Doctor of Philosophy (Ph.D.), Bowling Green State University, 2020, Biological Sciences

Many taxa suffer from habitat loss, spread of invasive species, and climate change; however, reptiles are especially vulnerable because they are constrained physiologically from their ectothermic nature in addition to global population declines. Like other taxa, reptilian basic ecology requirements are influenced by ecological neighborhoods, which shape the abundance of critical resources and their movement patterns. My goal was to better understand reptilian movement patterns across spatial and temporal scales to facilitate conservation efforts within Oak Openings Region (OOR), of northwestern Ohio and southeastern Michigan, using a combination of field surveys, remote sensing data and modeling. My research examined (1) climate change, (2) distribution patterns, (3) habitat use, and (4) movement patterns. At the regional scale, we found moderate increases in suitable habitat for box turtles within the future scenarios based on climatic suitability models. Individuals may be more displaced or vulnerable from temperature change during the driest quarter of the year. Dispersal is feasible; box turtles, based on tracking of individuals, can move large distances within their lifetime but the physical barriers, like roads, on the landscape may greatly hinder these movements. I modeled habitat suitability for a suite of reptiles based on occupancy data and climate, habitat, elevation, and structural features. Currently, suitable habitat was less than half the area within OOR and was more restricted for two species of concern. There is a need to examine the range of limitations, i.e., minimum and maximum models, when planning conservation efforts for a suite of species, especially emphasizing the protection of wet and dry forest. At the local scale, I found using radio telemetry that box turtles displayed typical average home range sizes; however, some were much larger than other studies. This is likely a difference in landscape heterogeneity where home range size increase (open full item for complete abstract)

Committee: Karen Root PhD (Advisor); Michael Decker PhD (Other); Kevin McCluney PhD (Committee Member); Helen Michaels PhD (Committee Member); Enrique Gomezdelcampo PhD (Committee Member) Subjects: Biology; Conservation; Ecology

Doctor of Philosophy, The Ohio State University, 2015, Mechanical Engineering

Cerebral palsy (CP) is a developmental disorder of movement and posture that occurs due to damage to the developing nervous system. The current standard of care for gait treatments includes stretching, night splinting, serial casting, physical therapy, medication, botulinum toxin injections and orthopedic surgeries. In some cases therapists may visit homes of patients to prescribe exercises but only for few hours a week. The equipment used for therapy is large and stationary requiring the patients to visit clinics. Motor learning therapies require repetitive voluntary movements and it may be difficult to achieve this during the limited outpatient hours available. Wearable sensors that trigger interactive feedback can be easily portable and used at home enabling repetitive movements as per the patients' schedule. They reduce personnel and equipment demand and require minimum clinical supervision. Auditory feedback has been helpful in gait retraining in past studies. A portable, musical device could have therapeutic effect by promoting gait retraining. Here, we describe a prototype of a heel-strike real-time feedback system that has been developed as part of the dissertation. It records the number of heel strikes during gait and also promotes them through real time auditory feedback. The prototype was tested with healthy children and children having CP. The accuracy in detecting heel strikes in case of healthy children was 97.8% (S.D. = 5.541) and when tested with children having CP it was 96.78% (S.D. = 5.243). The device is lightweight and does not affect the maximum force with which neurologically impaired children strike their feet during walking. The heel strike timing detected by the device was compared to the time at which heel strike was detected by a force plate. Based on statistical analysis, the difference between heel strike detected by a force plate and the heel-strike detected by the device was on an average 14 ms. In some trials, there was minimal d (open full item for complete abstract)

Committee: John Bolte PhD (Advisor); Necip Berme PhD (Committee Member); Manoj Srinivasan PhD (Committee Member); Blaine Lilly PhD (Committee Member); Lise Worthen-Chaudhari MFA, MS, CCRC (Committee Member) Subjects: Biomechanics; Dance; Engineering; Mechanical Engineering; Neurosciences; Rehabilitation; Therapy

Doctor of Philosophy, The Ohio State University, 2009, Civil Engineering

Damage caused by landslides exceeds $ 3 billion annually in the U.S and more than $ 10 billion each year worldwide, making losses attributed to landslides greater than any other natural disaster except hurricanes. Along with massive property loss, thousands people are killed and injured every year as the result of landslides. Potentially, much of this property damage and many of the injuries and deaths can be avoided with an operational landslide warning system. The goal of this research is to develop a wireless sensor network to predict the onset of landslides. The system will work by recording orientation changes from tiltmeters deployed on the surface of landslide prone slopes. Basic wired detection systems have been installed but, due to the high costs, monitoring systems can only cover a limited portion of a slope and requiring pre-existing knowledge of the most likely slide locations. Wireless landslide detection systems today have many problems limiting their practicality. Current limitations include subsurface sensor installation costs, high energy consumption and actual validation at the network level. In this research software having the capacity to interpret signals and generate failure alerts is being developed. To validate the above system, measured displacement data using wired extensometers from select sites are converted to tilt values and for the same sites, failure modes showing vector plots are generated using a numerical analysis program. These failure modes will be compared with various non critical movements. These comparisons related to surface movement patterns will provide essential characteristics for the stable landslide warning algorithm. For demonstration of this system, a slope with forty nodes consisting of eight columns and five rows is considered to be representative a typical hill slope. This demonstration shows how to implement the proposed algorithm based on a simple on and off sensor which will perform similarly to a tiltmeter. (open full item for complete abstract)

Committee: William Wolfe PhD (Advisor); Fabian Tan PhD (Committee Member); Tarunjit Butalia PhD (Committee Member) Subjects: Civil Engineering