Salt marsh ecology

Salt marshes are highly dynamic ecosystems that rely on multiple environmental and physical drivers that determine species distribution and soil property distribution. However, climate change and human interference are threatening the delicate ecosystem. One of the easiest ways to monitor marsh dynamics is through remote sensing. Traditional methods may not handle the large, non-parametric datasets well and often do not spatially determine areas of uncertainty. This dissertation research developed a framework to map marsh species and predict ground soil properties using multiple remote sensing data sources by integrating modern Object-based Image Analysis (OBIA), machine learning, data fusion, and band indices techniques. It also sought to determine areas of uncertainty in the final outputs and differences between different spectral resolutions. Five machine learning classifiers were examined including Support Vector Machine (SVM) and Random Forest (RF) to map marsh species. Overall results illustrated that RF and SVM typically performed best, especially when using hyperspectral data combined with DEM information. Seven regressors were assessed to map three different soil properties. Again, RF and SVM performed the best no matter the dataset used, or soil property mapped. Soil salinity had r as high as 0.93, soil moisture had r as high as 0.91, and soil organic an r as high as 0.74 when using hyperspectral data.

As sea levels continue to rise, the projected damage that will ensue presents a great challenge for conservation and management of coastal ecosystems in Florida. Since Juncus roemerianus is a common marsh plant throughout Florida with unique growing characteristics that make it a popular restoration plant, this study implemented a 20 week greenhouse split plot experiment to examine the effects of sea level rise on J. roemerianus and ultimately determine its tolerance ranges to salinity and inundation in a high nutrient environment. Overall, salinity level and the interaction effect of salinity level and water level had the greatest effects on measured growth parameters including average mature height, maximum height, density, basal area, root length, and biomass. An inverse relationship between increasing salinity and the measured growth variables was observed with the greatest growth and survivability in 0 ppt water, survivability and reduced growth in 20 ppt water, survivability and little growth in 30 ppt water, and nearly complete senesce in 40 ppt water. This was the first laboratory study to determine the effect of 40 ppt water on J. roemerianus. Elevated water levels resulted in higher growth variables in the 20 ppt, 30 ppt, and 40 ppt treatments while inundated water levels produced higher growth variables in the 0 ppt treatment despite previous research finding inundation to have completely adverse effects on J. roemerianus. It is likely that the high nutrient environment provided for this study is the cause for this anomaly. The results of this study have major implications for the future of coastal ecosystems that are dominated by stands of J. roemerianus in South Florida and can be used in conjunction with studies on bordering marsh plants to predict shifts in the ecosystems of Florida that are responding to sea level rise scenarios.

Plant interactions (e.g., competition, facilitation) are critical drivers in
community development and structure. The Stress Gradient Hypothesis (SGH)
provides a predictive framework for how plant species interactions vary inversely
across an environmental stress gradient, predicting that facilitation is stronger with
increasing levels of stress. The SGH has been supported in numerous ecosystems
and across a variety of stress gradients, but recent research has demonstrated
contradictory results. These discrepancies have led to SGH revisions that expand its
conceptual framework by incorporating additional factors, such as other stressor
types and variations in species life history strategies. In this dissertation, I examine
a further modification of the SGH by proposing and testing a Multiple Stress
Gradient Hypothesis (MSGH) that considers how plant interactions vary along a continuous gradient of two co-occurring stressors using mangrove and salt marsh
communities as a case study. In Chapter 1, I outline the predictive framework of a
MSGH, by creating a series of predictions of species interactions. The components
of the MSGH predict that stressors of similar types (e.g., resource and nonresource)
will have similar effects and be additive. On the other hand, varying
species life history strategies and life stages will lead to extremes of plant
interactions. In Chapter 2, I performed a series of experiments to test the various
components of the MSGH. In Chapter 3, I performed a large-scale observational
study to test whether multiple co-occurring stressors altered the cumulative effects
on plant interactions, and if these stressors should be grouped (e.g., resource and
non-resource, abiotic and biotic, etc.) to enhance predictability. From a series of
studies conducted herein, I concluded that co-occurring stressors are important
factors that control complex species interactions as shown in my MSGH modeling
approach. Further, future theories need to incorporate species-specific and stressor specific
grouping when modeling how species interactions shape communities.