A revised Holocene coral sea-level database from the Florida reef tract, USA

• Main Authors: Anastasios Stathakopoulos, Bernhard M. Riegl, Lauren T. Toth
• Format: Article
• Language: English
• Published: PeerJ Inc. 2020-01-01
• Series: PeerJ
• Subjects: Holocene; Sea level; Coral reefs; Acropora palmata; Florida reef tract; Western atlantic
• Online Access: https://peerj.com/articles/8350.pdf

The coral reefs and mangrove habitats of the south Florida region have long been used in sea-level studies for the western Atlantic because of their broad geographic extent and composition of sea-level tracking biota. The data from this region have been used to support several very different Holocene sea-level reconstructions (SLRs) over the years. However, many of these SLRs did not incorporate all available coral-based data, in part because detailed characterizations necessary for inclusion into sea-level databases were lacking. Here, we present an updated database comprised of 303 coral samples from published sources that we extensively characterized for the first time. The data were carefully screened by evaluating and ranking the visual taphonomic characteristics of every dated sample within the database, which resulted in the identification of 134 high-quality coral samples for consideration as suitable sea-level indicators. We show that our database largely agrees with the most recent SLR for south Florida over the last ∼7,000 years; however, the early Holocene remains poorly characterized because there are few high-quality data spanning this period. Suggestions to refine future Holocene SLRs in the region are provided including filling spatial and temporal data gaps of coral samples, particularly from the early Holocene, as well as constructing a more robust peat database to better constrain sea-level variability during the middle to late Holocene. Our database and taphonomic-ranking protocol provide a framework for researchers to evaluate data-selection criteria depending on the robustness of their sea-level models.