Title: Forecasting Natural Regeneration, Functional Properties, and Species Abundance in Semi-Arid Shrublands of the Great Basin
Program: Doctor of Philosophy in Ecology, Evolution, and Behavior
Advisor: Dr. Trevor Caughlin, Biological Sciences
Committee Members: Dr. Megan Cattau, College of Innovation and Design; Dr. David S. Pilliod, Ecology; and Dr. Rongsong Lui, Biological Sciences
Natural regeneration is the foundation of resilient ecosystems. Anthropogenic pressures across the globe have caused the decline of ecosystem diversity, distribution, and functional properties essential to global and local communities. An example of the Great Basin ecosystems in the Western United States demonstrates how foundational shrub species decline due to intensified wildfires and other anthropogenic pressures, undermining the resilience across ecological and social-ecological systems. Ecological restoration aims to assist ecosystems in recovery after wildfires. However, restoration efforts in the Great Basin have had limited success in great part due to highly unpredictable rates of natural regeneration and spatially variable ecosystem functions that underlie resilience. This work approaches the problem by developing forecasting capacity and understanding natural regeneration using satellite and airborne remote sensing technology. It further demonstrates how remote sensing, including unoccupied aerial systems (UAS), provides the means to improve the monitoring of extensive landscapes by capitalizing on population abundance estimates. Chapter 1 aimed to estimate the post-wildfire recovery of Artemisia spp., foundational shrub species in the region, using a decadal remote sensing data source at a relatively coarse resolution. Using a spatially explicit approach to find similarities between > 400 wildfires undergoing natural regeneration, we found that a simple population model can forecast the recovery trajectory without using local data. In other words, insights about natural regeneration are transferrable among ecologically similar landscapes. Chapter 2 focused on high resolution UAS imagery as a data source to understand the relationship between functional (resistance and recovery) and structural properties of shrubland communities. In establishing this relationship, structural properties based on 3D vegetation reconstruction depend on the spatial scale of the summary metric. We used scale decomposition of structural heterogeneity to show how resistance and recovery manifest scale-specific associations with structure. The results establish a scale-explicit dependence on key functional ecosystem properties and structural patterns. Chapter 3 extends the utility of aerial imagery as a source for population abundance estimates of Artemisia tridentata. The key challenge to estimating abundance from aerial imagery is the presence of observation errors that can significantly bias predictions and inference. Observation errors can include missed and false detections, which require different analytical solutions to avoid biased results. We use a hierarchical state-space model that accounts for both types of observation errors to evaluate the landscape drivers of observation errors and abundance. Applying state-space models to aerial imagery allowed the pursuit of predictive and inferential questions that may otherwise be impossible due to observation errors. Overall, this work advances our capacity to better forecast, understand, and monitor natural regeneration in the Great Basin. By expanding the scales of data acquisition, remotely sensed data can be as valuable in meeting management-oriented monitoring and predictive goals as they are in advancing ecological theory.