Graduation Year

2025

Document Type

Thesis

Degree

M.S.

Degree Name

Master of Science (M.S.)

Degree Granting Department

Marine Science

Major Professor

Frank Muller Karger, Ph.D.

Committee Member

Ana Carolina Peraltova, Ph.D.

Committee Member

Marcus Beck, Ph.D.

Keywords

Google Earth Engine, Habitat monitoring, remote sensing, Seagrass extent, Turbidity, US Virgin Islands Salt Bay

Abstract

Seagrass beds provide important services in many coastal ecosystems. They provide habitat and food for a wide range of marine species, stabilize sediments, and prevent coastal erosion. These services support local fisheries and tourism-driven economies. Yet, seagrass ecosystems are threatened by changes in water quality, sedimentation, and other stressors. This study investigated the interannual variability in seagrass extent in Salt River Bay, United States Virgin Islands (USVI), from 2018 to 2024. The goals were to better understand temporal dynamics in seagrass extent and evaluate the influence of turbidity as a driver of seagrass health and distribution. The overarching research question guiding this work was: Was there interannual variability in seagrass extent in Salt River Bay between 2018 and 2024, and is this variability related to changes in water turbidity? The central hypothesis is that temporal changes in seagrass extent are strongly influenced by variability in turbidity, which constrains light availability and limits seagrass growth. To address this, the objectives were to: (1) quantify seagrass extent and turbidity using a time series of Sentinel-2 satellite images, (2) evaluate interannual variation and possible trends in seagrass and turbidity, and (3) assess potential relationships between seagrass change and turbidity. Understanding this relationship helps managers know when and where seagrass is most at risk from turbidity, so they can better plan protection, and restoration. The approach was to select high-quality atmospherically corrected Surface Reflectance Sentinel-2 imagery between 2018-2024. Images collected between June and August every year were excluded from the analysis, as Saharan dust reduced visibility and altered spectral reflectance. Normalized Difference Water Index (NDWI) and a Support Vector Machine (SVM) algorithm were applied to identify pixels that contained seagrass, sand, hardbottom, and deep-water, generating annual maps. The analysis focused on the whole region of the Salt River Bay (-64.76° to -64.75° W and 17.78° to 17.79° N), and in more detail on four quadrants of that region: Northwest (NW), Southwest (SW), Northeast (NE), and Southeast (SE). Seagrass areal extent was estimated by summation of the area of pixels identified as containing seagrass in these areas and converting the area to hectares (ha). Water turbidity was also estimated using the same Sentinel-2 product, using the red band (665nm, band 4) as a proxy for turbidity, as red reflectance is sensitive to sediments suspended near the surface of the water.

The results show that total seagrass extent declined in Salt River Bay from 2018 to 2024 at a rate of -2.57 ha per year (linear regression r = -0.78, R2 = 0.60, two-tailed p = 0.04, n = 7). The NE quadrant showed the steepest decline at over -1.2 ha per year but low significance (r = -0.66, R² = 0.43, two-tailed p = 0.11). The SE quadrant exhibited a declining trend at -0.92 ha per year, approaching statistical significance (r = -0.74, R² = 0.54, two-tailed p = 0.06). Weaker or negligible trends were observed in the NW and SW quadrants. Considering the whole region, the overall seagrass extent in Salt River Bay declined from 90.8 ha in 2018 to 77.3 ha in 2024, representing a 13.5 ha reduction (14.9% decrease) over the study period.

Turbidity, measured at the mouth of the bay (-64.76° to -64.75° W and 17.78° to 17.79° N), varied seasonally and interannually over the study period. Turbidity was averaged annually to compare it with the seagrass annual decreasing trend. Although annual turbidity showed an increasing trend with time (2019-2024, not enough data for an annual mean on 2018), this was not significant (r = 0.50, R² = 0.25, two-tailed p = 0.31, n = 6). There was an inverse correlation between the annual seagrass extent and annual turbidity averages (r =-0.64, R² = 0.41, two-tailed p = 0.172, n = 6). Although that correlation was non-significant, the correlation together with the visual inspection of the interannual variation of seagrass and turbidity suggest that fluctuations in water clarity influence seagrass dynamics in Salt River Bay, indicating that periods of elevated turbidity often corresponded with reduced seagrass extent and vice versa.

This research aligns with the objectives of the Blue-Green Action Platform (BlueGAP, 2024), which aims to promote sustainable management and conservation of marine and coastal resources by monitoring watershed sources of pollution. By integrating seagrass monitoring into a broader framework for assessing water quality, this study gives policymakers and conservation groups clearer information on when and where seagrass is most at risk from turbidity, helping guide restoration and management efforts. Future research could include integrating seagrass density and extent data into regional monitoring programs using in situ surveys, evaluating water runoff volume, suspended sediment load, and nutrient content from land-based sources to help manage water quality to promote seagrass growth.

Download

Share

COinS