Understanding complex and dynamic inter-specific relationships is key for informing and developing conservation policies. Accurately identifying the diet of various predators across Washington can provide insight into these relationships in terms of resource needs and predator-prey dynamics. DNA metabarcoding studies present a comprehensive way to surmise the complete diet profile of the American black bear (Ursus americanus), an opportunistically omnivorous predator that provides several important ecosystem services. As one of the only large omnivores in Washington, they also require large home ranges, making them an umbrella species that can be used as a guideline for overall conservation efforts. Our preliminary metabarcoding study with scat comparing the American black bear’s diet in northeastern and central Washington using a mitochondria-specific marker suggested their diet could be greatly influenced by human activity and that diet preference could delineate along geographic and ecological means. As the samples containing atypical food sources were within proximity to anthropogenic structures, these results suggest the pervasiveness of human-based food source availability to wildlife. These results were corroborated in our comprehensive study using both mitochondria and chloroplast specific markers. Analysis of the herbivorous portion identified the native and introduced plant species and their frequency of occurrence. This allowed us to understand the American black bear’s reliance on specific native species, as well as the extent of how local bear populations have incorporated non-native plant species into their diet, thus identifying their ecosystem intrusion. Additionally, the study’s sample collection spans a multiyear period, providing information on the seasonality of the American black bear’s diet, giving insight into seasonal prey switching and prey partitioning. This total diet profile allows for the examination of the influence of the wildlife-urban interface on food availability and resource selection in the Washington American black bear population, increasing our understanding of Washington predator dynamics.

After an 80-year absence, gray wolves have been returning to Washington state over the last decade. Mapping their population growth and reproductive activity is key to understanding their recovery and to assisting wildlife conservation management. Recognizing an established pack requires identification of the presence of breeding females. Accurate noninvasive identification of pregnant wolves through scat could greatly assist such efforts. A reliable index of pregnancy in most mammals is progesterone levels (ng/g) in feces; progesterone rises post-ovulation, but will only be elevated above a “pregnancy-threshold” level among pregnant females. However, this metric is less definitive in canids. Progesterone levels remain elevated in all post-ovulatory females, regardless of whether the females become pregnant or not. Since gut microbiome diversity has also been shown to differ between pregnant and non-pregnant large mammals, in this study I examined whether the combination of progesterone levels and gut microbiome diversity can refine this pregnancy metric in free-ranging wolves. Fecal samples from female wolves (n=62) with known progesterone levels were provided by the Center for Conservation Biology from a 2015-2017 study in Northeast Washington. Samples with progesterone levels above 2000 ng/g were defined to be from potentially pregnant females. I generated gut microbiome profiles by amplifying and sequencing the V4 16S rRNA gene region in each sample. I analyzed the sequences using Qiime 2 and R with the Silva reference database for microbial taxonomy classification. Gut microbiome profiles were compared to progesterone levels. Pregnant wolves are expected to have a different phyla diversity in bacterial communities from non-pregnant wolves. Results may guide future studies to focus on identifying a particular bacteria species or a certain proportion of diversity combined with progesterone levels to further refine pregnancy diagnosis. Accurate identification of pregnancies in wolf packs can improve population growth estimates, leading to informed Washington wildlife conservation policies.

Astronomy of the 21st century is driven by large data sets collected by large automated sky surveys. The largest survey project currently being built is the Legacy Survey of Space and Time (LSST), which will be a 10-year survey of the southern sky, expected to discover 5.5 million small bodies in our Solar System. The greater scientific community needs to know what the research potential and scope of the data LSST will collect. I am building a database -- accessible at http://ls.st/ssdb -- of simulated LSST observations of asteroids in our Solar System. My work delves into simulation accuracy, big data analysis, and database management. This dataset consists of individual observations, an orbit catalog, and a catalog of physical and observational characteristics. Using simulations from the University of Washington’s Data Intensive Research in Astrophysics and Cosmology Institute (DiRAC) and scripting in Python, I am attempting to accurately present this data and areas of possible research before LSST becomes operational. This project will be integral to preparing for research projects that will analyze actual LSST data.