Sea surface temperatures and seagrass mortality in Florida Bay: Spatial and temporal patterns discerned from MODIS and AVHRR data

New publication by Daniel Frazier Carlson, Laura Yarbro, Sheila Scolaro, Michael Poniatowski, Vicki McGee-Absten and Paul Carlson.

2018.03.12 | Peter Schmidt Mikkelsen


Two major episodes of seagrass mortality have occurred in Florida Bay in the past 30 years: The first occurred between 1987 and 1991 and the second began in 2015. In both episodes, dense beds of turtle grass (Thalassia testudinum) were decimated. Elevated water temperature and hypersalinity have been implicated as contributing factors in both mortality events. During both die-off events seagrass beds on shallow (Z≈0.3 m) mudbanks in western Florida Bay were disproportionately affected when compared to neighboring seagrass in deeper basins (Z≈1.5 m). A network of in situ monitoring stations has measured sea surface temperature and salinity at hourly intervals in 16 basins in Florida Bay from August 2009 to the present day and these data aided in diagnosing the 2015–2016 die-off event. However, very few in situ observations of sea surface temperature have been collected on Florida Bay's shallow mudbanks as most are inaccessible by boat. As a result, our understanding of the role of elevated SST in shaping the spatial patterns of seagrass mortality is hindered by the spatial distribution of the in situ monitoring data and its record length. Therefore, we turn to remotely sensed SST data to expand our spatial coverage to include the entire Florida Bay ecosystem and to extend the record length to include the 1987–1991 die-off event. 1 km MODIS SST shows that shallow mudbanks were consistently warmer (by up to 6°C) than nearby deeper basins. While water depth is likely the primary driver of spatial variability in SST, Landsat-8 surface reflectance data suggest that shallow seagrass beds could have suffered from the added influence of low surface reflectance, which might have further contributed to their thermal stress. Daily 0.25° AVHRR SST shows that the August maximum SST has increased by 1°C from 1981 to 2016, which is a cause for concern for the future of seagrass in Florida Bay. Correlation of monthly AVHRR SST anomalies with the multivariate ENSO index shows that ENSO can only partially explain the anomalous temperatures. When viewed together, the in situ and remotely sensed SST suggests that both extended exposure to anomalously warm temperatures and large, rapid changes in temperature could have contributed to seagrass mortality during both events.

Remote Sensing of Environment, Bind 208, 23.02.2018, s. 171-188.

Arctic Research Centre