| Popis: |
Advancements in biologging technology allow terabytes of data to be collected that record the location of individuals but also their direction, speed and acceleration. These multi-stream data sets allow researchers to determine movement and behavioural patterns at high spatiotemporal resolutions and in turn quantify fine-scale changes in state along with likely ecological causes and consequences. The scope offered by such data sets is only increasing there is potential to gain unique insights into a suite of ecological and life history phenomena. We use multi-stream data from global positioning system (GPS) and accelerometer (ACC) tags to quantify breeding events remotely in an Arctic breeding goose. From a training set of known breeders we determine the movement and energy expenditure patterns indicative of incubation and use these to classify breeding events in individuals with unknown reproductive status. Given that researchers are often constrained by the amount of biologging data they can collect due to tag weights, we carry out a sensitivity analysis. Here we explore the relative merits of GPS vs ACC data and how varying the temporal resolution of the data affects the accuracy of classifying incubation. Classifier accuracy deteriorates as the temporal resolution of GPS and ACC are reduced but the reduction in precision (false positive rate) is larger in comparison to recall (false negative rate). Precision fell to 95.5% were as recall rarely fell below 98% over all sampling schedules tested. Our dataset could have been reduced by c.95% while maintaining precision and recall >98%. The GPS-only classifier generally outperformed the ACC-only classifier across all accuracy metrics but both performed worse than the combined GPS and ACC classifier. GPS and ACC data can be used to reconstruct breeding events remotely, allowing unbiased, 24-hour monitoring of individuals. Our resampling-based sensitivity analysis of classifier accuracy has important implications with regards to both tag design and sampling schedules for study systems where tag size is constrained. It will allow researchers with similar aims to optimize tag battery, memory usage and lifespan to maximise the ability to correctly quantify life history events. |
