Date of Award

Spring 2026

Abstract

Climate change is accelerating the frequency of extreme heat events in the world’s oceans. These heatwaves can pose critical challenges for marine organisms, especially ectotherm foundation species who have experienced increasingly frequent and severe mass mortality events (MMEs). Predicting when, where, and to what extent populations experience heat-induced mortality is a critical knowledge gap in our ability to anticipate, mitigate, and recover from catastrophic MMEs. Here, I test the application of physiological nonlinear thermal tolerance landscapes (TTLs) and dynamic tolerance algorithms to translate dynamic heatwave events into ectotherm mortality predictions. In chapter 2, I conducted a simulation exploration of how modeled organism mortality post-heatwave is mismatched from traditional climatological metrics of marine heatwave severity. Climatological metrics were sensitive to seasonal timing and variability of annual temperature, either under- or over-stating the severity of the event compared to post-event mortality severity. Further, I highlight the importance of considering thermal tolerance on both acute and chronic time scales, as described with TTLs. Relying on static, acute temperature thermal tolerance measurements to determine winners and losers in multi-species assemblages risks inaccurately projecting resilient populations under chronic heatwave conditions. In chapter 3, I conducted static thermal death-time experiments and built a novel dynamic tolerance algorithm that accommodates submerged and aerially exposed TTLs to predict intertidal mortality of the eastern oyster, Crassostrea virginica. Exposed oysters, like those at low tide, were able to persist at extremely hot temperatures for longer periods than oysters at the same water temperatures, but at lower stressful temperatures submerged oysters survived for longer time periods. Initial field-validation of these two TTLs to predict summer intertidal oyster mortality at two sites in Great Bay, NH showed low model skill (RMSE = 6.4, 10.37). However, after we accounted for potential acclimation of field oysters to warm and variable intertidal conditions through a simulation-based parameter calibration analysis, model performance improved markedly (RMSE = 0.93, 2.29). In chapter 4, I expand the application of TTLs to hindcast mass mortality of the gorgonian octocoral Paramuricea clavata in the western Mediterranean. Through this seascape-scale analysis, I obtained mortality predictions over 250,512 MHW events across 38 years and two depth strata. Through an elasticity analysis, I determined that small changes in MHW duration and magnitude have the largest impact on mortality during acute and chronic events, respectively. By integrating the density of observed MHWs into different mortality severity categories, I showed that less severe mortality events typically arise through increasing magnitude of short MHWs, with more extreme mortality events becoming primarily driven by increasing MHW duration. Finally, I validated the modelling approach against monitoring data at two locations in the western Mediterranean. Model skill varied between locations and sites within each location, suggesting complex interplay of choice of temperature forcing, microclimate, and acclimation plasticity in the spatial variation of survival. Rather than predicting climate risks to marine populations on the basis of temperature means and physiological thresholds, it will be imperative that biological prediction becomes more event-based and mechanistic under a climate future without historical analogues.

Document Type

Dissertation

First Advisor

Easton R. White

Second Advisor

Brittany M. Jellison

Third Advisor

Jennifer A. Dijkstra

Department or Program

Biological Sciences

Degree Name

Doctor of Philosophy

Download

Share

COinS