My work pairs pressing conservation questions with the quantitative methods needed to answer them, often in systems and taxa that would otherwise go understudied. Below, my main research areas are followed by the methodological approaches that cut across them.
Research Areas
Methods for studying biodiversity at scale
Biodiversity data are growing exponentially, yet coverage remains deeply uneven across taxa, regions, and time—and much of what we most need to observe still lies beyond the reach of traditional fieldwork. I build the methods to close both gaps. On the sensing side, I develop remote, non-invasive approaches—passive acoustic monitoring, computer vision, and radar—to observe wildlife and quantify animal movement and cave emergences at scales fieldwork cannot reach. I complement these with endogenous markers (stable isotopes) which can forensically infer animal diet and geographic origins, and scale from individual samples to population-level inference.
On the data side, I develop methods and tools to harness participatory-science platforms like iNaturalist, diagnose biases in global biodiversity databases, and fill knowledge gaps in under-studied regions. A central future direction is quantifying global biodiversity sampling completeness across the Global Biodiversity Information Facility (GBIF), to find where ecological knowledge and evidence-based conservation policy is most constrained by missing data.
-
Remote rainforest inventorying and human-in-the-loop ML for bat biodiversity survey — Team Limelight. XPRIZE grand-prize winner
I led bat bioacoustics and bioinformatics for Team Limelight (Limelight Rainforest), grand-prize winner of the $5M XPRIZE Rainforest competition (2024). A manuscript is in preparation.
XPRIZE Rainforest Team Limelight -
Campbell, CJ; et al. (2023). Identifying the identifiers: how iNaturalist facilitates collaborative, research-relevant data generation.
BioScience 73(7):533–541.Showed that a small set of highly active, expert users drives most research-grade identifications on iNaturalist, a pattern that matters for how community-science data are used.
Article PDF -
Grady, E; Campbell, CJ; Callaghan, C; Guralnick, R. (2026). iNaturalist users exhibit distinct spatiotemporal sampling preferences, with implications for biodiversity science and project planning.
Citizen Science: Theory and Practice 11(1):3.Found that when and where users choose to record wildlife systematically shapes the biodiversity datasets those records become.
Article -
McCleery, R; Guralnick, R; Beatty, M; Belitz, M; Campbell, CJ; et al. (2023). Uniting experiments and big data to advance ecology and conservation.
Trends in Ecology & Evolution 38(10).Set out how controlled experiments and large observational datasets can complement each other rather than compete.
Article PDF
-
Weather-surveillance radar to monitor mass-emerging bat activity. Current project
Using weather-surveillance radar to detect and quantify the activity of mass-emerging bat colonies across broad regions.
-
Bioacoustics and computer-vision monitoring of bat emergence dynamics. Current project
Combining acoustic recordings and computer-vision video counts to resolve the timing and intensity of nightly bat emergences.
Movement & ecology of understudied taxa
Seasonal migration is a globally ubiquitous process that governs ecosystem structure and links distant regions through the exchange of genes, energy, and disease—yet for many animals the where and how of that movement remain undetected. My research spans bats, birds, and insects to reveal these hidden dynamics and their consequences for conservation. Because wind-energy development disproportionately affects migratory species, I study how migratory behavior drives fatality risk at wind-energy facilities and develop tools to target interventions toward the times and places migrants are most likely to be present.
Bats are a particular focus: they are among the most diverse and least-studied groups of mammals, and much of their basic natural history remains undescribed. I work to fill those gaps, focusing on migratory routes, seasonal ranges, roosting and emergence behavior, diet and trophic ecology, and responses to a changing climate and threats like white-nose syndrom. This work integrates data across scales, from historical museum specimens to community-science observations and automated video counts of nightly emergences.
-
Campbell, CJ; et al. (2025). Migratory strategy is a key factor driving interactions at wind energy facilities in at-risk North American bats.
Ecology Letters 28:e70202.Showed that how a bat migrates, not just where it lives, governs its overlap with and fatality risk at wind-energy facilities.
Article PDF -
Wilson, J; True, M; Campbell, CJ. (2025). Autumn migration to higher latitudes in Seminole bats (Lasiurus seminolus) redefines seasonal ranges.
Ecology & Evolution 15:e71657.Used stable isotopes and fatality data to show that Seminole bats head north in autumn, redrawing the species' known range.
Article PDF -
Carpenter, B; Campbell, CJ; Fanning, A; McBride, M. (2024). Migratory mixing of Gallinago delicata (Wilson's snipe) in wintering areas highlights the need for international coordination for monitoring and management.
Ornithological Applications duae064.Found that wintering snipe mix widely across breeding origins, so monitoring and management need international coordination.
Article PDF -
Campbell, CJ; et al. (2022). White-nose syndrome pathogen Pseudogymnoascus destructans detected in migratory tree-roosting bats.
Journal of Wildlife Diseases .Detected the white-nose syndrome pathogen on migratory bats outside winter, including hoary bats, widening where the disease may spread.
Article PDF -
Smith, LM; Gore, JA; Doonan, TJ; Campbell, CJ. (2022). Tricolored bats at a southern range edge exhibit partial migration northward in autumn.
Movement Ecology 10(1):1–13.Found that only part of a southern tricolored-bat population migrates north in autumn, a subtlety that matters for management.
Article PDF -
Campbell, CJ; Nelson, DM; Ogawa, NO; Chikaraishi, Y; Ohkouchi, N. (2017). Trophic position and dietary breadth of bats revealed by nitrogen isotopic composition of amino acids.
Scientific Reports 7:15932.Used amino-acid nitrogen isotopes to read the trophic position and dietary breadth of bats from their tissues.
Article PDF
Science-to-conservation
Research protects biodiversity only when it informs decision-making, and much of my work aims to close that gap. A central thread is renewable energy and transmission: the shift to renewables is essential, yet its biodiversity impacts fall disproportionately on migratory bats and birds. By resolving which species are most at risk, and when and where that risk is concentrated, my research helps target the times and places where curtailment and other interventions do the most good. I apply the same translational approach to habitat management, developing concrete guidance for managing power-line corridors; and to continental-scale syntheses, including an assessment of the conservation status of North America's bats.
Because the species I study move across state and national lines, effective conservation rarely stops at a single jurisdiction. My isotope-based work on rosy-finches and Wilson's snipe connectivity shows populations mixing and cross national borders widely between seasons, so safeguarding them depends on coordination among the states and countries that share them. Looking ahead, I am increasingly interested in conservation that operates at the level of whole assemblages rather than one species at a time, which is a more scalable response to a biodiversity crisis that is unfolding at a global scale.
-
-
-
-
-
Belitz, M; Campbell, CJ; et al. (2025). A case for assemblage-level conservation to address the biodiversity crisis.
Nature Reviews Biodiversity 1:134–143.Argued for protecting whole communities of species, not just one at a time, as a scalable strategy for conservation policy and practice.
Article PDF
Methodological Approaches
Spatial Modeling
Understanding how organisms interact with, and are limited by, their environment is a central question of ecology. Correlative species distribution models (SDM) can be used to project the expected past, current, or future geographic distributions of a species based on the environmental conditions where they are known to occur. I am particularly interested in leveraging big data approaches to understand and identify animal migration at the population level given shifts in seasonal distributions.
-
Rubin, J; Campbell, CJ; et al. (2025). Strong bat predation and weak environmental constraints predict longer moth tails.
Proceedings of the Royal Society B 292(2046).I built species distribution models for 157 of the world's insectivorous bats to map global bat-predator richness, then used it to show that bat predation likely drove the evolution of the elaborate hindwing tails of Actias moon moths.
Article PDF Code -
Campbell, CJ; et al. (2025). Migratory strategy is a key factor driving interactions at wind energy facilities in at-risk North American bats.
Ecology Letters 28:e70202.Modeled the seasonal distributions and movements of at-risk North American bats and showed that migratory strategy, not geographic range alone, governs their overlap with, and fatality risk at, wind-energy facilities.
Article PDF -
Wilson, J; True, M; Campbell, CJ. (2025). Autumn migration to higher latitudes in Seminole bats (Lasiurus seminolus) redefines seasonal ranges.
Ecology & Evolution 15:e71657.Used stable-isotope and wind-energy fatality data to show that Seminole bats migrate to higher latitudes in autumn, redrawing the species' known seasonal range.
Article PDF
Data Science
The availability of biodiversity data is increasing exponentially. I research its generation and develop methods to help integrate data from multiple sources and account for data biases.
-
Grady, E; Campbell, CJ; Callaghan, C; Guralnick, R. (2026). iNaturalist users exhibit distinct spatiotemporal sampling preferences, with implications for biodiversity science and project planning.
Citizen Science: Theory and Practice 11(1):3.Found that iNaturalist users have distinct spatiotemporal sampling preferences that shape where and when biodiversity data accumulate, with practical implications for project design and bias correction.
Article -
Idec, J; Campbell, CJ; Belitz, M; Vinod Anand, A; Guralnick, R. (2025). Using citizen science data to estimate trait and climate drivers of daily activity patterns in temperate butterflies.
PLOS One 20(11):e0335856.Used iNaturalist observations to estimate how species traits and climate drive the daily activity patterns of temperate butterflies.
Article -
McCleery, R; Guralnick, R; Beatty, M; Belitz, M; Campbell, CJ; et al. (2023). Uniting experiments and big data to advance ecology and conservation.
Trends in Ecology & Evolution 38(10).Made the case for uniting controlled experiments with big observational data, showing how the two paradigms complement each other to advance ecology and conservation.
Article PDF -
Campbell, CJ; et al. (2023). Identifying the identifiers: how iNaturalist facilitates collaborative, research-relevant data generation.
BioScience 73(7):533–541.Showed that a small, highly active community of expert identifiers generates most research-grade records on iNaturalist—clarifying how research-relevant biodiversity data are produced.
Article PDF
Machine Learning & Quantitative Methods
Modern ecology increasingly depends on extracting reliable inference from data that are large, noisy, and imperfectly sampled. I build machine-learning and statistical models to meet that challenge. This includes computer-vision pipelines that detect and count individual animals in video (for example, quantifying bat emergences frame by frame), and human-in-the-loop classifiers for bioacoustic data, developed as part of the XPRIZE Rainforest grand prize–winning Limelight Rainforest team. I pair these with hierarchical and Bayesian models that make uncertainty and sampling structure explicit rather than assumed.
-
Remote rainforest inventorying and human-in-the-loop ML for bat biodiversity survey — Team Limelight. XPRIZE grand-prize winner
Built a human-in-the-loop machine-learning pipeline for bat bioacoustics as part of Team Limelight (Limelight Rainforest), grand-prize winner of the $5M XPRIZE Rainforest competition (2024). A manuscript is in preparation.
XPRIZE Rainforest Team Limelight -
Campbell, CJ; Fitzpatrick, MC; Vander Zanden, H; Nelson, DM. (2020). Advancing interpretation of stable isotope assignment maps.
Animal Migration .Developed and validated methods to compare and summarize probabilistic origin assignments, released as the open-source isocat R package.
Article PDF Code -
Rubin, J; Campbell, CJ; et al. (2025). Strong bat predation and weak environmental constraints predict longer moth tails.
Proceedings of the Royal Society B 292(2046).Fit species distribution models for 157 insectivorous bats to build a global map of predator richness for downstream analysis.
Article PDF Code
-
Computer-vision detection and counting of bat emergences. Current project
Developing computer-vision pipelines that detect and count individual bats in video to quantify nightly cave emergences frame by frame.
Remote Monitoring & Bioacoustics
Understanding wildlife at scale means observing animals where and when traditional field methods fall short. I develop and apply remote, non-invasive monitoring (passive acoustic sensing and bioacoustic detection, camera- and video-based counting, and weather-surveillance radar) to measure activity, abundance, and movement across broad areas and long time spans. These tools turn otherwise-inaccessible phenomena, from nightly cave emergences to continental-scale aerial movements, into data we can analyze.
-
Passive acoustic monitoring of bat emergence dynamics. Current project
Using full-spectrum recordings to characterize the timing and intensity of nightly bat emergences, from near-silence to peak activity.
-
Weather-surveillance radar for broad-scale aerial movement. Current project
Adapting radar remote sensing to observe the movement of bats and other aerial wildlife across broad regions and long time spans.
Stable Isotopes & Endogenous Markers
Animal tissues reflect the components synthesized during their formation. Put another way, you are what you eat. Analyzing the stable isotope ratios of inert tissues like fur can reveal the past conditions under which they were synthesized, yielding information about the geographic origin and diet of an animal when the tissue was formed. I use stable hydrogen analysis of animal tissues to study broad-scale migratory strategy, and I have developed and validated new techniques to help scale that understanding from individuals to populations.
-
Campbell, CJ; Fitzpatrick, MC; Vander Zanden, H; Nelson, DM. (2020). Advancing interpretation of stable isotope assignment maps.
Animal Migration .Introduced ways to compare and summarize probabilistic origin assignments from known-provenance bats, packaged as the open-source isocat R package.
Article PDF Code -
Campbell, CJ; Nelson, DM; Ogawa, NO; Chikaraishi, Y; Ohkouchi, N. (2017). Trophic position and dietary breadth of bats revealed by nitrogen isotopic composition of amino acids.
Scientific Reports 7:15932.Used compound-specific amino-acid nitrogen isotopes to read bat trophic position and dietary breadth directly from tissue.
Article PDF -
Carpenter, B; Campbell, CJ; Fanning, A; McBride, M. (2024). Migratory mixing of Gallinago delicata (Wilson's snipe) in wintering areas highlights the need for international coordination for monitoring and management.
Ornithological Applications duae064.Applied feather hydrogen isotopes to show that wintering snipe draw from a wide sweep of breeding origins.
Article PDF -
Software Development
I am passionate about developing open-source software to promote accessible and reproducible science.
-
SDMetrics — Metric to compare temporally-explicit species distribution models (2022). Maintainer
An R package providing a quantitative metric for comparing species distribution models that shift over time.
Code -
-