Browsing by Author "Chapman, Phillip, committee member"

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    Open Access
    A penalized estimation procedure for varying coefficient models
    (Colorado State University. Libraries, 2015) Tu, Yan, author; Wang, Haonan, advisor; Breidt, F. Jay, committee member; Chapman, Phillip, committee member; Luo, J. Rockey, committee member
    Varying coefficient models are widely used for analyzing longitudinal data. Various methods for estimating coefficient functions have been developed over the years. We revisit the problem under the theme of functional sparsity. The problem of sparsity, including global sparsity and local sparsity, is a recurrent topic in nonparametric function estimation. A function has global sparsity if it is zero over the entire domain, and it indicates that the corresponding covariate is irrelevant to the response variable. A function has local sparsity if it is nonzero but remains zero for a set of intervals, and it identifies an inactive period of the corresponding covariate. Each type of sparsity has been addressed in the literature using the idea of regularization to improve estimation as well as interpretability. In this dissertation, a penalized estimation procedure has been developed to achieve functional sparsity, that is, simultaneously addressing both types of sparsity in a unified framework. We exploit the property of B-spline approximation and group bridge penalization. Our method is illustrated in simulation study and real data analysis, and outperforms the existing methods in identifying both local sparsity and global sparsity. Asymptotic properties of estimation consistency and sparsistency of the proposed method are established. The term of sparsistency refers to the property that the functional sparsity can be consistently detected.
  • Thumbnail Image
    Item
    Open Access
    Cross country comparison of genetic diversity by merging microsatellite data from pigs
    (Colorado State University. Libraries, 2014) Ibeka, Cyril O., author; Whittier, Jack, advisor; Blackburn, Harvey, advisor; Chapman, Phillip, committee member
    The need to develop a comprehensive method to extend genetic diversity to a global level has never being more important given the increase in the amount of genetic information being generated within various countries using different genotyping procedures, hence the need to merge datasets. Merging also enables a country to compare the genetic diversity of its animals to other countries. Many genetic diversity studies have been limited either due to the use of manual merging techniques, which ignores the effect of different laboratories with different laboratory protocols and are obviously time consuming or due to allele size inconsistencies in the case of automated merging processes. Our objective was to develop a method that extends genetic diversity globally using inexpensive methods based on Bayesian approach. Thirty microsatellite markers were originally genotyped from 220 Brazilian pigs and thirty-five microsatellite markers were genotyped from 179 pigs from the United States. Fourteen microsatellite markers were common between pigs in both countries. However, twelve microsatellite markers with posterior probability greater than 0.550 were successfully merged using a Bayesian cluster method. GENALEX and FSTAT results showed that the Brazilian pigs were significantly different from the Chinese (P< 0.01667) and US (P < 0.01667) pigs but were genetically closer to the US breeds (0.25) than the Chinese breeds (0.42). Population structure results obtained from STRUCTURE software showed peaks at K = 2, 4, 13 and 15. STRUCTURE result with K = 4 showed evidence of geographic differentiation of breeds into Brazil, China and the United States. STRUCTURE result at K = 9 revealed evidence of admixture within countries. The Chinese pigs showed evidence of genetic differentiation within the breeds. Many Brazilian pigs are not unique pigs but are admixtures developed by crossing local pigs with commercial pigs. The genetic diversity of the US swine population may need to be increased to prevent loss of biodiversity in the event of disease outbreaks or natural disaster.
  • Thumbnail Image
    Item
    Open Access
    Driving change: the 2011 National Beef Quality Audit
    (Colorado State University. Libraries, 2012) Igo, Jessica Leigh, author; Belk, Keith, advisor; Tatum, Daryl, committee member; Woerner, Dale, committee member; Chapman, Phillip, committee member
    The National Beef Quality Audit - 2011 evaluated the current status and progress being made towards quality and consistency of cattle, carcasses, and beef products produced by the U.S. fed beef population since the introduction of the National Beef Quality Audit in 1991. The objectives of this research were to determine how each beef market sector defines seven identified quality categories, estimates willingness to pay (WTP) for specified quality categories within each beef market sector, and establishes a best-worst (BW) scaling for the identified quality attributes. Face-to-face interviews were conducted using a modern, dynamic routing instrument over an 11-mo period (February to December 2011) with representatives of the following beef market sectors: Government and Allied Industries (n = 47); Feeders (n = 59); Packers (n = 26); Food Service, Distribution, and Further Processors (n = 48); and Retailers (n = 30). To accomplish the objectives, all responses were characterized using seven pre-established quality categories as the basis for asking interviewees the WTP and BW scaling questions. To determine WTP of the beef market sectors for U.S. fed beef, it was first important to understand what "quality" meant to each sector as it related to the U.S. fed beef products they purchase. To achieve this, "quality" was divided into seven pre-established categories: (1) How and where the cattle were raised, (2) Lean, fat, and bone, (3) Weight and size, (4) Cattle genetics, (5) Visual characteristics, (6) Food safety, and (7) Eating satisfaction, and interviewees in each beef market sector were asked to explain iii exactly which quality-related details/practices were important within each category. Overall, "Food safety" was the attribute of greatest importance to all beef market sectors except Feeders, who ranked "How and where the cattle were raised" as the most important. "Eating satisfaction" was the attribute of second most importance to all beef market sectors, except Feeders. Feeders ranked "Weight and size" as the second most important. Overall, "How and where the cattle were raised" had the greatest odds (0.25) of being considered a "non-negotiable requirement" before the raw material for each sector would be considered at all for purchase, and differed (P < 0.05) from "Visual characteristics" (0.14), "Lean, fat, and bone" (0.12), "Eating satisfaction" (0.12), "Cattle genetics" (0.10), and "Weight and size" (0.06). Of all market sectors combined, "Eating satisfaction" calculated the highest average percentage premium (11.1%), but only differed (P < 0.05) from "Weight and size" (8.8%). Most notably, when a sector said that "Food safety" was a "non-negotiable requirement," no sector was willing to purchase the product at a discounted price if the "Food safety" of the product could not be assured.
  • Thumbnail Image
    Item
    Open Access
    Influence of abiotic and biotic factors on the response of benthic macroinvertebrates to metals
    (Colorado State University. Libraries, 1995) Kiffney, Peter Michael, author; Clements, Will, advisor; Fausch, Kurt D., committee member; Kondratieff, Boris C., committee member; Chapman, Phillip, committee member
    Stream ecologists are well aware that chemical, biological, and physical characteristics of lotic systems vary spatially and temporally. With this in mind, I designed a series of experiments and field studies to examine the role of spatial variation in stream benthic macroinvertebrate communities in response to metals. Specifically, I tested the hypothesis that stream invertebrate communities from pristine streams of different size and altitude varied in their response to metals. To evaluate how metals affected biotic interactions, I manipulated invertebrate density, predation intensity, and metals in stream microcosms. Using stream invertebrate communities, I also designed an experiment and field survey to identify reliable bioindicators of metal contamination in western streams. Results from microcosm experiments and field studies showed that benthic invertebrate populations from high-altitude streams were more sensitive to the effects of metals than invertebrate populations from low-altitude streams. For example, Baetis sp. and Rhithrogena hageni from Little Beaver Creek (LBC), Colorado, (high-altitude stream) were significantly more sensitive to zinc than the same species from the South Fork of the Poudre River (SFP) (low-altitude stream) in stream microcosms. Results from field surveys showed that densities of most groups of aquatic insects (e.g., Ephemeroptera, Plecoptera, Trichoptera) were lower at high-altitude metal contaminated streams than those same groups at low-altitude streams. Other factors, such as variation in water temperature or nutrient concentrations between high- and low-altitude streams, could be lll responsible for these differences. However, because invertebrate responses were similar under controlled and field conditions, I hypothesize that differences in abundances between reference and contaminated locations was a result of metals. To determine if body size contributed to the variation in sensitivity of insects to metals, size measurements were made on species collected from LBC (high-altitude) and SFP (low-altitude). Measurements were also made on species from control and metal-treated stream microcosm. Most species were smaller at LBC (high-altitude stream) than the same species from SFP (low-altitude stream). For example, the mayfly Baetis tricaudatus and the caddisfly Arctopsyche grandis were significantly larger at SFP than LBC. In addition, insect body size was larger in metal-dosed microcosms than in controls. Brachycentrus sp., B. tricaudatus, R. hageni, Eohemerella infrequens, and P. badia were significantly larger in metal-treated microcosms than in controls. Logistic regression indicated survival in metal-dosed microcosms was less for small individuals than for larger individuals of the same species. These results suggest that some insect species from high-altitude streams were smaller than those from low-altitude streams, and that survival was greater for larger lifestages. Thus the variation in response of macroinvertebrates between different altitude streams observed in earlier studies may be due to differences in body size. The effects of low levels of metals (half the chronic levels of Cd, Cu, and Zn) on some species varied in relation to invertebrate density (low and high density) and invertebrate predation (no predators added and predators added). The abundance of Hydropsyche sp. was significantly lower in metal-dosed, high density treatments than in control, high density treatments. Moreover, the effects of an invertebrate predator (Hesperoperla pacifica) on Hydropsyche sp. was significantly greater in metal-dosed microcosms than in controls. These results suggest that metals interact with biotic factors to influence stream invertebrate community structure, and that effects occurred at metal concentrations lower than chronic criteria value. Toxicity experiments in stream microcosms showed that the abundance and species richness of aquatic insects were significantly reduced at 1x, 5x and 10x the United States Environmental Protection Agency chronic levels of cadmium, copper, and zinc (1x=1.1, 5.0, and 110 μg/L Cd, Cu, and Zn, respectively). Mayflies were the most sensitive group, as the abundance of Baetis sp. and Rithrogena hageni were significantly reduced in the Ix treatment. The response of Drunella grandis was size dependent, as small lifestages were significantly more sensitive than large lifestages. Stoneflies were also affected, but their response was more variable with abundances of some species (Pteronarcella badia) being reduced in the Ix treatment, whereas other species were unaffected (Sweltsa sp.). Heptageniid mayflies were consistently less abundant downstream of sources of metal contamination in the Arkansas and Eagle rivers, whereas the response of other measures were more variable. For instance, species richness and total density were greater at a metal-contaminated site on the Arkansas River compared with an upstream reference site. Therefore, results from this experiment and field survey suggest that changes in abundance of heptageniid mayflies may provide a reliable indicator of metal-contamination in western streams.
  • Thumbnail Image
    Item
    Open Access
    Sex determination using the first thoracic vertebra in 19th century American and ancient Nubian humans
    (Colorado State University. Libraries, 2010) Cundiff, Charles, author; Magennis, Ann, advisor; France, Diane L., advisor; LaBelle, Jason, committee member; Chapman, Phillip, committee member
    Many metric sex determination methods exist and have proven to be useful (Albanese 2003; Albanese et al. 2008; France 1998; Freiman et al. 2008; Frutos 2002; Frutos 2005; Gapert et al. 2009; Giles and Elliot 1963; Ozer and Katayama 2008; Ozer et al. 2006; Pastor 2005; Phenice 1969; Yu et al. 2008). Most metric sex determination methods rely on differences in stature and musculature between the sexes. The first thoracic vertebra is of interest because of its ease of identification and location at the boundary of many muscle groups. Linear measurements were taken on 161 T 1s from two osteological collections housed in Colorado. The first population, housed at CU Boulder, is derived from cemeteries excavated in Kulubnarti, Sudan. Burials in from this sample range from c.a. 550 AD to c.a. 1500 AD. The second population, housed at CSU Fort Collins, are remains from the cemetery of an asylum locate in Pueblo Colorado in the late 19th century. A linear function was used to determine the best classifying features of the T 1. From the 4 best classifiers (length of the transverse process, length of the spinous process, body diameter and coronal breadth of the vertebral foramen) a discriminant function was created for purposes of classification. Cross-validated results for the entire population give an accuracy of 86.76% for females and 89.25% accuracy for males. For the CSU (American) population 92.31% of females were classified correctly while 95.56% of males were correctly classified. For the CU (Nubian) population 92.59% of females were correctly classified as female with 85.71% being correctly classified male.