Grasses are a diverse group of plants that play a significant role in many terrestrial ecosystems globally. Despite their importance, very little is known about where and when grasses originated. In particular, it remains unclear in which environment early grasses evolved. Current hypotheses, based primarily on phylogenetic work, suggest that early grasses emerged in closed habitats, such as forests, or in more open habitats along forest margins. However, there is little direct paleobotanical evidence to support either option. To understand the environmental context grasses evolved in, I will be reconstructing canopy openness using phytoliths from Argentinian fossil sites. Phytoliths are microscopic silica bodies deposited in or around plant cells, which can be preserved in the fossil record and used to reconstruct past vegetation. The phytoliths that I am analyzing are from the Las Violetas Formation (57.9-50.6 Ma) of Argentine Patagonia, a rock formation within an area known to hold the oldest records of grass phytoliths in South America. As a proxy for vegetation structure, I am using reconstructed Leaf Area Index (rLAI), which takes the area to perimeter ratio of non-grass phytoliths originated in the leaf epidermis and determines a value that corresponds to how much light passes through a canopy. A canopy with a high leaf coverage correlates to a high rLAI value and vice versa. Based on preliminary rLAI results, I expect to find that ancestral grasses lived in habitats with semi-open canopies, similar to modern shrublands. Investigating the ancestral environment of grasses can help us better understand their evolutionary history and potential environmental drivers that led to their success in terrestrial ecosystems. This information can help us gain insight on the vulnerability of grasses and grass-dominated habitats to environmental changes in the past, present and future.

Trait-based plant ecology can serve as a means to better understand shifts in ecological strategies within plant communities and how that affects greater ecosystem processes, such as productivity, across a period of major climate change. The most recent major global warming event prior to modern anthropogenic influences was the Miocene Climatic Optimum (MCO) ca. 17-14 million years ago. This event was a short aberration to a long-term cooling trend of the last 53 million years, with global temperature averages up to 8°C warmer than preindustrial averages. Changing climate conditions during the MCO may have led to plant community reshuffling, with many ecosystems possibly restabilizing with notably different optimal trait distributions. Functional traits such as leaf shape, size, and toothedness can be indicative of a plant’s ecological strategy. These leaf morphology variables have been shown to closely correlate with climate, highlighting their role in plant function and strategy, and can thus be used to statistically analyze community diversity across the MCO. I hypothesize that the MCO caused an overall increase in functional trait diversity through an increase in favorable environments, allowing plant reshuffling or migration of plants with new ecological strategies into existing communities. We expect to see this through trait distributions in a community diversifying as the MCO progresses. This study uses leaf functional trait data measured digitally from a range of Miocene fossil sites to assess trends and variances from before, during, and after the MCO. Statistical analysis will make use of a previously developed R package to assess functional diversity. These results will be crucial information in understanding the ecological response to today’s far more rapid climate change, as well as humanity’s response to the possible need for human-assisted plant community reshuffling by providing an example of how global warming affected vegetation in Earth’s past.

Eelgrass (Zostera marina) is an ecologically important species of flowering marine plant that inhabits sublittoral sediments in the northern hemisphere. Eelgrass beds provide critical habitat for many species of invertebrates, birds, and larval fish. As a true plant with roots, leaves and a vascular system, eelgrass is capable of extracting heavy metals from its environment, potentially making it a sentinel accumulator for heavy metal contamination and sourcing for ecological monitoring. Previous studies of eelgrass beds found that metal concentrations were higher in eelgrass tissues than sediment; however, it is uncertain whether the source of the Pb is the sediment, the water, or both. To determine metal sources and accumulation in eelgrass, we used the natural and anthropogenic variability of lead (Pb) isotopes to fingerprint the source of Pb in bottom sediments and in eelgrass of the Puget Sound. We collected eelgrass and sediment from five sites in Possession Sound, WA between 2018-2019. After purification of Pb from the samples, we analyzed Pb isotope ratios by multi-collector ICP mass spectrometry in the Dept. of Earth & Space Sciences. Significant variation was observed between sites – Hat Island and Whidbey sediments were undifferentiated but Mount Baker Terminal (MBT) sediments have different Pb isotope ratios. Eelgrass from the MBT site also has Pb isotope ratios distinct from other eelgrass and sediment samples, indicating a higher proportion of anthropogenic lead. Site MBT has more pollutant input from the city of Everett, whereas the Hat and Whidbey sites, which are adjacent to uninhabited land, may have more Snohomish River influence. Paired sediment and eelgrass samples from the same sites did not have the same lead isotope ratios, implying that eelgrass accesses multiple sources of Pb during growth. Testing this observation at other sites is important to understanding metal cycling in sublittoral environments.

The geologic record has become an increasingly important source of information for scientists to observe how plant communities of the past have responded to disturbance. Currently, there is a limited ability to recognize disturbance as a primary driver of plant community change, as there is limited evidence of how functional traits – plant traits that relate directly with plant function and ecological strategy – that can be measured in fossil leaves vary across succession. In this study I will measure a functional trait to help better identify disturbance in the fossil record, the carbon stable isotopic composition (δ13C) of bulk organic matter in leaves sampled across a successional gradient following a disturbance. This trait is often preserved during leaf fossilization and is representative of a plant's water use efficiency (WUE), or the amount of carbon dioxide used by the plant during photosynthesis for a given amount of water that is lost during transpiration. It is currently not known the extent to which carbon isotopes measured at the community-scale reflect the successional stage of a plant community. In an effort to develop this tool, I hypothesize that the WUE of plant species within a community will become more conservative in later successional stages. In support of this hypothesis, I predict that the abundance-weighted community average of leaf δ13Cwill increase through succession. In addition, I hypothesize that δ13C as a proxy for WUE will be most confounded in early succession, before a tree canopy forms, due to seedling utilizing water resources more rapidly without having established root systems and thus predict a higher variance of δ13Cvalues in this earliest stage of succession (Cernusak 2020). This research will help develop a method of identifying disturbances within geologic records which can give guidance on management decisions regarding modern ecosystems.

Control of fungal pathogens remains a significant challenge facing agriculture, causing plant death and post-harvest spoilage. Existing methods for fungal control, such as chemical fungicides or transgenic crops, are often non-specific or come with environmental concerns. Prior research indicates that some bacterial symbionts, naturally occurring in plants, can increase resistance to pathogens. In the Sharon L. Doty Plant Microbiology lab, I explored the use of endophyte bacteria to inhibit the growth of three fungal pathogens common among fruit and fruit trees. I investigated fungal growth inhibition through the secretion of volatile chemicals without direct contact. To do this, I grew putative antifungal bacteria together with the fungal pathogens, so that the two cultures were not in direct contact but shared airspace. I identified six strains that were highly successful against multiple fungal pathogens. I cultured these strains to collect any volatiles produced for gas chromatography-mass spectrometry analysis. Several of these strains produce similar volatiles; although the exact structure is not known, the chemical formula is the same among them. In addition, some of these strains produce carbon dioxide, which may be responsible for the inhibitory effect. I also used various bioinformatics pipelines to identify genes responsible for the biosynthesis of these inhibitory volatiles, allowing for deeper understanding of how these strains are able to inhibit fungal growth. These strains have strong potential for use as biocontrol agents. This use would allow for greater pathogen resistance without many of the drawbacks associated with chemical fertilizers or genetically modified organisms.

Trees are sessile organisms and require effective root architecture for water and nutrient uptake. Despite their importance, research in plant root systems is limited in comparison to shoots as roots are difficult to access. Rhizotrons, flat plexiglass pots, enable direct imaging of roots, but are restricted in their capabilities since plant material is started from seed with a limited growth period. These restrictions also limit the ability to model interactions between stress and tree root architecture. Effective experimental systems and accurate biological models are needed to predict how trees will respond to a rapidly changing environment. Here, I demonstrate the ability of a rhizotron using vegetatively propagated material and extended growth times to phenotype the effects of nutrient stress on root architecture in the model tree species, Populus trichocarpa. I then show how Populus traits generated from this system can parameterize a 3D functional-structural plant model (FSPM) for roots called CropRootBox. Populus cuttings were transplanted to rhizotron pots containing high contrast soil media and experienced a nutrient stress treatment. Images collected throughout the growth period were analyzed with RhizoVision and ImageJ to extract root traits including root depth, angle, branching frequency, and lateral root density. These traits advised Populus’ addition to CropRootBox, incorporating the response to nutrient stress. Using this platform, I expected nutrient deficit to decrease primary root elongation and increase lateral root elongation, branching, and density. I also expected these results to improve CropRootBox’s predictions of Populus root responses to nutrient stress. This accessible phenotyping platform accelerated my study on root architecture. CropRootBox showcases modeling’s ability to extend the derived functionality beyond just early growth nutrient stress. In combination, these tools have great potential to aid our understanding of trees and their rhizosphere.