Unsupervised classification of Blanding’s turtle (Emydoidea blandingii) behavioural states from multi-sensor biologger data
Bibliographic record
Abstract
Classifying animal behaviors in their natural environments is both challenging and ecologically important, but the use of biologgers with multiple sensors has significantly advanced this research beyond the capabilities of traditional methods alone. Here, we show how biologgers containing an integrated tri-axial accelerometer, GPS logger and immersion sensor were used to infer behavioural states of a cryptic, freshwater turtle, the Blanding's turtle (Emydoidea blandingii). Biologgers were attached to three males and five females that reside in two undisturbed coastal marshes in northeastern Georgian Bay (Ontario, Canada) between May and July 2023. Raw acceleration values were separated into static and dynamic acceleration and subsequently used to calculate overall dynamic body acceleration (ODBA) and pitch. The unsupervised Hidden Markov Model (HMM) successfully differentiated five behavioural states as follows: active in water, resting in water, active out of water, resting in water, and nesting. Overall accuracy of the classification was 93.8%, and except for nesting (79%), all other behaviours were above 92%. There were significant differences in daily activity budgets between male and female turtles, with females spending a greater proportion of time active out of water, and inactive out of the water, while males spent a greater proportion of time active in water. These differences were likely a result of large seasonal life-history requirements such as nesting and mate finding. Accurate classification of behavioural states is important for researchers to understand fine-scale activities carried out during the active season and how environmental variables may influence the behaviours of turtles in their natural habitats.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".