Biology subjects

Miller, E. T.

Publications and source records attributed to Miller, E. T..

13 recordsLinked to original sources

bioRxiv2026

Large language models unlock the ecology of species interactions

Species interactions shape population dynamics, geographic distributions, evolutionary trajectories, and responses to environmental change. Yet data on these interactions remain scarce across broad spatial, temporal, and taxonomic scales because they are difficult to collect in the field. One promising source of interaction data is citizen science platforms, which contain billions of biodiversity observations, often accompanied by unstructured text comments that may document interactions among organisms. Advances in large language models (LLMs) make it increasingly feasible to identify, extract, and categorize biotic interactions from these unstructured data at scale. Here, we present an LLM workflow that collects species interaction observations from multilingual citizen science comments. Using two case studies--bird-bird and plant-pollinator interactions--we show that LLMs can rapidly extract interaction types and participating species with high accuracy. These data can greatly expand the spatial, temporal, and taxonomic coverage and resolution of species interactions data, enable new tests of long-standing ecological questions, and improve our ability to track ecological changes. With appropriate validation, expert review, and attention to data privacy for both users and sensitive species, this approach opens new opportunities to characterize, forecast, and conserve biodiversity under global change.

ecology↗

bioRxiv2025

The BirdsPlus Index, a novel method for assessing site-level conservation values.

While there is growing interest in sustainable management practices to mitigate the biodiversity impacts of agriculture, logging, and other critical societal needs, implementation of such practices is often hindered by a lack of cost-effective, fine-scale metrics that directly link management actions to conservation outcomes. We introduce the BirdsPlus Index (BPI), a novel, scalable approach that integrates monitoring data, remote sensing, and conservation-weighted species scores to quantify deviations of observed site scores from spatiotemporally explicit expectations. Using nearly 29,000 recordings from the Macaulay Library, we generated acoustic checklists with the Merlin and BirdNET sound identification models under multiple detection thresholds. We matched these acoustic checklists with species-specific conservation values (BirdsPlus species scores), then trained random forest models to predict total, site-level biodiversity (the sum of these species scores) given environmental and effort covariates. The resulting model also enabled us to map expected BirdsPlus site scores across the landscape. These scores integrate information on species conservation status, ecological roles, and phylogenetic and functional uniqueness. BPI was calculated as the residual between observed and expected site scores, thereby providing a direct site-level measure of conservation value. Across 30 sites, we found that BPI values were consistent across acoustic models and detection thresholds, with high-scoring sites supporting regionally uncommon breeders and habitat specialists. While acoustic- and observer-based (eBird) models showed differing spatial patterns, both aligned with known ecological drivers such as urban density, elevation, and wetland cover. Our results demonstrate that acoustic checklists can be used to model expected biodiversity over time and space, and that the BPI provides a robust, interpretable metric for evaluating the ecological integrity of local sites. Beyond its immediate application to conservation planning, this framework lays the foundation for global, real-time biodiversity monitoring that leverages automated acoustic classifiers, citizen science, and remote sensing to integrate conservation value into development and management decisions.

ecology↗

bioRxiv2025

Aligning conservation status, vulnerability factors, and ecological and evolutionary uniqueness to produce integrated assessments of the world's birds

A growing awareness, now enshrined in the Kunming-Montreal Global Biodiversity Framework, of the need to monitor biodiversity effectively at scale has led to a proliferation of novel solutions for doing so. Although global biodiversity encompasses all life, from tiny nitrogen-fixing bacteria to emergent rainforest trees, birds have several characteristics that make them a frequent focus of such monitoring efforts. In particular, birds frequently give diagnostic, species-specific vocalizations that simplify monitoring, they perform a number of critical ecosystem services, they are widely distributed in most ecosystems with strong representation on all continents, and the basic ecology, conservation status, populations, and distributions of many species is well known; birds thus provide a window into the underlying health and habitats of the systems under study. How best to summarize biodiversity monitoring results is a research question that has led to the development of approaches that incorporate species IUCN Red List threat assessments into site-level biodiversity scores. Notably, birds vocal behavior means that they can be effectively surveyed at scale with passive acoustic monitoring, and the potential to link such monitoring with automated identification and therefore quickly generate site-level biodiversity scores is an appealing approach to implement rigorous evaluations of global biodiversity. Yet, while many of the worlds birds are suffering worrisome population declines, the vast majority of species (78%) are still ranked "Least Concern" by the Red List. In an effort to develop a species scoring system that would be more conducive to such site-level valuations, we integrated key databases of species population status assessments, exposure to known vulnerability factors, and their functional and phylogenetic uniqueness to provide quantitative summaries of their conservation significance. We augmented these databases with two novel data sets available for most of the worlds birds: quantitative measurements of migration distances, and species-level phylogenetic and functional uniqueness values comparing each species to those it co-occurs with throughout its range. While the resulting BirdsPlus species scores also inherently reflect our own scientific expertise and judgement, our approach is transparent, dynamic, easily updated, and readily modified by users with different goals or values.

ecology↗

bioRxiv2024

A complete and dynamic tree of birds

We present a complete, time-scaled, evolutionary tree of the worlds bird species. This tree unites phylogenetic estimates for 9,239 species from 262 studies published between 1990 and 2024, using the Open Tree synthesis algorithm. The remaining species are placed in the tree based on curated taxonomic information. The tips of this complete tree are aligned to the species in the Clements Taxonomy used by eBird and other resources, and cross-mapped to other taxonomic systems including the Open Tree of Life (Open Tree), National Center for Biotechnology Information (NCBI), and Global Biodiversity Information Facility (GBIF). The total number of named bird species varies between 10,824 and 11,017 across the taxonomy versions we applied (v2021, v2022 and v2023). We share complete trees for each taxonomy version. The procedure, software and data-stores we used to generate this tree are public and reproducible. The tree presented here is Aves v1.2 and can be easily updated with new phylogenetic information as new estimates are published. We demonstrate the types of large scale analyses this data resource enables by linking geographic data with the phylogeny to calculate the regional phylogenetic diversity of birds across the world. We will release updated versions of the phylogenetic synthesis and taxonomic translation tables annually. The procedure we describe here can be applied to developing complete phylogenetic estimates for any taxonomic group of interest. Significance statementBirds are charismatic - well loved, and highly studied. Many new phylogenies elucidating avian birds evolutionary relationships are published every year. We have united phylogenetic estimates from hundreds of studies to create a complete evolutionary tree of all birds. While a variety of resources aggregate huge collections of trait, behavior and location data for birds, previously the barriers to linking data between these data resources and bird evolutionary history have limited the opportunities to do exciting large scale analyses. We have bridged that gap, and developed a system that allows us to easily update our understanding of bird evolution as new estimates are generated. We share a workflow and the software needed to create a complete evolutionary tree for any group.

evolutionary biology↗

bioRxiv2022

Molecular early burst associated with the diversification of birds at the K-Pg boundary

Complex patterns of genome and life-history evolution associated with the end-Cretaceous (K- Pg) mass extinction event limit our understanding of the early evolutionary history of crown group birds [1-9]. Here, we assess molecular heterogeneity across living birds using a technique enabling inferred sequence substitution models to transition across the history of a clade [10]. Our approach identifies distinct and contrasting regimes of molecular evolution across exons, introns, untranslated regions, and mitochondrial genomes. Up to fifteen shifts in the mode of avian molecular evolution map to rapidly diversifying clades near the Cretaceous-Palaeogene boundary, demonstrating a burst of genomic disparity early in the evolutionary history of crown birds [11-13]. Using simulation and machine learning techniques, we show that shifts in developmental mode [14] or adult body mass [4] best explain transitions in the mode of nucleotide substitution. These patterns are related, in turn, to macroevolutionary shifts in the allometric scaling relationship between basal metabolic rate and body mass [15, 16]. In agreement with theoretical predictions, this scaling relationship appears to have weakened across the end-Cretaceous transition. Overall, our study provides evidence that the Chicxulub bolide impact [17] triggered integrated patterns of evolution across avian genomes, physiology, and life history that structured the evolutionary potential of modern birds.

evolutionary biology↗

bioRxiv2022

Re-assessing niche partitioning in MacArthur's Warblers: foraging behavior, morphology, and diet metabarcoding in a phylogenetic context

Due in large part to MacArthurs classic 1958 paper, wood-warblers (Parulidae) are ecological icons, textbook protagonists of a story of competition and niche partitioning. As the story goes, subtle differences in foraging behavior are the principal means by which these nearly morphologically indistinguishable species are able to co-occur and avoid extinction. Yet, MacArthurs study was in fact quite limited in scale, and he said little about the relevance of evolution to the study system. Here, we reassess MacArthurs conclusions across an expanded set of syntopic warbler species in a forest in northern New York. We combine morphometrics, quantitative foraging data, and fecal metabarcoding--a direct measure of warbler diet--to study competition and niche partitioning in an evolutionary framework. We find close and kinematically realistic relationships between morphology and foraging behavior, but little connection between warbler ecomorphology and the 2,882 invertebrate taxa detected in their diets. Instead, diet remains phylogenetically conserved--closely related warblers eat similar suites of invertebrates, regardless of where they forage. Finally, we present evidence that these species not only partition niche space in the present day, but that competition has shaped their behaviors over evolutionary time. MacArthur (1958) may have drawn a few incorrect inferences, but his overall conclusion that evolved differences in foraging position, driven by competition among close relatives, does indeed appear to be a key reason these warblers can occur in such close sympatry.

evolutionary biology↗

bioRxiv2022

The effect of sociality on competitive interactions among birds

Sociality can provide many benefits, including increased foraging success, reproductive opportunities, and defence against predation. How does sociality influence the dominance hierarchies of ecological competitors? Here, we address this question using a large citizen science dataset of competitive interactions among birds foraging at backyard feeders, representing a network of over 55,000 interactions among 68 common species. We first show that species differ in average group size (the number of conspecifics observed together) as a fundamental measure of sociality. When analyzing heterospecific competition, we find that sociality is inversely related to dominance. On average, a single individual from a solitary species is more likely to displace a size-matched opponent than a single individual from a social species. Yet, we find that social species gain an increase in their competitive advantage when in the presence of their conspecifics, which may occur as a result of dynamics within their groups. Finally, we show that more social species have relatively fewer dominance interactions with heterospecifics, and more with conspecifics. Overall, these results demonstrate that sociality can influence competition in ecological networks. More social species have decreased competitive ability as individuals, but they may gain competitive ability in groups.

animal behavior and cognition↗

bioRxiv2022

Distilling complex evolutionary histories with shiftPlot

Phylogenies form the backbone of many modern comparative methods and are integral components of contemporary science communication. Recent years have seen drastic increases in both the size and complexity of phylogenetic data as computational resources and genetic/trait databases expand. Graphical representations of these massive phylogenetic datasets push against the limits of legibility, often veering closer to artwork than scientific figures optimized to communicate results. While attractive scientific illustrations are certainly a laudable goal, researchers may want to opt for simpler representations to communicate results more concisely. Here, we introduce a new R package, shiftPlot, which implements methods for simplifying and plotting phylogenetic comparative data on discrete traits. Specifically, shiftPlot automatically finds and collapses clades exhibiting the same character state, effectively creating smaller phylogenies that may be more legibly rendered on standard page sizes. Further, these visualizations more clearly communicate evolutionary dynamics by emphasizing state shifts over tip states. While there are undoubtedly situations where this graphical approach will not be suitable (e.g., continuous traits), we believe shiftPlot will prove useful for modern researchers faced with the task of communicating the results of complex phylogenetic analyses.

evolutionary biology↗

bioRxiv2022

Host phylogeny, but not diet diversity, influences gut microbiome composition in breeding wood-warblers

Understanding the factors that shape microbiomes can provide insight on the importance of host-symbiont interactions and on co-evolutionary dynamics. Unlike for mammals, previous studies have found little or no support for an influence of host evolutionary history on avian gut microbiome diversity and instead have suggested a greater influence of the environment or diet due to fast gut turnover. Because effects of different factors may be conflated by captivity and sampling design, examining natural variation using large sample sizes is important. Our goal was to overcome these limitations by sampling wild birds to compare environmental, dietary, and evolutionary influences on gut microbiome structure. We performed fecal metabarcoding to characterize both the gut microbiome and diet of fifteen wood-warbler species across a four-year period and from two geographic localities. We find host taxonomy generally explained [~]10% of the variation between individuals, which is [~]6-fold more variation of any other factor considered, including diet diversity. Further, gut microbiome similarity was more congruent with the host phylogeny than with host diet similarity and we found little association between diet diversity and microbiome diversity. Together, our results suggest evolutionary history is the strongest predictor of gut microbiome differentiation among wood-warblers. Although the phylogenetic signal of the warbler gut microbiome is not very strong, our data suggest that a stronger influence of diet (as measured by diet diversity) does not account for this pattern. The mechanism underlying this phylogenetic signal is not clear, but we argue host traits may filter colonization and maintenance of microbes.

evolutionary biology↗

bioRxiv2019

Does variation in glucocorticoid regulation predict fitness? A phylogenetic meta-analysis.

Glucocorticoid hormones (GCs) are central mediators of metabolism and the response to challenges. Because circulating levels of GCs increase in response to challenges, within-population variation in GCs could reflect individual variation in condition or experience. At the same time, individual variation in the degree to which GCs increase in response to challenges (which is relatively consistent within individuals over time) could have causal effects on stress coping capacity, and the ability to survive and reproduce. Although a number of studies in vertebrates have tested whether within-population variation in GCs predicts components of fitness, it is not clear whether there are consistent patterns across taxa. Here we present the first phylogenetic meta-analysis testing whether within-population variation in GCs is associated with components of fitness across vertebrates. At the same time, we introduce and test predictions about an overlooked but potentially important mediator of GC-fitness relationships: life history context. We suggest that strong context-dependence in the fitness benefit of maintaining elevated GCs could obscure consistent patterns between GCs and fitness across taxa. Meta-analyses revealed that across vertebrates, baseline and stress-induced GCs were consistently negatively correlated with reproductive success. This relationship did not differ depending on life history context. In contrast, the relationships between GCs and survival were highly context dependent, differing according to life history strategy. Both baseline and stress-induced GCs were more strongly negatively associated with survival in longer-lived populations and species. Stress-induced GCs were also more negatively associated with survival in organisms that engage in relatively more valuable reproductive attempts. Fecal GCs did not predict survival or reproductive success. We also used a meta-analytic approach to test whether experimental increases in GCs had consistent causal effects on fitness. Experimental increases in GCs reduced both survival and reproductive success, although the latter relationship was not significant when accounting for phylogeny. Overall, these results support the prediction that GC-fitness relationships can be strongly context dependent, and suggest that incorporating life history may be particularly important for understanding GC-survival relationships. Future studies that explore the role of other aspects of context (e.g., the nature and frequency of stressors, environmental variation) within and across species could provide important insights how and when variation in GC regulation predicts fitness.

evolutionary biology↗

bioRxiv2019

Best practices for making reliable inferences from citizen science data: case study using eBird to estimate species distributions

Citizen science data are valuable for addressing a wide range of ecological research questions, and there has been a rapid increase in the scope and volume of data available. However, data from large-scale citizen science projects typically present a number of challenges that can inhibit robust ecological inferences. These challenges include: species bias, spatial bias, and variation in effort. To demonstrate addressing key challenges in analysing citizen science data, we use the example of estimating species distributions with data from eBird, a large semi-structured citizen science project. We estimate two widely applied metrics of species distributions: encounter rate and occupancy probability. For each metric, we assess the impact of data processing steps that either degrade or refine the data used in the analyses. We also test whether differences in model performance are maintained at different sample sizes. Model performance improved when data processing and analytical methods addressed the challenges arising from citizen science data. The largest gains in model performance were achieved with: 1) the use of complete checklists (where observers report all the species they detect and identify); and 2) the use of covariates describing variation in effort and detectability for each checklist. Occupancy models were more robust to a lack of complete checklists and effort variables. Improvements in model performance with data refinement were more evident with larger sample sizes. Here, we describe processes to refine semi-structured citizen science data to estimate species distributions. We demonstrate the value of complete checklists, which can inform the design and adaptation of citizen science projects. We also demonstrate the value of information on effort. The methods we have outlined are also likely to improve other forms of inference, and will enable researchers to conduct robust analyses and harness the vast ecological knowledge that exists within citizen science data.

ecology↗

bioRxiv2018

Climate, habitat, and geographic range overlap drive plumage evolution

Organismal appearances are shaped by selection from both abiotic and biotic drivers 1-5. For example, Glogers rule describes the pervasive pattern that more pigmented populations are found in more humid areas 1,6,7, and substrate matching as a form of camouflage to reduce predation is widespread across the tree of life 8-10. Sexual selection is a potent driver of plumage elaboration 5,11, and species may also converge on nearly identical colours and patterns in sympatry, often to avoid predation by mimicking noxious species 3,4 To date, no study has taken an integrative approach to understand how these factors determine the evolution of colour and pattern across a large clade of organisms. Here we show that both habitat and climate profoundly shape avian plumage. However, we also find a strong signal that many species exhibit remarkable convergence not explained by these factors nor by shared ancestry. Instead, this convergence is associated with geographic overlap between species, suggesting strong, albeit occasional, selection for interspecific mimicry. Consequently, both abiotic and biotic factors, including interspecific interactions, are potent drivers of phenotypic evolution.

evolutionary biology↗

bioRxiv2017

Fighting over food unites the birds of North America in a continental dominance hierarchy

The study of aggressive interactions between species has, to date, usually been restricted to interactions among small numbers of ecologically close competitors. Nothing is known about interspecific dominance hierarchies that include numerous, ecologically varied species. Such hierarchies are of interest because they could be used to address a variety of research questions, e.g. do similarly ranked species tend to avoid each other in time or space, and what will happen when such species come into contact as climates change? Here, we propose a method for creating a continental-scale hierarchy, and we make initial analyses based on this hierarchy. We quantified the extent to which a dominance hierarchy of feeder birds was linear, as intransitivities can promote local species coexistence. Using the existing network of citizen scientists participating in Project FeederWatch, we collected the data with which to create a continent-spanning interspecific dominance hierarchy that included species that do not currently have overlapping geographic distributions. Overall, the hierarchy was nearly linear, and largely predicted by body mass, although there were clade-specific deviations from the average mass-dominance relationship. Most of the small number of intransitive relationships in the hierarchy were based on small samples of observations. Few observations were made of interactions between close relatives and ecological competitors like Melanerpes woodpeckers and chickadees, as such species often have only marginally overlapping geographic distributions. Yet, these species ranks--emergent properties of the interaction network--were usually in agreement with published literature on dominance relationships between them. Interspecific dominance hierarchy, aggression, displacement, citizen science\n\nLAY SUMMARYWhen it comes to fighting over food, bigger is better but woodpeckers are best. The outcome of aggressive encounters between birds frequently determines which individual gains access to contested resources like food, but until now, little was known about such encounters between individuals of different species. We partnered with citizen scientists to record interspecific behavioral interactions at bird feeders around North America, and assembled these interactions into a continental dominance hierarchy.

ecology↗