Award: OCE-1916870

Ecosystems can exhibit ‘tipping points’ whereby an environmental disturbance pushes an ecosystem into a degraded state from which it does not recover, even when the environment normalizes. This may have happened to valuable oyster reefs in Northwest Florida in 2012, when drought and low river flow allowed predators of oysters to flourish and consume nearly all the oysters. Despite subsequent years of normal rainfall and river flow, oysters have not recovered, suggesting the ecosystem may have crossed a tipping point. In this project investigators hypothesized that the strength and timing of Hurricane Michael may have provided a sufficient physical and salinity disturbance to shift the system back into its original, healthy state. Prior observations of hurricane effects supported this prediction, but that earlier research was unable to quantify the underlying ecological mechanisms. In this study, investigators collected water quality data to show that Hurricane Michael did lower the salinity of Apalachicola Bay in a manner that could have initiated system recovery by alleviating predation on juvenile oysters. Immediately after this disturbance, we repeated predator-exclosure experiments throughout the bay. In contrast to our older pre-hurricane experimental results, post-hurricane experiments showed a significant reduction in predation on juvenile oysters that had not been observed since the fishery collapsed in 2012. In the absence of predation, the number of juvenile oysters on oyster reefs increased significantly, suggesting a potential return shift to a healthy state. However, it remains unclear if this recovery will be sustained because Hurricane Michael did not re-expose previously buried reef substrate, which larval oysters require for settlement. A mathematical model of the oyster population based on the lab and field data suggests that pulses of low salinity (e.g., hurricanes) that increase recruitment and reduce predation can lead to increases in the oyster population. Additionally, if managers add additional shell to the reef (termed 'cultching') to provide more surface area for settling larval oysters, and those additions are timed to coincide with that increased recruitment, then the population increase can be substantially greater. However, the model suggests that this is not a system with clear alternate states. Rather, the long-term status of the oyster population depends on salinity conditions. Thus shell additions can provide a short-term improvement in the population, moving it away from low abundance, but that improvement can be reversed by subsequent higher-salinity conditions. Intellectual Merit: Identifying and predicting tipping points for ecosystems between healthy and degraded states remains a central goal in ecology. A new approach was recently proposed: replicated field observations and experiments are compared to model simulations in which the system has been pushed past the tipping point. While this approach has provided improved predictions in tests with historical datasets and highly controlled experiments, it has not been applied in the field to identify the mechanisms underlying the bi-directional shift of a natural system beyond a tipping point. This is mainly because observations, experiments, and models are typically not in place to do so at the time of a shift. Hurricane Michael provided a unique opportunity to address this knowledge gap. The observational and experimental data generated were combined with a mathematical model to determine whether there were different stable states in the system (healthy and degraded). The model suggests that salinity has a strong effect on the system but there are not necessarily stable states; a degraded reef can slowly recover if salinity conditions are favorable, but adding shell can hasten that recovery. Broader Impacts: This research can inform the restoration and management of the Apalachicola Bay oyster fishery, which crashed in 2012. Specifically, this research highlights how post-hurricane alterations to salinity may represent an optimal time period for enhancing the amount of substrate on oyster reefs. This project also identified the optimal amount of substrate to deploy per unit area (manuscript published in the journal Restoration Ecology). Given a limited supply of restoration substrate material and restoration funds, it is important to identify the optimal time and amount of substrate to deploy. To date, this project has produced a nearly eight year time series on oyster density and size, substrate availability, predation, and larval recruitment on oyster reefs in Apalachicola Bay, with associated meta-data deposits at the Biological and Chemical Oceanography and Data Management Office of the National Science Foundation (links). In coordination with the Apalachicola National Estuarine Research Reserve, meta-data were also deposited at the Statewide Ecosystem Assessment of Coastal and Aquatic Resources Data Discovery of the Florida Department of Environmental Protection (https://dev.seacar.waterinstitute.usf.edu/programs/details/5075). This project also supported the training of two graduate students and one postdoctoral scholar. This training was highly cross-disciplinary, with a ecology student learning mathematical modeling tools and an applied mathematician developing expertise in oyster ecology. Results of this work will continue to be disseminated in scholarly and public-facing outlets beyond the lifetime of the award. Last Modified: 08/25/2021 Submitted by: James W White