Author ORCID Identifier

Date of Award

Winter 2023

Document Type

Thesis (Ph.D.)

Department or Program

Ecology, Evolution, Environment and Society

First Advisor

Kathryn L. Cottingham


Temperature is a nearly ubiquitous driver of ecological and evolutionary processes, especially for organisms whose body temperature changes with the environment. The overarching biological importance of environmental temperature and increased availability of temperature datasets have led to its inclusion in spatially continuous models of species’ performance, abundance, and distribution. However, variation in both environmental temperature and animal behavior at finer scales than captured by these models may reduce the accuracy of their scientific conclusions and conservation recommendations. Here I contribute to an improved understanding of thermal variability in river systems and its potential biological consequences for an iconic coldwater species, brook trout (Salvelinus fontinalis). First, I compared the accuracy of two regional stream temperature models currently used for conservation planning to my newly developed community science-based model using observational data from >200 sites in New Jersey’s Brook Trout Conservation Zone. Both regional models explained little variation in observed temperature, likely because of local effects of small dams and groundwater discharge. By contrast, my model considerably increased explanatory power and may be a useful, widely implementable approach to produce relatively accurate environmental temperature maps at the scale of conservation planning. Second, I evaluated whether alternative forms of reproductive behavior across fine-scale, groundwater-induced temperature gradients had qualitatively different outcomes for brook trout offspring exposure to demographically impactful flood events. After developing the most extensive breeding system description with nest-level habitat data currently available for brook trout, I simulated the effects of observed, random, and classic territorial reproductive behaviors. Observed spawning behavior generated the lowest modeled egg losses relative to perfectly size- and density-dependent or completely random habitat choice. Notably, accounting for observed behavior produced qualitatively different, less dire, predictions about the consequences of environmental change than behavior-free models. Finally, I assessed whether a stream restoration project improved the thermal environment of overwintering trout. There were beneficial increases in wintertime temperatures only in an area of known groundwater discharge, suggesting that biologists, hydrologists, and conservation practitioners should continue working together to strategically identify project opportunities in areas expected to generate or augment thermal refugia at biologically relevant scales.