There are 12 species of bats that occur in Illinois; 5 of these species can be found hibernating in abandoned mines and caves in southern Illinois. Due to the destruction of their natural hibernacula, caves, many species of bats have found abandoned mines to be suitable replacement habitat. A complex of abandoned underground microcrystalline silica mines in southern Illinois owned by Unimin Specialty Minerals Corporation now provides hibernacula for 5 species of cavernicolous bats: the federally endangered Indiana bat (Myotis sodalis), little brown bat (M. lucifugus), eastern pipistrelle (Perimyotis subflavus), big brown bat (Eptesicus fuscus), and the northern long-eared bat (M. septentrionalis). Within the last 10 years the number of bats using these mines has increased dramatically, especially the Indiana bat which has increased from just over 9,000 to 43,000 hibernating in Magazine Mine. One concern of having so many endangered bats hibernating in one mine is stability. Mines were created relatively recently and are still in the process of settling. Thus, these mines might act as a potential sink, drawing in hibernating bats but potentially collapsing and killing them. Thirteen mines were surveyed for bats and for the amount of spalling that occurred over the 16-month study period from September 2006 to December 2007. Factors that could increase the amount of spalling were quantified, including temperature, moisture, and moisture variability in the material of the walls in the mines, and temperature variability. Number of hibernating bats in the mines was also documented. Data were analyzed with logistic regression. Temperature was a significant predictor of spalling (W2 = 12.76, p = 0.0004) when considered as a univariate variable, as was temperature variation (W2 = 21.89, p = <0.0001). Considering multiple logistic regression analyses, moisture was the best predictor. For the 13 mines surveyed, number of hibernating bats ranged from 0 to 3,755. Whereas all three variables were important at predicting the presence of bats, temperature variation (W2 = 35.98, p =<0.0001) was a better predictor than temperature or moisture. In a multiple logistic regression, temperature (W2 = 46.75, p = < 0.0001) and temperature variation (W2 = 20.56, p = < 0.0001) were better at predicting presence of bats then was moisture. The less variation in temperature the more likely that bats will be present. Because bats prefer stable temperatures and spalling occurs more often at high variability of temperatures and very low temperatures, bats were usually in areas that exhibited little or no spalling.

Understanding the health of rare and at-risk plant species is crucial for the conservation of biodiversity. Rare and at-risk species are often specialists with very specific requirements for growth and reproductive success. Some parameters associated with growth and reproductive success of rare aquatic species may include biotic and abiotic factors such as herbivory, competition, water depth, pH, and other factors associated with habitat quality. In this thesis, the population health of the rare and at-risk aquatic species Plantago cordata in Southern Illinois streams is assessed in relation to stream quality of sites containing Plantago cordata and sites without Plantago cordata.The objectives of this study were to (1) Determine if there is a relationship between population size structure and biotic habitat factors among populations and sites, and (2) Determine if there is a relationship between population size structure and abiotic habitat factors among populations and sites. Hypotheses were constructed using the hierarchy-of hypotheses method and most narrow alternative hypotheses are as follows: (1) There will be significant relationships between biotic and abiotic factors and Plantago cordata growth and reproductive success. (2) Levels of herbivory will have significant relationships with growth and reproductive success of Plantago cordata. (3) The abundance of co-occurring species individually and considered in functional groups (graminoid, herb, woody) will have significant relationships with growth and reproductive success of Plantago cordata. (4) Stream depth will have a unimodal relationship with growth and reproductive success of Plantago cordata, with maximum performance at an intermediate stream depth. (5) Stream pH will have a unimodal relationship with growth and reproductive success of Plantago cordata. (6) Stream quality (as outlined by the QHEI assessment) will have significant relationships with growth and reproductive success of Plantago cordata. General trends in herbivory suggest that Plantago cordata could be at greatest risk of herbivory during warmest average temperatures of the study season. There is evidence to suggest that flowering individuals could have a greater threat of herbivory than other growth stages based on Kendal correlation results. A Wilcox test determined that there were no statistical differences in species richness for plots with versus without Plantago cordata. There were relatively low values for percent cover of co-occurring species when considered in functional groups and there was overall relatively low species richness across all communities. Despite this, non-metric multidimensional scaling (NMDS) of community data provided evidence to suggest that there were statistical differences in the species composition of co-occurring individual riparian species among P. cordata populations. A significant negative correlation for water depth and flowering individuals and a significant positive correlation with water depth and pH occurred. A significant positive correlation for stream pH with the numbers of seedlings and dehiscing individuals was also found. One-way analyses of variance (ANOVAs) showed that some populations containing Plantago cordata were found to have stream pH values statistically different from nearby IEPA control streams without Plantago cordata. One population of Plantago cordata was found to be significantly different than two other populations in regards to stream pH using a linear mixed model. A principal components analysis (PCA) of stream Qualitative Habitat Evaluation Index (QHEI) assessments demonstrated distinct clustering of Plantago cordata streams and control IEPA streams without Plantago cordata. These findings provide partial evidence to suggest that growth and reproductive success of Plantago cordata could potentially be impacted by the abiotic and biotic parameters explored in this thesis.

Wood ducks are thought to depend on mature hardwood forests juxtaposed with palustrine wetlands but these habitats have been continuously degraded or destroyed since European settlement. Wood ducks are adaptable and the use of marginal habitats and nest boxes has extended their range and probably was important to the recovery of wood duck populations. Until now no study has analyzed the contribution of upland nesting to local population growth and maintenance. I investigated predation effects on nest site selection using wood duck nests and simulated nests placed in natural cavities. Using demographic data collected on wood ducks in southern Illinois, I created a population model to compare growth rates of population segments that nested in upland and floodplain habitats. During 1993-1998, 179 of 244 radiomarked hens remained on the study area as resident hens. One hundred-four nests were located by following radiomarked hens and 66% of nests were found in upland forests. Nests in the floodplain were initiated earlier than nests in the upland. Thirty-six percent of known nest cavities were used in subsequent years but <10% were used by the same hen. Nest success was greater in upland habitats (0.78 ± 0.10) than in floodplain habitats (0.54 ± 0.18). Hen survival through the nesting season was 0.80 ± 0.03 and did not differ between habitats or age classes. Sixty-five percent (n = 43) of simulated floodplain nests were destroyed compared to 33% (n = 45) in the upland. Logistic regression models of simulated nest data indicated cavity security could be important in the fragmented floodplain forests of Union County Conservation Area (UCCA). No physical characteristics of wood duck nest cavities differentiated successful and unsuccessful nesting attempts. The growth rate of the local population was positive and estimated to be about 3%. Lambda was most sensitive to upland nesting parameters and floodplain parameters appeared to have little impact on lambda. Hens nesting in the area appear to have adapted to predation pressure by nesting in more secure floodplain cavities at UCCA or by nesting in the upland habitats. Floodplain and upland habitats are ecologically intertwined and the local wood duck population would not survive if either habitat were destroyed or severely degraded.

Southern Illinois University at Carbondale (SIUC) is a site of linguistic diversity where speakers of three major dialects of American English--Northern, Midland, and Southern--are brought into contact with one another. The speech of undergraduates at SIUC is subject to the processes of dialect contact and accommodation; as a result, regional speech features are lost in favor of an overarching SIUC dialect norm or koiné. The linguistic contact that takes place at Southern Illinois University at Carbondale is unique. Previous studies of dialect contact involve situations created by migrations of large populations of settlers moving to a new area. These --migrants‖ settle permanently in the new area and become isolated from their original anchor dialects. The dialect mixture that arises from countless single instances of interpersonal accommodation will, under many circumstances, lead to koinéization or new dialect formation. However, the dialect contact situation at SIUC is different from these previous studies. First, the contact situation at SIUC is made up of fluid populations of highly mobile individuals--undergraduates. While the groups in contact remain consistent, individual students comprising the populations of these groups come and go every year. Additionally, rather than permanently relocating, the contact between the different groups at SIUC is interrupted by students leaving for three months of summer break and one month of winter break every year, thereby preventing speakers of the displaced dialects from becoming isolated from their original anchor dialects. The presence of these factors at SIUC provides a way to test and expand our existing models of language use and language attitudes in regards to dialect contact, accommodation, self- and group- categorization, and individual- and community-level notions of linguistic variation and language change. / text