Spatial Patterns in the Response of California’s Kelp Forests to Climate Variability and Extremes
Skip to main content
eScholarship
Open Access Publications from the University of California

UCLA

UCLA Electronic Theses and Dissertations bannerUCLA

Spatial Patterns in the Response of California’s Kelp Forests to Climate Variability and Extremes

No data is associated with this publication.
Abstract

Kelp forests, like many marine ecosystems, are being exposed to more frequent and intense disturbances. Marine heatwaves are major drivers of widespread and sustained kelp forest losses, as temperature directly affects the distribution and persistence of populations. Heatwave effects have been examined on large scales using satellite-derived data, but there is also variability in kelp responses on local to regional scales, driven by marine microclimates such as localized upwelling. Field and drone-based monitoring efforts have been employed along the California coastline to capture local dynamics, but these surveys only cover spatially discrete samples from selected kelp beds. Consequently, field and drone-based datasets represent one end of the spatial spectrum for observing kelp canopy response and recovery after disturbance events, while satellite observations represent the other end. Neither method alone can fully resolve both local and regional to global-scale dynamics.The development of CubeSat constellations has enabled a workaround for these trade-offs, with global imagery available near-daily at meter-scale. In my first chapter, I develop methods to create spatially continuous time series of California kelp population dynamics from high-resolution CubeSat data. In my second chapter, I employ these time series to highlight local-scale variability in California kelp populations that have been impacted by marine heatwave events, helping to better identify areas of concern and understand the drivers of heatwave-mediated loss and stability. In my third chapter, I scale these methods to effectively map kelp canopy across the state of California. I develop deep learning models for estimating kelp canopy presence and apply these models to a time series of CubeSat data. My findings contribute new perspectives for addressing and understanding kelp forest disturbance regimes. Mapping kelp populations with CubeSat constellation data allowed for the identification of areas of vulnerability, providing utility in prioritizing areas for strategic protection and restoration. Classifying sections of coastline that supported kelp forests amidst marine heatwave events showed the presence of refugia, where subpopulations have been able to persist in habitats that are buffered from disturbance. Further, mapping these populations across California will allow for the development of new indicators of ecosystem functioning and health.

Main Content

This item is under embargo until June 14, 2025.