Biology subjects

Beasley, J. C.

Publications and source records attributed to Beasley, J. C..

4 recordsLinked to original sources

bioRxiv2025

Generalizing animal movement predictions across landscapes: a scalable framework grounded in empirical telemetry data

BackgroundUnderstanding animal movement across broad ecological contexts is essential for guiding management decisions in natural resource and animal disease management. However, most movement studies collect data over small spatial extents, which challenges using the data for landscape-scale management applications. MethodsWe collated relocation data from 564 wild pigs spanning 25 studies across 10 states in the United States. We derived net and daily displacement distributions for individuals in each of 4 seasons. We fit gamma distributions to displacement distributions and derived mean and dispersion parameters as metrics of displacement magnitude and variation. We modeled mean and dispersion variables using gradient-boosted regression, including landscape and weather predictors, and predicted daily and net displacement distributions across the species range. We used integrated step selection analysis to estimate habitat selection for wild pig groups sharing a common composition of habitat availability across the studied populations. ResultsFemale average ({+/-} 95% confidence interval) net displacement means varied by season, peaking in Jul-Sept at 1,152 {+/-} 140 m along with daily displacement mean distances of 650 {+/-} 41 m. Average net displacement means for males peaked in Oct.-Dec. at 1,651 {+/-} 160 m with daily displacement mean distances of 1,092 {+/-} 95 m. Extreme net displacement values ranged up to 9,649 m and 4,080 m for male and female, respectively, and extreme daily displacement distribution means of 6,616 m for males, and 3,855 m for females. Almost half of the individuals included a similar composition of habitat availability, regardless of study site location. Herbaceous wetland was the most consistently favored land type and was available for more than 70% of all individual pigs across all studies. Preference for other land types depended strongly on habitat availability. ConclusionsBy comparing movement and habitat selection across a large geographic extent, our work provides new insights about the environmental drivers and limits of wild pig movement behavior. Our workflow digests variation wild pig movement distributions and habitat selection into parameters that can be used to predict movement on heterogeneous landscapes for downstream applications such as invasive species elimination campaigns or disease management.

ecology↗

bioRxiv2025

Wild raccoons are more innovative but not bolder than another ecological generalist, the Virginia opossum, on a pull-string task

Humans are altering natural environments at unprecedented rates. Ecological generalism is one of the strongest predictors of survival in light of these changes which, in animals, may be shaped by bold and innovative behaviours. Species with greater habitat generalism are proposed to behave more boldly (e.g., touching and ingesting novel food). Species with greater dietary generalism are proposed to behave more innovatively (e.g., novel problem-solving to access food). Support for both hypotheses exists but remains largely restricted to broad comparisons between generalists and specialists. Further comparative data are needed to understand the extent to which these behavioural patterns might shape more nuanced ecological variation between species, such as species with finer-scale niche differences. We compared bold and innovative behaviour in two wild sympatric generalists, the Northern raccoon (Procyon lotor) and Virginia opossum (Didelphis virginiana), using a pull-string task with strings and wires attached to vertical cups containing high-value food. Although both species are broadly classified as generalists, raccoons show relatively greater habitat and dietary flexibility than opossums. Because they live sympatrically, it offers a unique opportunity to test - at the same time and locations - whether these finer-scale niche differences are reflective of differences in their bold and innovative behaviour. We predicted that raccoons would display bolder behaviour (in terms of touching our novel task) and more innovative behaviour (in terms of solving it). We found that both species were equally likely to acknowledge and touch the task, but raccoons were more likely to display innovation to access the food. Raccoons propensity for using innovation to solve novel foraging challenges may contribute to their greater ecological flexibility compared to opossums. More broadly, our findings may help explain differences in how each species adapts to environmental changes. HighlightsO_LIBold and innovative behaviour may help some species adapt to environmental changes C_LIO_LIWe administered a novel pull-string task to wild raccoons and opossums C_LIO_LIRaccoons displayed more innovative, not bolder, behaviour than opossums C_LIO_LIBoth species are generalists, but raccoons are more flexible than opossums C_LIO_LIDifferences in innovation may shape their finer-scale niche differences C_LI

animal behavior and cognition↗

bioRxiv2020

Improving the accessibility and transferability of machine learning algorithms for identification of animals in camera trap images: MLWIC2

O_LIMotion-activated wildlife cameras (or "camera traps") are frequently used to remotely and non-invasively observe animals. The vast number of images collected from camera trap projects have prompted some biologists to employ machine learning algorithms to automatically recognize species in these images, or at least filter-out images that do not contain animals. These approaches are often limited by model transferability, as a model trained to recognize species from one location might not work as well for the same species in different locations. Furthermore, these methods often require advanced computational skills, making them inaccessible to many biologists. C_LIO_LIWe used 3 million camera trap images from 18 studies in 10 states across the United States of America to train two deep neural networks, one that recognizes 58 species, the "species model," and one that determines if an image is empty or if it contains an animal, the "empty-animal model." C_LIO_LIOur species model and empty-animal model had accuracies of 96.8% and 97.3%, respectively. Furthermore, the models performed well on some out-of-sample datasets, as the species model had 91% accuracy on species from Canada (accuracy range 36-91% across all out-of-sample datasets) and the empty-animal model achieved an accuracy of 91-94% on out-of-sample datasets from different continents. C_LIO_LIOur software addresses some of the limitations of using machine learning to classify images from camera traps. By including many species from several locations, our species model is potentially applicable to many camera trap studies in North America. We also found that our empty-animal model can facilitate removal of images without animals globally. We provide the trained models in an R package (MLWIC2: Machine Learning for Wildlife Image Classification in R), which contains Shiny Applications that allow scientists with minimal programming experience to use trained models and train new models in six neural network architectures with varying depths. C_LI

ecology↗

bioRxiv2018

Machine learning to classify animal species in camera trap images: applications in ecology

1. Motion-activated cameras (\"camera traps\") are increasingly used in ecological and management studies for remotely observing wildlife and have been regarded as among the most powerful tools for wildlife research. However, studies involving camera traps result in millions of images that need to be analyzed, typically by visually observing each image, in order to extract data that can be used in ecological analyses.\n\n2. We trained machine learning models using convolutional neural networks with the ResNet-18 architecture and 3,367,383 images to automatically classify wildlife species from camera trap images obtained from five states across the United States. We tested our model on an independent subset of images not seen during training from the United States and on an out-of-sample (or \"out-of-distribution\" in the machine learning literature) dataset of ungulate images from Canada. We also tested the ability of our model to distinguish empty images from those with animals in another out-of-sample dataset from Tanzania, containing a faunal community that was novel to the model.\n\n3. The trained model classified approximately 2,000 images per minute on a laptop computer with 16 gigabytes of RAM. The trained model achieved 98% accuracy at identifying species in the United States, the highest accuracy of such a model to date. Out-of-sample validation from Canada achieved 82% accuracy, and correctly identified 94% of images containing an animal in the dataset from Tanzania. We provide an R package (Machine Learning for Wildlife Image Classification; MLWIC) that allows the users to A) implement the trained model presented here and B) train their own model using classified images of wildlife from their studies.\n\n4. The use of machine learning to rapidly and accurately classify wildlife in camera trap images can facilitate non-invasive sampling designs in ecological studies by reducing the burden of manually analyzing images. We present an R package making these methods accessible to ecologists. We discuss the implications of this technology for ecology and considerations that should be addressed in future implementations of these methods.

ecology↗