- Climate change is predicted to affect ecosystems, including systems already stressed by human impacts. One ecosystem that is already highly impacted by human land use is the cold headwater stream system of the Pacific Northwest. One method of assessing the function of an ecosystem is by using an indicator species. Rhyacotriton variegatus is one such indicator species, sensitive to disturbance, and especially to temperature elevation. This study combines field measurements from the warmest edge of the range of R. variegatus, laboratory determination of thermal tolerance, and modeling. These diverse experimental sources combine to clarify the potential risks of climate change on R. variegatus, and the headwater streams they occupy. Abiotic factors are important determinates of the range of species. Predicted range shifts under climate change are based on the assumption that temperature increases will make habitat at the edge of the known range unsuitable in the future. In order to accurately predict such changes, a quantification of the current thermal boundary is needed. In Chapter Two, I placed temperature loggers and measured other environmental variables in 28 streams: 8 in the cool core of the range of R. variegatus, 10 as far east and south as R. variegatus has ever been found, and 10 outside the known range of R. variegatus. The variables which best defined the range edge were degree days (number of days over specific temperatures), and the slope of the stream bed. Specific physiological tolerance information is also essential for accurate modeling of species habitats. Physiological limits should be determined experimentally using procedures that mimic natural conditions as closely as possible, so that the results will be applicable to natural systems. Forecasting the effects of human activities on populations also requires an understanding of how specific abiotic changes will impact different life stages. I used a realistic cycling temperature treatment in Capters Three and Four, based on the data collected in Chapter One. I tested the survival of larval R. variegatus at a chronic exposure (21 days), and the level of stress as measured by corticosterone in adult R. variegatus. Larval R. variegatus survived up to a daily maximum of 23° C, beyond this the larvae died (LT 50 value of 24° C). I found that daily maximum temperatures over 18° C caused a doubling of corticosterone. There are many ways of modeling future climate change and the effect of this change on species' distribution. I chose to use large array of potential climate futures, modeling methods, and time periods to forecast the change in R. variegatus' range. This allowed me to compare the variation between the predictions for climate change, and find averages across the models. I used two correlative models, and one mechanistic model. The mechanistic model incorporated the relationship between air and water temperature from Chapter Two, and the physiological limits from Chapters Three and Four. All models predicted decreases in areas of the map classified as excellent habitat for R. variegatus. As expected, the reduction in range was most severe at longer time periods into the future, with higher CO₂ amounts in the atmosphere, and in models that incorporated more abiotic variables. R. variegatus are sensitive indicators for headwater stream ecosystem function, and will have a reduced range under climate change.