Browsing by Subject "Spatial Analysis"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    The Landscape of Farming: An Exploration of Spatial Bio-Economic Characterization Approaches
    (2015-11) Joglekar, Alison
    This dissertation consists of three inter-related but standalone papers focused on the theme of measuring the spatial, bio-economic attributes of production agriculture. Hitherto most of the agricultural development literature dealing with production agriculture has relied on data delineated in geopolitical (i.e., administrative district) boundaries of varying spatial resolutions, with some (increasingly of late) data reported for farm households. In some cases the household level data are geo-referenced, but in a majority of the studies the data are essentially aspatial. Many of the realities facing farmers however, including the agro-ecological (climate, soil, terrain and so on) attributes with which farmers have to work and their proximity to markets, are intrinsically spatial. Thus the location of farms and their physical and economic access to markets have a whole host of agricultural production and consumption implications that profoundly affect the economic circumstances of farm families. Spatially delineated data to facilitate analysis of the effects of location and its associated attributes on farm economies is still limited, but beginning to grow. This dissertation casts a critical eye over the nature and empirical plausibility of some key, spatially explicit datasets, including efforts to form spatially granular estimates of the location of crop production, area and yield worldwide; estimates of the proximity of African crop production to markets of varying sizes; and finally, the retail-level prices of key inputs (specifically fertilizer) faced by farmers throughout Tanzania.
  • Thumbnail Image
    Item
    Loss of Hospital Obstetric Services in the United States
    (2017-08) Hung, Yeiyin
    The loss of obstetric services has been a pressing concern in urban and rural areas nationwide. The first chapter used discrete-time multinomial logit model to identify urban-rural differences and factors associated with obstetric unit or hospital closures. During 2003-2013, 632 of 3,551 nationwide hospitals with obstetric services in 2002 stopped their obstetric services- through closures of hospitals or obstetric units. Loss of obstetric services has disproportionately affected areas served by low-volume hospitals, rural noncore, non-white and low-income communities, and counties with fewer family physicians, signaling a need to address maternity care access in these communities. The second chapter used generalized logistic regression models to examine associations of maternal residence rurality and insurance type with whether a woman lived in a zip code area where the nearest hospital obstetric unit and advanced neonatal care units were farther than 30 miles away in 2002 and 2013. The uneven geographic distribution of hospital obstetric care increased and varied by maternal residential rurality and insurance type. Women in rural noncore areas, with Medicaid or no insurance, and living in counties with lower income and educational attainment levels, had to travel farther to the nearest hospital with obstetric services or neonatal care than their counterparts. The third chapter employed propensity score matching to estimate the associations between service loss at the nearest hospital during 2003-2013 and outcomes, stratifying by maternal residence rurality and residential proximity to care prior to the services loss. Hospital obstetric service loss increased an average of 12 miles (10.6 minutes) to the nearest hospital obstetric unit for rural women. Service loss was not significantly associated with postpartum hemorrhage and prolonged lengths of stay for women who had access to hospital obstetric services within 30 miles, which accounted for nearly 98% of women with childbirth hospitalizations. Loss of nearby hospital obstetric services increased distances to care for rural women, and was also associated with a higher risk of non-indicated cesarean and postpartum hemorrhage for women living more than 30 miles from hospital-based obstetric services, highlighting the need to address efforts to improve maternal health in rural isolated populations
  • Thumbnail Image
    Item
    Spatial approaches to site formation and carnivore-hominin interaction at Dmanisi, Georgia
    (2016-06) Coil, Reed
    Spatial analyses are valuable tools for examining abiotic and biotic site formation processes that contribute to the accumulation of archaeological and paleontological material. When these analyses are used in tandem with taphonomic and geologic interpretations, archaeologists can make stronger arguments for site reconstruction and behavioral inferences. Spatial analyses have been used for decades to understand how human behaviors and the deposition of material are linked. Similarly, spatial behaviors by carnivores are apparent, with some species, namely hyenas, differentially using space for various behaviors. Through the lens of spatial analysis, this dissertation examines the degree to which abiotic and biotic agents contributed to site formation at the Early Homo site of Dmanisi, Georgia. Taphonomic and geologic studies at Dmanisi have interpreted the hominin-bearing deposits in Block 2 as resulting primarily from biotic agents, mainly carnivores, and not from fluvial or colluvial action (Lordkipanidze et al., 2006, 2007; Tappen et al., 2007). This dissertation expands on these interpretations by providing spatial evidence from two excavation areas (M6 and Block 2) and multiple strata used in conjunction with taphonomic analyses by Martha Tappen and geologic and archaeological analyses by Reid Ferring to compare relatively horizontal strata to the complex pipe/gully fill strata in which the highest concentration of bones are found. Orientation and dip, fragmentation and winnowing, and skeletal refits are examined and the results for each support the interpretation that the assemblages are not a product of fluvial and colluvial deposition. What is more, an analysis of spatial patterning indicates that higher densities of carnivores are located in the pipe/gully fill deposits than elsewhere. Also, coprolites are prevalent in these deposits, but follow a different spatial distribution than the skeletal material. Taphonomic analyses demonstrate that carnivores contributed greatly to the consumption of carcasses, and several species of carnivores are present at Dmanisi during this time. Each species of hyena has been observed to defecate in specific areas, both inside of the dens and outside in latrines, which could explain the cluster of coprolites in the B1 strata in Block 2. Two of the pipe/gully fill phases of B1 (B1x and B1y) are within a complex basalt formation that could have provided a secluded area for carnivores to consume carcasses and potentially have dens. These coprolites provide potential insight into carnivore space use in the Early Pleistocene, but further analysis is necessary to attribute the coprolites to specific animals. In any case, carnivores appear to be the main influence on site formation at Dmanisi, while the hominin imprint is minimal. In order to understand the potential carnivore-hominin interaction and the timing of carcass access at Dmanisi, this dissertation also provides new analytical methodology for studying bone fracture angles produced during the marrow acquisition process. In lieu of agent-specific surface modifications (i.e. carnivore tooth marks or hammerstone percussion marks), green breaks on long bones are difficult to attribute to a specific agent of breakage. By quantifying fracture angles, perhaps archaeologists can infer the causal agent of breakage and interpret the influence of carnivores and hominins on bone breakage. This dissertation contributes the results of a controlled breakage experiment where bones were broken by hyenas and hammerstones. Based on the results of the fracture angle analysis, different long bones create different fracture angle assemblages and it is necessary to identify the long bone fragments to at least limb portion (upper, middle, lower) if not to skeletal element. In addition, the preliminary results of the hyena created assemblage shows that hyena broken bones result in fracture angles further from 90° than hammerstone broken bones. Carnivores played an important role in the accumulation of faunal material at Dmanisi, Georgia. Evidence for abiotic site formation processes, such as fluvial and colluvial deposition, is minimal and likely did not affect the original depositional context of the stones and faunal material. The hominin contribution to site formation is also minimal, but perhaps understanding fracture angles created during the marrow acquisition process can allow for further testing of this hypothesis in the context of Early Homo. By examining these different factors, this dissertation broadens the knowledge of how carnivores and hominins contributed to site formation during the earliest expansion of our genus outside of Africa.
  • Thumbnail Image
    Item
    Spatial vegetation-environment relationships and distributional changes in the presettlement Minnesota prairie-forest boundary.
    (2009-02) Danz, Nicholas P
    The prairie-forest boundary region in Minnesota spans 650 km along a northwest-southeast axis and is often considered one of the most abrupt grassland-forest transitions in the world. Historically, this region separated tallgrass prairie vegetation to the south and west from forest vegetation to the north and east, while land conversion since presettlement has eliminated over 95% of original prairie and continues to convert and fragment forested lands. Ecological boundaries such as the prairie-forest transition are considered critical landscape elements that control the fluxes of organisms, materials, and energy between ecosystems. While the significance and characteristics of ecological boundaries has been often discussed in scientific literature, there are few studies that provide empirical support for boundary concepts. In particular, studies are lacking that evaluate vegetation-environment relationships across boundaries. In this thesis, I use the presettlement prairie-forest boundary as an example of an ecological boundary to address the following issues: 1) the influence of spatial scale and spatial heterogeneity in the controls of vegetation, 2) the spatial nature of a vegetation-climate relationship across the boundary, and 3) range distributional shifts since presettlement in tree taxa common to the boundary region.