Evaluating the accuracy of wolverine identification from photographs of snow tracks by expert observers in North America
Bibliographic record
Abstract
North American wolverines Gulo gulo are a species of conservation concern across much of their range. The remote and rugged terrain they occupy has led to the development of various remote detection methods. While visual identification of wolverine tracks has a long history of use, it has recently been considered less reliable than other remote identification methods. We evaluated the ability of 29 observers to identify wolverine from photographs of snow tracks. Each observer reviewed 99 observations: 48 of wolverine and 51 of species whose tracks are confusable for wolverine. The identification of each observation used was independently verified by visual observation of the animal that made the tracks or genetic samples collected from the tracks that confirmed field identification. We compared the performance of observers based on their demographic characteristics and the details present in each observation. Observers demonstrated a low false‐positive error rate, cumulatively misidentifying observations of other species as definitively wolverine less than 1% of the time. The only clear demographic predictor of an observer's skill was their level of tracking certification via Cybertracker Conservation. Higher certification levels produced more positive detections of wolverine than did low or no certification. Observations with clear details in individual tracks produced the highest rate of positive detections. Depending on the level of certainty expected, observers detected 75–88% of wolverine observations that contained clear details in the tracks and clear track patterns, with a false‐positive error rate of 1.1–3.5%. While the error rate did not increase, positive detections occurred less in observations without morphological details in individual tracks or clear track patterns. Observers called these ‘unknown' more often. Our results indicate that observers can consistently distinguish wolverine from other species from photographs of snow tracks and refute the assumption that visual track identification is inherently unreliable.
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 | 0.001 |
| 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) | 0.000 | 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".