Ecology and Evolutionary Biology, 1992-2026

Since 1860, human body temperature (BT) has decreased by 0.4°C from mechanisms that remain poorly understood and implications for fever ranges and population health. This study tested whether individual factors such as athletic status, BMI, sex, sleep, meals can predict BT, and used compartmental modeling to analyze what types of shifts in infection dynamics (assuming that background infections elevate temperatures) could explain the decline in BT. Cross-sectional data were collected from 101 Princeton students through a self-reported survey, oral temperature measurements in the morning and evening, and athletic classification. SIR and SIS models were solved at endemic equilibrium to estimate the parameter changes required to produce the observed BT de- cline. Both mean morning BT (36.58 ± 0.16°C) and evening BT (36.63 ± 0.17°C) were below 37°C. Among pro-athletes, BMI was negatively associated with BT (𝛽 = -0.041, p = 0.017). Only non-athletes showed a significant rise in morning-to-evening BT. SIR modeling indicated that a 2- to 6-fold increase in recovery rate could plausibly explain the observed BT decline. The SIS model also indicated a plausible mechanism through an increase in the recovery rate of 1.2 to 1.7 for acute recurrent infections (illustrative 3-7 day infections), but not for chronic infections. In both SIR and SIS models, BT decline was found to be less likely due to transmission rate changes. These findings suggest that body composition, athletic training, and reduced bacterial infection all contribute to variation in human BT. Understanding the drivers of BT decline has direct clinical importance, as fever thresholds are anchored to the old 37°C baseline and lower baseline BT have been associated with longer human life expectancies.

Hawaiʻi imports 85–90% of its food and applies the majority of its nitrogen as synthetic fertilizer, despite a pre-contact agricultural record in which a population of nearly 800,000 was sustained on roughly a quarter of the archipelago's land area without external inputs. Indigenous Hawaiian colluvial systems integrated sugarcane, ʻulu, and kukui residues across stratified polycultures, suggesting that combinations of these species supplied biologically fixed nitrogen at a scale sufficient to support continuous cultivation. This thesis tests whether asymbiotic nitrogen fixation associated with sugarcane (Saccharum officinarum) decomposition is enhanced, suppressed, or unaffected by co-mulching with ʻulu (Artocarpus altilis) and kukui (Aleurites moluccanus), and identifies the mechanisms governing any non-additive response. Mulch mixtures spanning pure monocultures (CCCC, UUUU, KKKK) and four-species blends (CCUU, CCKK, CCKU, KUUU) were incubated in field litterbags for 275 days, with periodic acetylene reduction assays, cumulative nitrogen accounting via trapezoidal integration, and stoichiometric tracking of C:N decay. A two-way ANOVA revealed a significant Treatment × Day interaction (F = 19.56, p < 2 × 10⁻¹⁶), and a linear model identified the proportion of sugarcane in the mixture as the strongest predictor of cumulative nitrogen yield (β = 2.03, R² = 0.62, p = 0.0016), with non-additive residual analysis isolating CCUU as uniquely synergistic (+0.338 gN/kg above expectation) and KUUU and CCKU as antagonistic. Mechanistically, sugarcane functions as an electron-donor substrate for nitrogenase, while ʻulu enables the microbial colonization that retains fixed nitrogen against leaching; the spatial separation of kukui- and ʻulu-dominated agroforests in traditional colluvial systems mirrors the antagonism observed experimentally.

The Hawaiian monk seal (Neomonachus schauinslandi) is an endangered marine mammal endemic to the Hawaiian archipelago, with approximately 1,400 individuals remaining and exceptionally low genome-wide genetic diversity, largely due to a bottleneck associated with 19th-century hunting. Although intensive ecological management by the National Oceanic and Atmospheric Administration (NOAA) has supported recent population growth, no formal genetic monitoring program exists, and mating system dynamics, including the type of mating system along with potential inbreeding and sexual selection, remain unresolved. Because mating system dynamics can shape genetic diversity, effective population size (Ne), and therefore adaptive potential, they have important implications for species resilience in the face of natural and anthropogenic threats, including disease, habitat degradation, and climate change. I hypothesized that Hawaiian monk seals exhibit extremely low genetic diversity, weak island-level population structure, and an Ne trajectory consistent with the species’ demographic history. In terms of mating system dynamics, I hypothesized that the species exhibits mild polygyny consistent with closely related phocids, that recent active inbreeding is limited despite chronically small population size, and that offspring heterozygosity may exceed maternal heterozygosity if mate choice favors genetically diverse males. Using restriction site-associated DNA sequencing (RADseq) data from 112 individuals sampled across five Northwestern Hawaiian Islands (NWHI), I generated filtered SNP datasets optimized for pedigree-based and population genomic analyses. I used the R package SEQUOIA to reconstruct the first genomic pedigree for the species and validate parentage assignments against known mother-pup pairs; I then characterized relatedness, inbreeding, heterozygosity-based mate choice, population structure, and historical Ne. The pedigree correctly assigned 18 of 20 known mother-pup pairs, with the two unassigned pairs having low maternal genotyping rates, and identified two inferred sires, including one assigned to four offspring across multiple years and at least two confirmed dams. Island-level population structure was weak, genetic diversity was extremely low, and inbreeding coefficients approximated zero. Heterozygosity-based mate choice was not detected. These findings support mild polygyny or sequential monogamy rather than strict monogamy and establish a cost-effective genetic framework for integrating pedigree-based monitoring into Hawaiian monk seal conservation management.

Antibiotic resistance poses an increasingly urgent global health threat, driving rising mortality and healthcare costs worldwide. Bacteriophages, viruses that infect and lyse bacterial cells, have garnered significant interest as potential replacements or complements to antibiotics. A critical step in clinical use of phages, known as phage therapy, is the amplification of phage particles on a susceptible host strain prior to therapeutic use. Yet the impact of amplification host strain identity on phage phenotype remains poorly characterized. Furthermore, the repeated dosing of phage into already-coevolving bacterial populations during treatment has potential implications for resistance trajectories that are not well understood. Using Pseudomonas aeruginosa and the lytic phage ΦPA3 as a model system, this study demonstrates that amplification host strain identity modulates phage host range and killing efficacy, with exopolysaccharide content implicated as a potential mechanistic source of this modulation. A mathematical model of bacteria-phage coevolution additionally demonstrates that repeated phage influx accelerates the extinction of susceptible bacteria, but that this effect does not scale with increasing daily dose frequency, suggesting diminishing returns beyond a single daily administration. Together, these findings highlight amplification host identity as a biologically meaningful variable in phage therapy design, and suggest that dosing frequency may have less influence on resistance evolution than previously assumed.

Arboviral diseases such as dengue, Zika, and chikungunya remain significant global public health burdens, and traditional control methods are increasingly ineffective. The introduction of Wolbachia-infected Aedes aegypti mosquitoes has emerged as a promising alternative for reducing arboviral disease transmission. However, reported outcomes across studies vary substantially, highlighting the need to understand the factors driving variability. This systematic review examines variation in reported outcomes of Wolbachia-based interventions by synthesizing findings from empirical trials and mathematical modeling studies. A PRISMA-guided literature search was conducted across three major databases, and 42 studies were included in the final review. Selected studies were analyzed to compare patterns of Wolbachia establishment, spatial spread, and incidence reduction to identify key drivers of variation across studies. Across both empirical and theoretical approaches, substantial reductions in arboviral incidence were observed when Wolbachia achieved stable establishment. However, outcomes varied widely across implementations due to differences in environmental conditions, biological parameters, operational constraints, and surveillance methodologies. While modeling approaches capture key mechanisms, they often rely on fixed parameter assumptions that do not fully reflect the variability observed in field settings. These findings underscore the importance of integrating modeling and empirical approaches. Incorporating field-derived data into models, alongside the use of context-specific, model-informed intervention strategies in the field, can improve both predictive accuracy and implementation efficiency. A more integrated and context-dependent approach is essential for optimizing Wolbachia interventions and its applicability across diverse ecological and operational settings.

Highly Pathogenic Avian Influenza (HPAI) is an emerging zoonotic disease of global concern, driven by transmission between wild avian hosts and commercial poultry. Understanding where these populations overlap is critical for predicting and mitigating spillover risk. This study aimed to identify geographic hotspots of potential HPAI spillover across the contiguous United States using a geospatial analysis. Relative abundance data for 157 wild avian host species were integrated with county-level poultry density data to construct a multiplicative spatial Risk Index. To identify extreme spillover interfaces, county-level risk values were standardized using Z-score normalization, with thresholds of Z > 3 and Z > 4 used to define high and extreme-risk counties, respectively. Spatial statistical analyses, including Spearman’s rank correlation and Global Moran’s I, were conducted to assess relationships and clustering patterns. All data analysis and associated code can be accessed through the GitHub repository (Appendix A). In addition, a field study in South Texas collected 111 avian fecal samples to test for the presence of HPAI and Avian Bornavirus (ABV) in urban bird communities (Appendix B). Results revealed no significant global correlation between poultry and wild host densities (ρ ≈ 0, p = 0.953), but a strong, significant spatial clustering of high-risk interfaces (Moran’s I = 0.610, p < 0.001). The Z-score analysis identified 53 high-risk counties (Z > 3) and 28 extreme-risk counties (Z > 4), representing less than 1% of U.S. counties. These high-risk areas were concentrated in major agricultural and migratory regions, including the Mississippi River Basin, Mid-Atlantic, Southeast, and California Central Valley. All field samples tested negative for both pathogens. These findings demonstrate that HPAI spillover risk is highly localized rather than uniformly distributed, highlighting the importance of targeted, spatially informed biosecurity strategies.

Humans have been major drivers of evolutionary change. Anthropogenic influences like industrialization and urbanization have resulted in significant, negative impacts to the environment and ecosystem. These influences have exacerbated climate change and increased the severity and frequency of extreme weather events like hurricanes. As hurricanes are becoming more frequent in tropical regions, research on their effects for the species living in these areas is much needed. Our study aims to address a gap in the literature and understand how different modes of anthropogenic change act together in exerting selective pressures in a species of tropical lizards, the Anolis cristatellus or the Crested Anole.

For anoles to survive hurricanes, they have to display good clinging ability. In order to carry out necessary life functions, anoles also have to sprint well. As such, we conducted sprinting and hurricane clinging trials to understand the relationship between these two performance metrics. We found that since anoles have had to experience hurricanes throughout their evolutionary history, they have evolved a balance between sprinting ability and hurricane survival. They have developed morphological combinations such as larger forelimbs, pelvises, and fourth toes but smaller metacarpals and pectorals to perform well in clinging and sprinting.

However, their urban counterparts have additionally experienced the selective pressures posed by urbanization. Due to this, they have shifted in their morphology to evolve sprinting specialization. This has disrupted the balance between sprinting and clinging performance that anoles have ancestrally maintained. While larger forelimbs but smaller pelvises, fourth toes, metacarpals, and pectorals confer good sprinting ability, their clinging performance is insensitive to morphological variation. They display a trade-off as sprinting specialization has made them poor performers when it comes to hurricane clinging ability.

Behavioral mismatches are one of the leading causes of euthanasia and relinquishment of dogs. Understanding the demographic and biological factors that support successful human-dog relationships is critical to improving adoption outcomes. Unlike behavior which may change with age and environment, a dog’s genotype is stable over their life. Genetic analyses can thus provide insight into a dog’s social needs and compatibility with different home environments. Previous research has determined four retrotransposon insertions within three genes (WBSCR17, GTF2I, and POM121) that are associated with a dog’s propensity for prolonged social contact with humans (“hypersociability”). A higher copy number of the derived allele correlates with increased hypersocial behavior. In this study, I examine whether there are differences in genetic sociability and behavior between long-term owned dogs and dogs that are currently in shelters. DNA samples were collected from 317 owned dogs and 171 shelter dogs, along with demographic and behavioral information. A greater proportion of shelter dogs had 3 or fewer alleles, whereas owned dogs had 4 or more (p = 0.039). I found no significant difference in per-locus derived allele frequencies between groups (Bonferroni-corrected p > 0.05). This study also investigated the associations between genotype and other factors, including weight and behavior. Information on WBSCR17 and GTF2I showed strong associations with behavioral variation among all dogs. Overall, the results of this study serve as an important thought experiment for future work on shelter dogs and long-term owned dogs as well as the use of genetic surveys to improve adoption success and shelter resource allocation.

Many animals use acoustic signaling to mediate social interactions such as competition, courtship, and territory defense. Birds, with their diverse vocal repertoires, present an optimal study system to examine these signals. In this thesis, I focused on the chip call of the broad-tailed hummingbird (Selasphorus platycercus), a migratory species that stands out for its evolutionary loss of song. While the chip call has been observed in several species of hummingbirds, the structure of this monosyllabic vocalization across behavioral contexts is largely unknown. By collecting field recordings and running acoustic playbacks of the broad-tailed hummingbird chip call, I investigated the acoustic features and syntax of male and female chip calls to better understand their roles in courtship and territoriality. I found that male and female calls exhibit vocal sexual dimorphism in chip call maximum frequency, minimum frequency, and duration. I also identified a bimodal distribution in within-bout call interval consistent in both males and females and more prominent in territorial pursuit, vigilance, and courtship contexts. A closer examination of syntactic structure using Markov analysis revealed a bias toward highly stereotyped, fast, successive chip calls during male pursuit and vigilance. No such pattern was found in rufous hummingbirds. These results suggest that broad-tailed hummingbirds encode sex-specific and context-specific information in the acoustic structure of their chip calls in lieu of more complex song. This study highlights the importance of the acoustic landscape to animal communication and paves the way for future studies examining call patterning within and across hummingbird species.

Often seen as a detriment to conservation efforts, the advent of modern genomic analyses has shed a new light on the potential utility hybridization and introgres- sion may serve in preserving rare species. Under ideal ecological situations, hybrids and introgressed individuals may serve as repositories of genomic information for rare species. An ongoing case study of this concept has been revealed in the sequencing analyses of the critically endangered Canis rufus (Red Wolf) and abundant Canis latrans (coyote). Recent findings of high proportions of stored Red Wolf DNA within coyotes inhabiting the American Gulf Coast has questioned the stigma held on these evolutionary processes. Simulation models developed in this thesis reveal statisti- cally that coyotes are systematically biasing mate selection for conspecifics with high densities of Red Wolf ancestry, resulting in genotyped individuals nearing governmen- tal thresholds for Red Wolf taxa declaration. Additionally, newly sequenced coyote genomes serve as an expansion of Ancestry Informative Markers for new captive breed- ing efforts proposed between computationally optimized canid cohorts. Results from simulation models and prospects held from proposed breeding strategies informed by analysis of new genome sequencing data offer a key insight into how these once- avoided phenomena may serve conservation goals both through natural processes and through human intervention to spark hope for not only the Red Wolf, but for many endangered species in similar ecological situations.

Laboratory mice are commonly used to model oral diseases like periodontal disease. However, their lack of exposure to environmental pathogens may limit the translatability of these mouse models to real-world organisms. Rewilding solves this problem by exposing laboratory mice to outdoor conditions, but the effects on the oral microbiome and periodontal pathology remain poorly understood. In this study, I explored how rewilding alters the oral microbiome and shapes periodontal bone loss in C57BL/6J mice. To do so, I use culturing procedures, 16S rRNA sequencing, ITS sequencing, and distance measurements from the cementoenamel junction (CEJ) to alveolar bone crest (ABC) to compare groups of lab-housed and rewilded mice. I found that rewilded mice had greater absolute abundances of culturable fungi, aerobic bacteria, and anaerobic bacteria. Rewilded mice also had significantly different bacterial community compositions compared to their lab-housed counterparts. However, I found no significant differences for the alpha diversity of either bacteria or fungi. While rewilded mice did have slightly greater periodontal bone loss, the observed differences were not significant. Attempts to correlate specific bacterial taxa with increased bone loss distances identified emerging trends, but I found no significant associations after multiple testing corrections. These findings suggest that while rewilding alters the oral microbiome composition and increases microbial burden, the given 5 week timeframe may not have been sufficient to induce significant periodontal bone loss. As a result, future research should utilize longer rewilding studies to create more representative animal models and more comprehensively understand how the environment can shape the oral microbiome and disease progression.

Previous problems with character state constructions in maximum parsimony analyses cause resulting phylogenies to be affected by subjectivity of anatomical features and bias. A new taxonomic character matrix designed to better assess morphological diversity in Squamates is applied to specimens of M. missouriensis and Mosasaurus sp. from the Bearpaw formation in southern Canada, placing them within a larger assemblage of mosasaurines. Several polytomies and low bootstrap support in the resulting consensus tree suggest that a lack of available data due to preservation and taphonomic effects still hinders the resolution of maximum parsimony analyses, despite improved character construction. Most Bearpaw specimens formed a polytomy in a sister clade to a clade containing a separate polytomy of M. hoffmanii, M. missouriensis, M. lemonnieri, and M. mangahouangae, possibly indicating certain region-specific morphologies. However, taxonomic assignment and relative contribution of interspecific versus intraspecific variation among the Bearpaw specimens remains unclear. Construction of a new taxon specific to the Bearpaw formation is not strongly supported.

Non-malarial fever (NMF) characterizes a large, but understudied disease burden globally. Use of new rapid diagnostic tests (RDTs) to diagnose NMFs offers a promising opportunity to better understand the distribution of infectious diseases. However, cost effective deployment of such tests would be strengthened by characterization of the spatial distributions of NMF, and associated landscape (human population density, temperature) and individual (e.g., use of bednets etc) features. Using Demographic and Health Survey (DHS) data from Madagascar (2011–2021), this study examines the prevalence, predictors, and spatial distribution of NMF among children under five. Data from 29,371 individuals were analyzed using generalized additive models (GAMs) incorporating environmental and demographic predictors. Fever prevalence ranged from 12.4 - 16%, with NMF accounting for 88.4% of the cases tested across all survey years. Population density was the most consistent predictor of NMF, showing a non-linear (broadly increasing and saturating) relationship with fever risk. Predicted probability of NMF ranged from 2.9 to 19.4% across the island. Our results indicate that most fevers in children in Madagascar across this timespan are not attributable to malaria and are spatially heterogeneous. We conclude by discussing how increased diagnostic capacity and geographic targeting in disease surveillance programs could uncover the ecology of other important disease agents in Madagascar

As coral reef cover declines globally due to climate change and anthropogenic stressors, accurate survey methods are necessary to collect coral data and inform conservation efforts. However, current methods exhibit a tradeoff between resource efficiency and taxonomic detail. Utilizing citizen science for coral reef surveying has been proposed as a resource-efficient technique to gather both wide-scale and detailed coral reef data. Studies have demonstrated promising results in locations such as the Great Barrier Reef but have yet to thoroughly evaluate the method in other reef ecosystems such as those in Hawaii.

This study investigated the accuracy of citizen scientists in identifying coral coverage for five types native to the Hawaiian archipelago – antler, cauliflower, finger, lobe, and rice coral. By comparing citizen scientist cover estimates to expert estimates, this study found that citizen scientists can measure all individual coral types and total coverage within 20% accuracy of expert estimates, an ecologically acceptable range. Antler, cauliflower, finger, and rice coral all fell within 5% of expert values while lobe coral exhibited the greatest overestimation at 14.9%.
Additionally, most coral types reached a ±0.05 precision threshold in less than 50 images while lobe coral and total cover did not reach this threshold at any number with the given dataset.

While a larger dataset is required to draw definitive conclusions, these preliminary findings suggest that citizen science is a promising tool to accurately survey Hawaiian coral reefs and support conservation efforts.

The common preclinical mouse models lack the mature immune system of humans, making them poor models for clinical research. Commensal microbes and pathogens, through their co-evolutionary history with the mammalian immune system, can shift the mouse immune system toward that of humans. Using rewilding and Trichuris muris helminth infection, we examined the independent and combined effects of microbial exposure and infection on immune functions and phenotypes in C57BL/6J (B6) and B10.BR (B10) mice. For immune phenotypes, we found that rewilding shifted the mouse blood lymphocyte to neutrophil proportions toward those of humans, and that infection increased spleen mass in B6 mice only. For immune function, we measured Interferon-γ (IFN-γ) and Tumor Necrosis Factor (TNF-α) production by splenocytes stimulated in vitro with the CD28 superagonist (CD28SA) that caused life-threatening cytokine storms in humans but not in mice. Most notably, T. muris infection alone was sufficient to drive ~23-fold higher IFN-γ production to CD28SA stimulation, while rewilding alone drove a more modest ~4-fold increase, and the combined effect was the most dramatic at an ~111-fold increase. These results highlight the importance of environmental experience and host genotype as determinants of immune phenotype and function, and suggest that a single defined helminth infection may offer an accessible means to mature the mouse immune system for more accurate and safe clinical research.

Intraspecific variation in niche space is increasingly recognized as important to species coexistence. Males and females face divergent ecological pressures, primarily stemming from sexual size dimorphism and differing reproductive strategies, that may produce intersexual niche differences. Using fecal metabarcoding data to reconstruct the diet of East African herbivores, I tested: 1. the prevalence of intersexual niche differences, 2. whether the magnitude of intersexual difference was correlated with sexual size dimorphism, 3. whether intersexual diet differences in bushbuck (Tragelaphus scriptus) led to intersexual differences in diet quality, and 4. whether intersexual diet differences in bushbuck were produced by variation in resource availability or variation in resource preference. I found intersexual dietary differences to be widespread, but the magnitude of the difference to be negatively correlated with grazing, not positively correlated with sexual size dimorphism. I found that diet quality did not differ between bushbuck sexes, but that larger males had higher quality diets than smaller males because of higher quality territories. These results point to underlying mechanisms that drive intraspecific variation, shaping individual and species-level resource utilization.

Recent advances in microbial research have revealed the crucial role that the gut microbiome plays in maintaining host health across a multitude of physiological systems, especially during early life maturation. Correspondingly, these studies indicate that antibiotic prescriptions administered to infants, which disrupt the development of a healthy microbiome, often have immediate and lasting consequences for host health. While these investigations are important, this thesis explores an underappreciated source of prolonged and unavoidable antibiotic exposure: environmental antibiotic pollution. As this crisis continues to escalate globally, it is crucial to investigate how exposure to subtherapeutic concentrations of antibiotics—similar to those measured in polluted waterways—impacts the gut microbiome, especially in early life. To address this pressing gap in existing literature, early-life (4.29 – 11.14 weeks old) and mid-life (17.29 – 35.14 weeks old) wild-derived mice were subjected to a seven-day treatment of low-concentration ampicillin, and changes to their gut microbiota were analyzed from fresh fecal samples. We found that exposure to subtherapeutic concentrations of ampicillin had a significant but transient impact on species richness (α-diversity) and a persistent impact on microbial community composition (β-diversity) in both the early-life and mid-life murine gut microbiome. Crucially, early-life mice younger than eight weeks were particularly susceptible to antibiotic-induced perturbation and largely unable to recover either species richness or composition upon discontinuation of treatment, suggesting that early-life antibiotic exposure before maturation may pose a critical threat to host health. Our results highlight the pressing risk that the global rise in antibiotic pollution poses to wildlife health. Further research should work to confirm the applicability of these findings to the human gut microbiome and investigate the impact of antibiotic pollution on comorbidities associated with early-life gut dysbiosis, like systemic inflammation, immune dysregulation, and antimicrobial resistance.

Avian H5N1 influenza, especially clade 2.3.4.4b viruses, is an increasing public health threat in recent years, responsible for global, widespread deaths in birds, cattle, and other mammals. Though this virus is not yet transmissible between humans, an A/H5N1 virus with human-to-human transmission potential could have severe ramifications since humans have not been previously exposed to this influenza subtype. Reasons for the absence of human-to-human transmission remain unclear but likely include a combination of the lack of replication of H5N1 in the upper respiratory airways and broadly cross-protective pre-existing immunity. Recent serological data support the existence of pre-existing Hemagglutinin (HA) H5 Stem antibodies in the population, from shared epitopes with circulating seasonal influenza A viruses. This project aims to model the potential transmission dynamics and health burden of an A/H5N1 virus in the background of prior immunity to the H5 stem; we hope to understand the current and future immune landscape as well as the optimal age groups for vaccination to mitigate pandemic spread. By analyzing publicly available serological data, we show that prior immunity to the H5 stem depends on both birth year via the first influenza exposure in childhood (H1N1 HA stem imprinting) and age through repeat lifelong exposures to influenza. Next, we integrate this information on pre-existing immunity into a mathematical transmission model to anticipate the impact of an H5N1 outbreak until 2050, accounting for cohort aging and future imprinting patterns. We find that fluctuations in prior immunity due to birth cohort imprinting in 45 – 64-year-olds and the loss of prior immunity in 65+ year-olds suggest that the highest attack rates for an A/H5N1 pandemic would occur in 2025, followed by 2050, and then 2015. Based on this immune landscape and contact patterns, modeling suggests that the 6 – 12-year-old age group would have the highest attack rate in a potential H5N1 pandemic, and prioritizing vaccination for this age group would be optimal to reduce attack rates and hospitalizations through direct and indirect protection.

Multilevel societies, hierarchical social structures in which smaller subunits cluster to form larger groups, allow organisms to flexibly balance the costs and benefits associated with group living. Previously thought to require complex cognitive abilities unique to a few species, multilevel societies are well-studied in primates and other mammals, but their functions in other taxa are less understood. Palmchats (Dulus dominicus), communal-nesting passerines endemic to Hispaniola, construct multi-chambered compound nests that are clustered among neighboring palms and separated into distinct nest ‘neighborhoods.’ This forms the physical structure of a multilevel society with hierarchically nested tiers expanding from chambers to nests to neighborhoods to populations. This study sought to investigate whether palmchats tailor their investment in the cooperative behavior of mobbing based on their level of social association within this multilevel structure. Playback experiments were conducted at active nest sites and behavioral response variables were recorded for focal individuals and combined into an additive composite score measuring response strength. Distress calls from group mates elicited the strongest responses (mean additive composite score: 11.57 ± 2.76), those from unfamiliar individuals elicited intermediate responses (9.82 ± 3.71), and those from neighbors elicited the weakest responses (8.86 ± 4.26). Although limited sample sizes prevented most results from obtaining significance, the multivariate effect of social category approached significance (Pillai’s trace = 0.668, p = 0.060), primarily driven by differences in time to first flight after playback began. Future experiments are planned to increase the sample size. If this pattern holds, it would show that the ordering of response strength departs from a simple gradient that declines with social association and instead suggest that cooperative investment is shaped by the distinct dynamics at each social level: elevated for group mates due to higher relatedness, cooperative breeding, and shared investment, suppressed for neighbors due to territoriality and public information redundancy, and baseline for unfamiliar individuals. These findings provide preliminary evidence that vocal individual recognition mediates cooperation across the tiers of this avian multilevel society and contribute to the growing recognition that multilevel societies, and complex functioning within them, are not restricted to large-brained mammalian taxa.

Nest parasitism is a growing field within behavioral biology. Obligate brood parasites lay their eggs into the nest of host species, abandoning their young and offloading the costs of parental care onto the hosts. Interactions between hosts and parasites are interesting to evolutionary biologists because they can reveal how natural selection favors the evolution of host defenses. The current literature underrepresents the evolution of these early-stage defenses. In this study, we observed the relationship between the parasite pin-tailed whydah (Vidua macroura) and its host the scaly-breasted munia (Lonchura punctulata). Their interactions began recently and are solely observed in California. This system presents an opportunity to observe the early stages of co-evolution of these two birds, potentially leading to the development of host defense mechanisms. We sought to determine whether the munias recognized whydahs as a threat, an antecedent to identifying and later counteracting the parasite. We conducted a playback experiment in which we exposed munias to both whydah and control (house finch) songs. The timed behaviors we observed were classified into either “reaction” or “no reaction” categories. Results from a paired Wilcoxon Signed-Rank test revealed that the whydah song produced a significantly higher “reaction” time than the control, mostly consisting of vigilant behavior. Our observations also detailed aggressive behavior from the munias towards the whydahs, which is remarkable as munias are not aggressive in their native range. These results suggest that munias have learned that pin-tailed whydahs are a threat, an example of behavioral plasticity leading to a frontline defense against brood parasitism. We were unable to determine whether munias are aggressive towards pin-tailed whydahs because they perceive them as a nest parasite, or because they are reacting to aggression by the whydahs in other contexts. Even so, we expect that these results will aid in addressing the gap in our understanding of the emergence of frontline defense mechanisms and provide the foundation for future research on this new and understudied host-parasite relationship.