). Many breeding programs devote efforts to adapt diverse germplasm to U.S. growing conditions, but few are participating in continual enhancement. Incorporating tropical germplasm into U.S. breeding pools could broaden the maize germplasm base, while concomitantly providing favorable alleles for yield and disease resistance. Knowing the genomic regions, or quantitative trait loci (QTL), for disease resistance can enhance gain by permitting selection on marker genotypes in the absence of disease expression. In addition, accounting for spatial variability can improve the precision of experiments and aid breeders in line advancement decisions and QTL mapping. Recombinant inbred (RI) lines were derived from a cross between NC300, a temperate-adapted, all-tropical line, and B104, a Stiff-Stalk-synthetic line. The RI lines were topcrossed to the tester FR615.FR697 (a C103 sister line cross). Resistance QTL for Southern Rust (rust) (Puccinia polysora) were mapped in the topcrosses, while Gray Leaf Spot (GLS) (Cercospora zeae-maydis) QTL were mapped in both the RI lines and topcross populations. A major resistance gene for rust was identified on the short-arm of chromosome 10, while ten GLS QTL mapped to chromosomes 1, 2, 3, 4, 8, and 10. Similar markers on chromosome 1 and 8 flanked three GLS and flowering time QTL pairs, and the resistance alleles were associated with increased flowering time. No flowering time regions co-localized with rust-resistance loci. The major rust-resistance gene and three GLS QTL corresponded to regions mapped in prior populations. The tropical parental allele, NC300, increased resistance at three of these four loci. Extensively haplotyping germplasm at these four consensus regions could aid in forward breeding strategies to efficiently integrate resistance packages into U.S. maize breeding populations. Spatial analyses, such as trend and trend analysis with correlated errors models, can improve precision of genotype means estimates. These analyses often reduce the phenotypic variance among family means, and in doing so, increase the response to selection. A dynamic SAS program, entitled SPATIALPRO, was developed to implement spatial analytical techniques. The program constructs and optimizes several spatial models for each trait and single-environment-trial combination, and chooses a preferred model based on a specified criterion. Results from the preferred model are outputted into SAS data sets. A long term breeding effort was initiated in 1975 to adapt and subsequently enhance tropical germplasm. Founder germplasm included seven double-cross-tropical hybrids. Based on the poor per se performance of the first and second-cycle lines, at least five cycles of S1 recurrent selection (RS) for grain yield has been practiced on two populations derived from these lines. Cycles per se and cycle-topcrosses to LH132.LH51 were grown in separate yield trials to estimate responses to selection. In both instances, grain yield increased linearly across the cycles of selection for each population, but the yield responses across the cycle-topcrosses are approximately half the average annual gains of commercial breeding activities in the U.S. Corn Belt. To determine the current range in combining ability, ninety-six S1 families were sampled from the latest cycles of each population and topcrossed to LH132.LH51. Three topcross families did not differ significantly in yield from the commercial check hybrid average. Variance components estimated from the topcross families suggest that S1 topcross RS is more promising in maintaining relevancy, and appears to be a more favorable method of enhancement, as resources are devoted to families with superior combining ability.
Evaluation of Elite Exotic Maize Inbreds for Use in Long-term Temperate Breeding
(2007-05-10) Nelson, Paul Thomas; Major M. Goodman, Committee Chair; James B. Holland, Committee Member; Cavell Brownie, Committee Member
The U.S. maize (Zea mays L.) germplasm base is narrow. While maize is a very diverse species, that diversity is not represented in U.S. maize production acreage. Most elite U.S. maize inbreds can be traced back to a small pool of inbreds that were developed decades ago. Increased genetic diversity can be obtained through breeding with exotic germplasm, especially tropical-exotic sources. However, setbacks are often encountered when working with tropical germplasm due to adaptation barriers. Furthermore, the pool of available tropical germplasm is large and diverse, making choices of tropical parents difficult. The maize breeding program at North Carolina State University has begun a large-scale screening effort to evaluate elite exotic maize inbreds, most of which are tropical-exotic in origin. The purpose of this research was to: 1) generate comparative yield-trial data for over 100 elite exotic maize inbreds, 2) determine the relative effectiveness of various testcross regimes, 3) identify sources of gray leaf spot (GLS) resistance among these elite exotic inbreds, and 4) promote the use of exotic maize germplasm to broaden the genetic base of U.S. maize. Over 100 elite exotic maize inbreds were obtained from various international breeding programs. They were tested in replicated yield trials in North Carolina as 50%-exotic testcrosses by crossing them to a broad-base U.S. tester of Stiff Stalk (SS) x non-Stiff Stalk (NSS) origin. The more promising lines additionally entered 25%-tropical testcrosses with SS and NSS testers and were further evaluated in yield-trials. A dozen tropical inbred lines performed well overall—CML10, CML108, CML157Q, CML258, CML264, CML274, CML277, CML341, CML343, CML373, Tzi8, and Tzi9. Inbred lines CML157Q, CML343, CML373, and Tzi9 did not show significant line x tester interaction. Furthermore, it was determined that testcrossing to a single broad-based tester will suffice for initial screening purposes, allowing for elimination of the poorest performing lines. Testcrossing to additional SS and NSS testers may be of value when determining where the better performing materials will fit into a breeding program. It was further determined that most tropical lines can effectively be evaluated at the 50%-tropical level because many of the problems typically associated unadapted tropical material were minimized through a single testcross to an adapted tester. Each of the exotic lines was screened for GLS resistance either as inbreds per se, as testcrosses, or both. Many of the inbreds showed high levels of GLS resistance, including several lines that have good yield potential. These lines include CML108, CML258, CML274, CML277, CML343, and Tzi16. The results presented in this thesis provide temperate breeders with information on a sizable pool of potentially useful exotic maize inbred lines. These lines certainly deserve further attention in breeding efforts to broaden the U.S. maize germplasm base. Many are already being used at North Carolina State University in both exotic x temperate and exotic x exotic breeding crosses and populations.
Evaluation of Peanut (Arachis hypogaea L.) Germplasm for Resistance to Aflatoxin Production by Aspergillus flavus Link ex Fries
(2005-03-24) Xue, Huiqin; Tom Stalker, Committee Member; Thomas G. Isleib, Committee Chair; Gary A. Payne, Committee Member; Cavell Brownie, Committee Member
Aflatoxins are carcinogenic and toxic secondary metabolites produced primarily by the fungi Aspergillus flavus Link ex Fries and A. parasiticus Speare. Aflatoxin contamination of peanut is a serious worldwide problem. Aflatoxin resistant cultivars should be a component of an integrated program of aflatoxin management. Peanut genotypes with resistance to in vitro seed colonization (IVSCAF), field seed colonization (FSCAF) and preharvest aflatoxin contamination (PAC) have been reported, but no germplasm highly resistant to aflatoxin production has been found in cultivated peanut. A technique was developed to identify genotypes with resistance to aflatoxin production when subjected to post-harvest conditions conducive to fungal growth and aflatoxin synthesis. This technique provides environmental control and generally results in a low coefficient of variation in the data. Using this technique, the effect of the high-oleate trait on aflatoxin production was tested by comparing normal oleic lines with high-oleic backcross-derived lines. High-oleate peanuts supported more aflatoxin than normaloleate lines, but the magnitude of the difference varied with background genotype. To determine if linoleate concentration in seed oil could be used to predict levels of aflatoxin production, seeds of genotypes representing a range of linoleate concentration were tested. Low-linoleate lines consistently contained more aflatoxin, while normal- to highlinoleate lines contained variable amounts of aflatoxin. Although fatty acid profiles accounted for significant portions of the genetic variation, fatty acid concentration was not a reliable predictor of aflatoxin production, especially for lines in the normal range for oleate and linoleate. The same technique was used to evaluate seven accessions of A. cardenasii Krapov. and W.C. Gregory, 29 accessions of A. duranensis Krapov. and W.C. Gregory, and 17 interspecific tetraploid lines derived from A. cardenasii. The two diploid wild species averaged significantly less aflatoxin contamination than A. hypogaea checks, but were not different from each other. Arachis duranensis accessions PI 468319, PI 468200, and PI 262133, and A. cardenasii accessions PI 262141 and PI 475997 had very low levels of aflatoxin contamination and should be valuable sources of resistance to aflatoxin contamination. Of the interspecific tetraploid lines, only GP-NC WS 2 supported aflatoxin production not significantly different from resistant parent A. cardenasii GKP 10017. It appears to be a line with reduced capacity for aflatoxin accumulation. To identify germplasm with more than one type of resistance, lines previously reported with resistance to IVSCAF, FSCAF or PAC were tested. The results suggested that there were no strong correlations of IVSCAF, FSCAF or PAC resistance with aflatoxin production resistance, so it should be possible to combine high resistance to IVSCAF, FSCAF, or PAC with aflatoxin production in a single genotype. PI 590325, PI 590299, PI 290626, and PI 337409 accumulated the lowest amounts of aflatoxin among all A. hypogaea lines tested. These genotypes and susceptible check Perry were used to test interactions between peanut genotype and strain of Aspergillus species. Aspergillus species and strain had more effect on aflatoxin production than did peanut genotype. There were significant interactions between peanut genotypes and strains within species. These results suggest that the technique might be improved by using a mixture of several aflatoxigenic strains of A. flavus and A. parasiticus to identify genotypes with stable and low aflatoxin contamination. A population developed by crossing resistant genotype PI 290626 with susceptible cultivar Gregory was used to study inheritance of resistance to aflatoxin production. The primary form of genetic variance in this population was nonadditive, suggesting that selection within the population will be ineffective in early segregating generations.
Evaluation of Robinia Pseudoacacia L. as Browse for Meat Goat Production in the Southeastern USA
(2004-11-25) Snyder, Lori June Unruh; Larry D. King, Committee Member; James T. Green, Committee Member; Cavell Brownie, Committee Member; Jean Marie Luginbuhl, Committee Member; J. Paul Mueller, Committee Chair
Demand for goat meat in the southeastern USA is steadily increasing as a result of preferences exhibited by expanding ethnic communities. Feeding systems can be developed to take advantage of the natural preference of goats for browse. A field study was undertaken in Raleigh, NC to measure the effects of spacing (1.0 or 0.5 m) and coppice height (0.25 or 0.5 m) of a 5-year old stand of black locust (BL; Robinia pseudoacacia L.) on growth characteristics such as herbage mass (HM), canopy height (H) and width, number and size of main branches, above ground woody biomass, and root collar diameter. A second objective was through regression analysis to identify one or more of the previously mentioned characteristics as a rapid method to estimate HM. The third objective was to determine the relationship between growing-degree-days (GDD) and HM, H, herbage quality indicators (N, in vitro true dry matter disappearance, neutral and acid detergent fiber (NDF and ADF), cellulose, and 72% sulfuric acid lignin) and anti-quality indicators (Folin-reactive phenolics, condensed and hydrolyzable tannins) of BL. The final objective was to evaluate the N metabolism of goats fed BL foliage. Results indicated that coppicing BL trees at 0.5 m and planting at the widest spacing (1.0 m) produced the greatest plant growth. Average HM (2,600 kg ha-1) was observed for the highest coppice height (0.5 m). The character most closely related to HM was size of main branches. In 1999, a dry year, there was a significant relationship between GDD and NDF, ADF (r2 =0.90 and 0.80, respectively). In 2000, a wet year, GDD was a poor predictor of NDF and ADF. For 1999 and 2000, GDD was a poor predictor of BL tannin concentrations. From the conclusions of the N metabolism trial, goats consuming BL had lower digestibilities and higher content of N in the feces. Overall, BL contributes well to a silvopastoral system.
Long-term Impacts of Changing Land-use Practices on Water Quality and Phytoplankton Assemblages in the Neuse Estuary Ecosystem, North Carolina
(2007-11-13) Rothenberger, Megan Beth; Thomas Wentworth, Committee Member; Cavell Brownie, Committee Member; JoAnn Burkholder, Committee Chair; Dave DeMaster, Committee Member
The goal of this research was to build upon present understanding of the eutrophication process in the Neuse Estuary ecosystem by evaluating linkages among land use practices, nutrient concentrations and ratios, and phytoplankton assemblage composition. First, geographic Information Systems (GIS) analysis was used to characterize 26 sub-basins throughout the Neuse watershed for changes in land use over the past decade. GIS was also used in concert with multivariate statistics to synthesize and integrate ten years of land cover and water quality data into a conceptual model. Second, a continuous, decadal record of the phytoplankton in the mesohaline Neuse Estuary, in conjunction with synoptic measurement of environmental variables, provided a unique opportunity to evaluate responses of the phytoplankton assemblages to changing environmental conditions. Ordination techniques were used to investigate potential environmental predictors of phytoplankton community patterns through the process of eutrophication. Analyses indicated that over the past 10 years, total phosphorus concentrations were significantly higher during summer months in sub-watersheds with high densities of wastewater treatment plants (WWTPs) and confined swine feed operations. Nitrate concentrations were significantly higher during winter in sub-watersheds with high WWTP densities, and both inorganic and organic forms of nitrogen were significantly higher in sub-watersheds with greater agricultural land use. Ammonium concentrations were significantly higher after high-precipitation periods, but were not significantly correlated with the land-use parameters included in this study. In the Neuse Estuary, among several important findings, abundance of the potentially toxic, bloom-forming dinoflagellate Prorocentrum minimum was positively related to low water temperatures (winter⁄spring) and organic nitrogen and suspended solids concentrations. In addition, abundance of other potentially toxic flagellated algae such as the raphidophyte, Heterosigma akashiwo, has increased over the past decade, and H. akashiwo was found to be an "indicator species" for high ammonium concentrations (> 50 μg⁄L). Overall, the data indicate that wastewater discharges in the upper Neuse basin and intensive swine agriculture in the lower basin have been the highest contributors of nitrogen and phosphorus to receiving surface waters. In the estuary, increased nutrients, especially ammonium, are promoting increased abundance of several potentially toxic, bloom-forming phytoplankton species.
Resistance To Powdery Mildew In Wheat Germplasm With Different Resistance Sources
(2007-09-07) Miranda, Lilian; David Marshall, Committee Member; Steven Leath, Committee Member; J. Paul Murphy, Committee Chair; Cavell Brownie, Committee Member; James Holland, Committee Member
Spatial Modeling of Detection and Abundance from Count Surveys of Animal Populations
(2006-12-27) Webster, Raymond Anthony; Kenneth H. Pollock, Committee Chair; Sujit K. Ghosh, Committee Member; Cavell Brownie, Committee Member; Kevin Gross, Committee Member
When analyzing data from surveys of animal populations, it has been common in the past to ignore important factors such as variation in animal detection probabilities across space, and spatial dependence in animal density. We present a unified framework for modeling animal survey data collected at spatially replicated survey sites in the form of repeated counts, "removal" counts, or "capture" history counts, that simultaneously models spatial variation in density and variation in detection probabilities due to changes in covariates across the landscape. The models have a complex hierarchical structure that makes them suited to Bayesian analysis using Markov chain Monte Carlo (MCMC) algorithms. To ensure that these algorithms are computationally efficient, we use conditional autogressive (CAR) models for modeling spatial dependence. We apply our models to two examples of animal survey data. In the first, an intensive repeated count survey of juvenile Coho Salmon in McGarvey Creek, Northern California, we detected moderate spatial dependence in density, and models which account for spatial dependence produced more precise predictions at unsurveyed habitat units, and thus more precise estimates of total stream abundance, than models which assumed spatial independence. Through a small simulation study, we show that ignoring heterogeneity in detection probabilities can lead to significant underestimation of total abundance. However, inclusion of heterogeneity using a random effect in the detection component of the model can lead to problems in Bayesian MCMC modeling for typical survey designs, and for this reason we stress the importance of accounting for heterogeneity by incorporating covariates in modeling detection probability. In our second example, we consider a large survey of birds in the Great Smoky Mountains National Park. We fit models to the three types of survey data, repeated counts, "removal" counts, and "capture" history counts. Our methods lead to maps of predicted relative density which are an improvement over those that would follow from ignoring spatial dependence. Modeling shows that variation in detection probability can also affect inference, particularly when a species is relatively difficult to detect. Our work also highlights the importance of good survey design for bird species modeling. We point out that these types of bird survey data, particularly removal and capture-recapture counts (which require individual birds to be identified), are prone to errors in bird identification. Although we obtain similar results for all three types of survey data, which implies that the effect of identification errors may be small, the consequences of such errors in the data requires further investigation. Finally, we present parametric models for combined distance and capture-recapture survey data from both line and point transect surveys that allow for two types of animal movement: permanent avoidance or attraction to a transect, or temporary displacement of animals in the vicinity of a transect. The models have a simple form, with parameters that quantify the impact transects and observers have on local density. We combine these density models with logistic-linear models for detection probability using the likelihood framework of Borchers et al. (1998) for combined distance and capture-recapture data. This allows us to separately estimate the parameters of both the density and detection components of the model, which is not possible using the standard methods of distance sampling. Through a simulation study, we show that, provided sufficient animals are detected, the model parameters have little bias, and lead to improved estimates of density over a simple uniform density model, particularly for line transect surveys. Model selection by AIC generally chooses the correct density model. We apply our models to the Great Smoky Mountains bird survey data, and find some evidence of observer effects on local bird density.
Topics Involving the Gamma Distribution: the Normal Coefficient of Variation and Conditional Monte Carlo.
(2007-01-19) Boyer, Joseph Guenther; William Swallow, Committee Chair; Dennis Boos, Committee Member; Cavell Brownie, Committee Member; Thomas Gerig, Committee Member; Michael Boyette, Committee Member
A transformation of the sample coefficient of variation ($CV$) for normal data is shown to be nearly proportional to a $chiˆ2$ random variable. The associated density is applied to inference on the common $CV$ of $k$ populations, testing $CV$ homogeneity across populations, and confidence intervals for the ratio of two $CV$s. The resulting tests and confidence intervals are shown via theory and simulation to be valid and powerful. In other work on the coefficient of variation, a sample of scientific abstracts is used to characterize the values of the $CV$ encountered in practice, point estimation for a common $CV$ in normal populations is studied, and the literature on testing $CV$ homogeneity in normal populations is reviewed. There is very little literature on the problem of conducting inference in models for continuous data conditional on sufficient statistics for nuisance parameters. This thesis explores Monte Carlo approaches to conditional $p$-value calculation in such models, including Dirichlet data generation, importance sampling, Markov chain Monte Carlo, and a method related to fiducial inference. Importance sampling is used to create a conditional test of $CV$ homogeneity in normal populations using the $chi^2$ approximation mentioned above. A Markov chain Monte Carlo solution is given to the long-standing problem of testing the homogeneity of exponential populations subject to Type I censoring. Conditional Monte Carlo algorithms are also applied to testing for an effect of a factor in an experiment with exponential data, testing for a dispersion effect in a replicated experiment with normal data, and testing a null value of a coefficient in exponential regression with an inverse link; brief consideration is also given to the problem of testing the homogeneity of $k$ $gamma$ distributions.
Tuning Variable Selection Procedures and Score Tests for Dose Effect in the Presence of Non-Responders.
(2002-09-22) Luo, Xiaohui; Cavell Brownie, Committee Member; Frederick T. Corbin, Committee Member; Dennis D. Boos, Committee Chair; Leonard A. Stefanski, Committee Co-Chair; David A. Dickey, Committee Member
There are two topics in this dissertation. The first topic is variable selection in linear regression, and the second topic is hypothesis testing in a regression setting with nonresponders. In the literature, there are many variable selection methods for linear regression whose performance depends critically on the stopping rule. But it appears that many of the rules used in practice do not adequately adapt to each particular data set. Thus we propose a general approach based on adding additional noise to the response variable that allows us to "tune" the stopping rule so that the selection method is not too greedy or too parsimonious and results in choosing a good model. We focus on a forward selection method due to an interest in handling large numbers of explanatory variables. Because the method is analytically intractable, we study it by Monte Carlo methods, compare it with some other methods, and find that it works very well except that it underfits models with large number of active predictors. For a mixture model where both the logit of the response rate and the response mean are linear functions of the covariate (dose level), we propose new score test statistics for treatment effect. If the linear coefficient for the logit response rate is β and d is the linear coefficient for the mean, then score statistics are derived for H0: β = d = 0 versus H1,β: β≠0, d=0, H0 versus H1,d: d≠0, β=0, and H0 versus H1: β²+d²≠0, respectively. For H0 versus H1 we propose a 2-degree-of-freedom score statistic and also the maximum of the individual score statistics for H0 versus H1,β and H0 versus H1,d, respectively. Permutation critical values are used, and the tests are compared with the simple linear regression method. A simulation study shows that under most of the circumstances considered, the 2-degree-of-freedom statistic gives the best performance, while the simple linear regression is very sensitive to the response rate. The five methods are also applied to several real data sets, and the 2-degree-of-freedom score statistic provides satisfactory results.
Use of Natural Tags in Closed Population Capture-Recapture Studies: Modeling Misidentification
(2007-08-14) Yoshizaki, Jun; Nicholas M. Haddad, Committee Co-Chair; Cavell Brownie, Committee Member; Joseph E. Hightower, Committee Member; James D. Nichols, Committee Member; Kenneth H. Pollock, Committee Chair