Defense Date

11-15-2021

Graduation Date

Spring 5-14-2022

Availability

One-year Embargo

Submission Type

dissertation

Degree Name

PhD

Department

Biological Sciences

Committee Chair

Brady A. Porter

Committee Member

Steven C. Latta

Committee Member

Kyle W. Selcer

Committee Member

Jan E. Janecka

Keywords

bird, diet, DNA metabarcoding, stable isotope analysis, logistic regression, Bayesian, trophic, Pennsylvania, Louisiana waterthrush

Abstract

Characterizing a species’ dietary composition presents an avenue to understand many facets of its ecological niche and can provide essential information for the species’ long-term conservation. To date, the vast majority of diet studies have relied on direct identification of prey during foraging observations or from diet samples to characterize the dietary habits of birds. However, advancements in laboratory-based approaches have revolutionized the field of trophic ecology by allowing researchers to indirectly infer dietary habits with higher resolution across greater time scales. Here, I apply two of these laboratory-based techniques, namely DNA metabarcoding and stable isotope analysis, to characterize the trophic niche of Louisiana Waterthrush (Parkesia motacilla), a migratory wood-warbler species of conservation concern due to their reliance upon unimpacted, riparian breeding territories. My findings suggest that the trophic niche of nestling Louisiana Waterthrush in the Laurel Highlands of Pennsylvania is highly centered on pollution-intolerant aquatic arthropods, namely Ephemeroptera and Plecoptera, as these prey orders exhibited the highest relative dietary contribution to nestling diets, and broods were found to reduce their dietary niche breadth when such prey were provisioned in higher quantities. I also present a novel approach to integrate DNA metabarcoding and stable isotope analysis under a unified framework by allowing the probability that a prey group was detected by DNA-based methods to dictate the probability that a prey group was included in each stable isotope mixing model iteration. This approach to inform Bayesian mixing models without informative priors, termed “Detection Probability”, yielded similar results as models applying generalized (i.e. “Diffuse Priors”) or DNA-based priors (i.e. “Scaled Priors”), but did so in a manner that preserves the probabilistic nature of DNA-based data. This work highlights the utility of integrating DNA-based and isotopic techniques in dietary analyses as their coupled application offers both high-resolution taxonomic information about prey composition as well as functional information about a species’ trophic niche dynamics; two distinct but equally important facets of an organism’s feeding ecology.

Language

English

Download

Share

COinS