Molecular Approaches for Analyzing Organismal and Environmental Interactions

Our planet is undergoing rapid change due to the expanding human population and climate change, which leads to extreme weather events and habitat loss. It is more important than ever to develop methods which can monitor the impact we are having on the biodiversity of our planet. To influence policy changes in wildlife and resource management practices we need to provide measurable evidence of how we are affecting animal health and fitness and the ecosystems needed for their survival. We also need to pool our resources and work in interdisciplinary teams to find common threads which can help preserve biodiversity and vital habitats. This dissertation showcases how improved molecular biology assays and data analysis approaches can help monitor the fitness of animal populations within changing ecosystems. Chapter 1 details the development of a universal telomere assay for vertebrates. Recent work has shown the utility of telomere assays in tracking animal health. Telomere lengths can predict extinction events in animal populations, life span, and fitness consequences of anthropogenic activity. Telomere length assays are an improvement over other methods of measuring animal stress, such as cortisol levels, since they are stable during capture and sampling of animals. This dissertation provides a telomere length assay which can be used for any vertebrate. The assay was developed using a quantitative polymerase chain reaction platform which requires low DNA input and is rapid. This dissertation also demonstrates how this assay improves on current telomere assays developed for mice and can be used in a vertebrate not previously assayed for telomere lengths, the American kestrel. This work has the potential to propel research in vertebrate systems forward as it alleviates the need to develop new reference primers for each species of interest. This improved assay has shown promise in studies in mouse cell line studies, American kestrels, golden eagles, five species of passerine birds, osprey, northern goshawks and bighorn sheep. Chapter 2 presents a machine learning analysis, using a topic model approach, to integrate big data from remote sensing, leaf area index surveys, metabolomics and metagenomics to analyze community composition in cross-disciplinary datasets. Topic models were applied to understand community organization across a range of distinct, but connected, biological scales within the sagebrush steppe. The sagebrush steppe is home to several threatened species, including the pygmy rabbit (Brachylagus idahoensis) and sage-grouse (Centrocercus urophasianus). It covers vast swaths of the western United States and is subject to habitat fragmentation and land use conversion for both farming and rangeland use. It is also threatened by increases in fire events which can dramatically alter the landscape. Restoration efforts have been hampered by a lack of resources and often by inadequate collaboration between stakeholders and scientists. This work brought together scientists from four disciplines: remote sensing, field ecology, metabolomics and metagenomics, to provide a framework for how studies can be designed and analyzed that integrate patterns of biodiversity from multiple scales, from the molecular to the landscape scale. A topic model approach was used which groups features (chemicals, bacterial and plant taxa, and light spectrum) into “communities” which in turn can be analyzed for their presence within individual samples and time points. Within the landscape, I found communities which contain encroaching plant species, such as juniper (Juniperus spp.) and cheatgrass (Bromus tectorum). Within plants, I found chemicals which are known toxins to herbivores. Within herbivores, I identified differences in bacterial taxonomical communities associated with changes in diet. This work will help to inform restoration efforts and provide a road map for designing interdisciplinary studies.