Dissertations, Theses, and Capstone Projects

Date of Degree

9-2025

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy

Program

Biology

Advisor

Ana Carolina Carnaval

Committee Members

Michael Hickerson

Robert P. Anderson

Matthew Aiello-Lammens

Brian T. Smith

Subject Categories

Biodiversity | Computational Biology | Evolution | Genetics | Ornithology | Population Biology | Zoology

Keywords

macrogenetics; macroecology; conservation; age-structured populations; comparative phylogeography; nucleotide diversity; polygamy; historical demography; individual-based models

Abstract

The field of comparative phylogeography aims at uncovering common underlying causes for shared patterns of diversity and diversification at the population level. However, it is now widely accepted that intraspecific genetic diversity patterns can differ considerably across co-occurring taxa due to ecological processes acting at the population level (and influencing individual movement and abundance). To model the processes underlying lineage diversification and demographic shifts in response to environmental changes, phylogeographers and population geneticists are now faced with the challenge of incorporating measurements of species ecological traits into their molecular studies. This dissertation aimed at investigating how different types of ecological traits may be incorporated in a quantitative comparative framework to advance our understanding of the processes shaping spatial patterns of genetic diversity and historical demography. For that goal, I developed and implemented novel analytical approaches to study the distribution of genetic and genomic DNA data from Neotropical bird species distributed in the Brazilian Atlantic Forest (chapters 1 and 3) and Amazonia (chapter 2). Chapter 1 studied the relative importance of different kinds of phenotypic traits (particularly those describing morphology vs. life history variation) as predictors of genetic divergence. For that, I mapped genetic differentiation within 28 bird species from the Brazilian Atlantic Forest and evaluated Random Forest models trained on those genetic data and multiple sets of ecological traits. My results showed that models trained on ecological trait data have consistently higher accuracy in predicting known measurements of genetic differentiation relative to models based on environmental descriptors alone, and that the inclusion of dispersal-related ecological traits leads to the highest increase in prediction accuracy. In Chapter 2, I analyzed how life history traits can influence population parameters (generation time and variance in reproductive success), measurements of genetic summary statistics and estimates of effective population size. This analysis allowed me to investigate how they can be integrated into simulation-based studies of historical demography. My results showed that life history traits have a large effect on population parameters and that life history traits differ in their impact on genetic summary statistics, with traits related to reproductive ages largely impacting generation time and genetic summary statistics. Estimates of population parameters from empirical genome-scale data from two species of Amazonian birds corroborated expectations that taxa with extreme polygamy, long adult longevity and later onset of reproduction will exhibit higher variance in reproductive success, longer generation time and smaller effective population sizes. Finally, Chapter 3 explored the relative contributions of descriptors of historical habitat shifts, species-specific responses to climate and phenotypic traits into models of demographic similarity across species. For that, I pooled genomic data from 19 species of Atlantic Forest bids. My results showed that differences in historical habitat shifts are a strong predictor of similarity in effective population sizes across species at large spatial scales, and that the explanatory value of phenotypic traits increases for smaller spatial scales. Overall, this dissertation highlights that ecological information can be efficiently incorporated into the analysis and interpretation of genetic and genomic patterns. Hypothesis generation and interpretation of results should nonetheless take into consideration the temporal and spatial scope of comparisons being made. Finally, the integration of different sources of ecological variation leads to more robust explanations of species response to environmental change.

Download

Share

COinS