Dissertations, Theses, and Capstone Projects

Date of Degree

9-2025

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy

Program

Biology

Advisor

Jessica Ware

Committee Members

Ana Carnaval

Christopher Beatty

Manpreet Kohli

Mary Blair

Subject Categories

Biodiversity | Bioinformatics | Entomology | Evolution | Other Ecology and Evolutionary Biology

Keywords

Dragonflies, Systematics, Phylogenetics, Niche Modeling, Systematics, Climate

Abstract

Dragonflies and damselflies (Odonata Fabricius, 1793) are among the most widely recognized and charismatic insect groups, whose significance have permeated throughout artwork, culture, and religion. They are highly mobile predators, currently consisting of ~ 6,500 species, and possessing some of the complex visual color systems, visual acuity, and flight patterns on Earth. Odonata are the among the earliest flying insects, with proto-odonates (Griffinflies) dating back to the Carboniferous (~360mya), with crown-odonata fossils stretching back to the Triassic (~250mya). Overall, dragonflies and damselflies are considered models of ecological and evolutionary research due to their long evolutionary history, range of habitats, and morphology.

The beginnings of Odonate systematics started approximately 180 years ago, with founding figures such Edmund de Selys Longchamp, Lucien-Victor Rambur, and Hermann August Hagen laying the groundwork for modern-day classifications of families, genera, and species. Traditional taxonomy relied predominantly on intuition based on morphological traits associated with wing venation, external genitalia, and nymphs. However, such traits are highly prone to convergence, obscuring the true relatedness of lineages.

The advent of Sanger sequencing provided higher resolution to the odonate tree of life, with improved identification of phylogenetic relationships among major lineages and the ability to target specific mitochondrial and nuclear genes (such as COI and 28S) that clarified previously ambiguous or unresolved taxonomic groupings. The benefit of Sanger sequencing was partially due to its high accuracy, cost effectiveness, and integration of morphological traits, to test for convergence. However, Sanger requires fresh material, limiting the utility of museum specimens, as well as the limitedness of genes, which are prone to hybridization, incomplete lineage sorting, and codon-site saturation. High-throughput molecular sequencing has enabled cost-effective, large-scale investigations into the phylogenetics and evolutionary history of Odonates by allowing the sequencing of more loci at faster rates and making it possible to sequence even century-old specimens.

The distributions of Odonata vary greatly, with the order inhabiting all continents except antarctica. Habitats can be extremely specific, some species being single stream endemics such as Apocordulia only residing in the Murray Darling Basin of Australia, to transoceanic pandemics such as the global skimmer Pantala flavescens which is found on multiple continents. Odonates possess aquatic nymphs (larvae), thus connecting between freshwater and terrestrial habitats. Adults can reside in forested or non-forested areas, while the nymphs reside either in lentic (still) or lotic (running) waters, both factors correlate with species richness. Finally, odonates are considered ecosystem indicators and/or keystone species, with numerous studies citing their importance in in the context of pest control, restoration and conservation.

Advances in genome-wide sequencing have allowed cost effective large-scale exploration into the phylogenetics and evolutionary history of Odonata, while the advent of citizen-science databases have enabled regional-scale studies into their ranges and niches. Throughout my doctoral work at the American Museum of Natural History and the City University of New York, I have had the opportunity to publish 13 manuscripts exploring dragonfly evolution and ecology using both methodologies. For this dissertation, I focus on four chapters whose aim is to answer two key questions pertaining to odonate biology: How are odonates related and how have they evolved over time? and what environmental factors shape odonate distributions? I plan to answer the first question by exploring the evolutionary relationships of odonate families using high-throughput molecular sequencing and the second question by investigating biogeographical patterns of Odonata using species distribution modeling (SDM).

How are odonates related and how have they evolved over time?

The first chapter utilizes Anchored Hybrid Enrichment (AHE) high-throughput molecular sequences and comprehensive morphological scoring of specimens based on wing, body, nymphal and genitalic characters to propose a new taxonomic classification for the dragonfly families Corduliidae (emeralds) and Synthemistidae (tigertails). Based on our results, we propose the creation of three new families, and the revision of six. We highlight the utility of Anchored Hybrid Enrichment sequencing as our molecular phylogeny recovers strong support for familial and intrafamilial relationships. Furthermore, our morphological analyses indicate high degrees of homoplasy, with most families being unable to be classified based on a single synapomorphy. Finally, ancestral state reconstructions estimate the ancestor of the superfamily Libelluloidea to possess a compact anal loop, prominent uniform teeth on the nymphal labial palps, and a reduced ovipositor, originating in the Late Cretaceous, with most families diversifying throughout the Cenozoic.

What environmental factors shape odonate distributions?

The subsequent chapters provide a multifaceted approach to species distribution modeling (SDM) of odonates as well as termites, by leveraging museum occurrence records and citizen-science databases alongside climatic, soil, and vegetation data.

Within Chapter 2, we explore habitat suitability for non-neoisopterous termite genera using a static modeling approach. Static models are a common form of SDM whereby environmental data is averaged over the course of several decades to encapsulate suitability across time. Static models are ideal for comparing distributions across species and provide not only a baseline for poorly known or rare species but can be applied for future predictions of the taxa in the face of habitat destruction and climate change. We conclude that temperature and precipitation extremes as well as soil type and pH are the predominant drivers if suitable habitat for non-neoisopterous termite species and highlights the utility of citizen-science based datasets.

Within Chapter 3 we assess the impact of recent bushfires on the distribution of the endemic Swamp Tigertail dragonfly (Synthemis eustalacta) using a dynamic modeling approach. Dynamic models incorporate temporally matched data, in which models are generated using occurrence and environmental data from the same year. Such modeling allows for the estimation of variation in habitat suitability over time, identify temporal trends, and determines if such trends are short-term habitat shifts or anomalies, or permanent range shifting. Overall, dynamic modeling captures fine-scale response of taxa. Our results indicate that dynamic models outperform static ones, with environmental variables pertaining to figure posess minimal importance in the distribution of S. eustalacta. We highlight not only the efficacy of dynamic ENMs to capture spatiotemporal variables such as vegetation cover for an endemic insect species but also providing a novel approach to mapping species distributions with sparse locality records.

Finally, within Chapter 4, we generate paleodistributions of the relictual dragonfly genus Epiophlebia to the Last Glacial Maximum. Paleoecological niche modeling (PaleoENM) projects modern-day static models onto paleoclimate layers reaching as far back as the Miocene. Such modeling can estimate historical range shifts of species, and links current distributions to past geologic events, and can provide clarity to the response of modern-day taxa to climate change over their evolutionary history. Finally, PaleoENMs can pinpoint historical refugia and niche stability, and can aid in future climate resilience predictions and potential for susceptible species. In the case of Epiophlebia, we estimate historical routes of gene flow, predict unsampled areas possessing the potential to harbor new populations of the genus, and more broadly highlight the utility of Paleo-ecological niche modeling (PaleoENM) in the context of odonate biogeography.

Conclusions

In total, this work provides a roadmap for the synthesis of genomic and ecological analyses in the context of odonate diversification and evolutionary history. The results of my first chapter highlight the importance of combining high-throughput molecular sequencing with detailed morphological assessments to uncover phylogenetic relationships within complex and historically misclassified groups. The use of Anchored Hybrid Enrichment (AHE) enabled strong resolution of familial and intrafamilial relationships, supporting the revision of six dragonfly families and the proposal of three new ones. Morphological homoplasy was widespread, emphasizing the limitations of traditional taxonomy and the necessity of genomic tools for accurate classification.

Chapters 2 through 4 demonstrate the power of species distribution modeling (SDM) to understand environmental drivers of odonate and termite distributions across temporal and spatial scales. Static models revealed that temperature extremes, precipitation, soil properties, and citizen-science data play crucial roles in defining habitat suitability for non-neoisopterous termites. Dynamic modeling in Chapter 3 showcased the effectiveness of temporally matched data in capturing fine-scale responses to recent environmental disturbances, such as bushfires, and revealed the limited predictive power of commonly used static variables for endemic species like Synthemis eustalacta. Finally, PaleoENM in Chapter 4 reconstructed paleodistributions of Epiophlebia, linking modern ranges to historical climate events and predicting areas with potential undiscovered populations.

Together, these findings underscore the complementary strengths of genomic and ecological approaches in illuminating both the evolutionary history and biogeographic patterns of Odonata. This dissertation not only enhances our understanding of odonate systematics and environmental adaptability but also offers tools and frameworks applicable to the broader study of insect biodiversity, conservation, and climate change resilience.

Download

This work is embargoed and will be available for download on Monday, March 30, 2026

Share

COinS