Assessing the Effects of Crude Oil Exposure on Avian Physiology

Photo of Myself at the Spring 2023 Graduation Ceremony.

Author: Jullianna McCauley | Major: Biology | Semester: Spring 2023

Hi, my name is Jullianna McCauley, and I’m currently a biology major working in the DuRant lab of Physiology in the Biological Sciences department as an undergraduate research assistant. Over the duration of the Spring 2023 semester, I have been given the opportunity to study the sub-lethal effects of crude oil exposure on zebra finches using a dynamic energy budget individual-based model (DEB-IBM) in the coding program, R Version 4.3.0. Crude oil exposure in birds has been shown to affect metabolic rates, activity patterns, fecundity, and overall fitness. DEB-IBMs can be used to simulate multiple individuals and changes to their physiology to extrapolate to broader scale impacts. A model-based approach allows me to benefit from existing data on crude oil exposure in birds without the need for large scale natural experiments while simultaneously informing environmental impact evaluation for future spills. Three treatments of oil exposure were randomly assigned to each simulated individual to compare the effect of concentration: no exposure (control), low exposure (3mL/kg), and high exposure (6mL/kg). I was interested in performing a series of simulations of a zebra finch using a DEB-IBM to address whether sublethal effects of crude oil exposure scaled with concentration, and the resulting impact on fitness and reproductive success.

I’ve always had an interest in avian conservation, so when my mentor, Dr. Sarah DuRant, introduced me to the idea of studying the effects of crude oil on zebra finch populations, I was eager to start the project. Through Dr. DuRant, I was introduced to Max-Carnes Mason, a PhD candidate in the Beaupre Lab of Physiological Ecology. Max guided me through the parameterization of a DEB-IBM through our weekly discussions on zebra finch physiology. My primary job throughout this project was to find values (i.e., metabolic rates, energy allocation data) searching through literature on Google Scholar that could be applied to building the model. At first, deciphering language in R was overwhelming, since I’ve never coded until this project. However, after getting advice from my mentors and taking plenty of time for practice, I was able to code some graphs that present my final excepted results.

As of right now, the model is still in the final stages of development for the simulations Max and I want to run. However, using the data I’ve collected from the literature, I can expect zebra finches in the high exposure treatment to have the lowest mass and reproductive success. This is likely due to a decrease in activity (i.e., foraging), so birds exposed to oil use more energy reserves in replace of frequent foraging. Additionally, since birds exposed to oil have less energy reserves, we expect clutch sizes to scale with exposure. We expect birds in the control treatment to have the highest clutch size (4-6 eggs), low exposure treatment to have the second highest mean clutch size (1-4), and birds with high exposure treatment to have the lowest mean clutch size (0-2). Knowing this, we can assume birds exposed to high doses will have a more detrimental impact on the overall population compared to individuals who are not exposed to oil.

Moving forward, I hope to further assist Dr. DuRant and Max in finding any additional literature that might be needed for parametrization of the model. The funding from this grant has allowed me to gain a more detailed understanding of avian physiology through interpreting relevant manuscripts, writing professional documents (i.e., proposals), and creating graphs using relevant quantitative data in R. These skills have helped me obtain a graduate position in Oklahoma State University’s Natural Resource Management and Ecology department for the Fall 2023 semester. There, I plan to apply the techniques I’ve learned over this project to study the effects of extreme hunting on Northern Bobwhite populations using a modelling exercise in R.