Rhinehart, T.A.; Turek, D.; Kitzes, J. 2022. A continuous-score occupancy model that incorporates uncertain machine learning output from autonomous biodiversity surveys. Methods in Ecology and Evolution. doi.org/10.1111/2041-210X.13905.
Machine learning models are increasingly applied to autonomous sensing devices to survey wildlife biodiversity. Models generate continuous-score data reflecting confidence a species is present in each autonomously sensed file. However, these data are not directly compatible with traditional methods to model species occupancy based on binary detection/non-detection data. In this manuscript, we presented a new occupancy model that models continuous scores to estimate species occupancy and detectability.
Chronister, L.M., Rhinehart, T.A., Place, A., & Kitzes, J. 2021. An annotated set of audio recordings of Eastern North American birds containing frequency, time, and species information. Ecology 102 (6), e03329. doi.org/10.1002/ecy.3329.
Strongly labeled bioacoustic datasets (i.e., soundscape recordings containing the time and frequency boundaries of all biotic sounds) are rarely published. These datasets are useful not only for the study of the soundscape itself, but also for training and validating machine learning models for species identification. In this data paper, we presented a fully-labeled autonomous recording dataset encompassing 385 minutes of dawn chorus recordings, 48 species, and 16,052 annotations
Kitzes, J., Blake, R., ..., Rhinehart, T.A., ..., Yule, K. 2021. Expanding NEON biodiversity surveys with new instrumentation and machine learning approaches. Ecosphere 12(11), e03795. doi.org/10.1002/ecs2.3795.
While the National Ecological Observatory Network (NEON) extensively uses automated instruments to collect environmental data, NEON’s biodiversity surveys are almost entirely conducted using traditional human-centric field methods. In this manuscript, we review previous research at the intersection of biodiversity, instrumentation, and machine learning at NEON sites and expand on five methods for automated biodiversity measurement that could potentially be employed at NEON sites in future, such as acoustic recorders for sound-producing taxa and camera traps for medium and large mammals.
Rhinehart, T.A., Chronister, L.M., Devlin, T., Kitzes, J. 2020. Acoustic localization of terrestrial wildlife: Current practices and future opportunities. Ecology and Evolution, 10(13), 6794-6818. doi.org/10.1002/ece3.6216.
A specialized use of autonomous recording units is acoustic localization, in which a vocalizing animal is located spatially, usually by quantifying the time delay of arrival of its sound at an array of time-synchronized microphones. In this manuscript, we describe trends in acoustic localization literature, identify considerations for field biologists who wish to use these systems, and suggest advancements that will improve the field of acoustic localization.
Frontiers in Ornithology: Data & Discoveries in the World of Birds. 2022. Birding, January 2022, 53 (8), 18-21 (Link; Inti Tanager painting by Dan Lane).
In this article I covered the description of the Inti Tanager, Heliothraupis oneilli, including interviewing several of the scientists involved in the description.
Eavesdropping on Birds: Bird conservation powered by breakthroughs in machine learning. 2020. Birding, April 2020, 52 (2), 44-49 (Link).
In this article, I described my lab's work developing machine learning classifiers to identify bird species in autonomous recordings, and how we're looking to apply these techniques to further bird conservation.