Geology topics

Molly Astell

Publications and source records attributed to Molly Astell.

3 recordsLinked to original sources

USGS2025

Modeling individual-level and population-level nest success of California Condors from movement data

The California Condor ( Gymnogyps californianus ) is a critically endangered species with populations that are not currently self-sustaining. Although understanding nest success is key to understanding trends in their populations, field monitoring of condor nests has become increasingly challenging as the number of nesting condors has increased and their range has expanded. We investigated whether California Condor nest fate could be accurately estimated from telemetry data with limited field observations. Our study focused on the southern California population of California Condors (2015–2022), and we used a recently published Bayesian hierarchical modeling framework that combines movement data and occasional field observations to estimate individual-level and population-level nest success. The model detected shifts in space use to categorize if each nest failed or if a young fledged. Estimated model parameters suggested that after nest failure, condors shifted toward more expansive space use. Additional field observations, not included as data in the model, provided evidence that we accurately categorized nest fate for 63 out of 65 California Condor nesting attempts. Finally, we scaled individual-level reproductive success to estimate annual population-level nesting success. These methods offer managers a way to reduce field monitoring efforts while still allowing for estimation of nest success, which will be key as the breeding populations of California Condors continue to grow and become more widely spread across the landscape.

California↗

USGS2023

Supervised versus unsupervised approaches to classification of accelerometry data

Sophisticated animal-borne sensor systems are increasingly providing novel insight into how animals behave and move. Despite their widespread use in ecology, the diversity and expanding quality and quantity of data they produce have created a need for robust analytical methods for biological interpretation. Machine learning tools are often used to meet this need. However, their relative effectiveness is not well known and, in the case of unsupervised tools, given that they do not use validation data, their accuracy can be difficult to assess. We evaluated the effectiveness of supervised ( n = 6), semi-supervised ( n = 1), and unsupervised ( n = 2) approaches to analyzing accelerometry data collected from critically endangered California condors ( Gymnogyps californianus ). Unsupervised K-means and EM (expectation–maximization) clustering approaches performed poorly, with adequate classification accuracies of <0.8 but very low values for kappa statistics (range: −0.02 to 0.06). The semi-supervised nearest mean classifier was moderately effective at classification, with an overall classification accuracy of 0.61 but effective classification only of two of the four behavioral classes. Supervised random forest (RF) and k-nearest neighbor (kNN) machine learning models were most effective at classification across all behavior types, with overall accuracies >0.81. Kappa statistics were also highest for RF and kNN, in most cases substantially greater than for other modeling approaches. Unsupervised modeling, which is commonly used for the classification of a priori-defined behaviors in telemetry data, can provide useful information but likely is instead better suited to post hoc definition of generalized behavioral states. This work also shows the potential for substantial variation in classification accuracy among different machine learning approaches and among different metrics of accuracy. As such, when analyzing biotelemetry data, best practices appear to call for the evaluation of several machine learning techniques and several measures of accuracy for each dataset under consideration.

Ecology and Evolution↗

USGS2017

Lack of observed movement response to lead exposure of California condors

Lead poisoning is an important conservation concern for wildlife, and scavenging birds are especially at risk from consumption of carcasses of animals killed with lead ammunition. Because current methods to identify lead exposure require animal capture and blood collection, management would benefit from the development of a less costly and noninvasive behavioral test for illness in wild animals. We attempted to design such a test to identify lead exposure in California condors ( Gymnogyps californianus ) that we tracked with global positioning system (GPS) telemetry in southern California, USA, 2013–2016. We measured blood-lead concentrations in tracked birds and expected that flight behavior would be influenced by lead exposure; thus, we measured the effect of blood-lead concentrations on 2 different types of movement rates and on the proportion of time condors spent in flight. We found no effect of lead exposure on any of these 3 behavioral metrics. Our work suggests that the measurements we took of flight behaviors were not a useful tool in predicting lead exposure in the mildly to moderately exposed birds we studied. Wild birds are effective at hiding illness, especially condors who have a strong social hierarchy in which showing weakness is a disadvantage. However, focusing on behaviors other than flight, expanding the sample studied to include birds with a wider range of lead concentration values, or analyzing tissues such as feathers (rather than, or in addition to, blood) may be more useful for identification of lead exposure and other diseases that may limit wildlife populations. © 2017 This article is a U.S. Government work and is in the public domain in the USA.

Journal of Wildlife Management↗