A field test of R package <scp>GPSeqClus</scp> : For establishing animal location clusters
Bibliographic record
Abstract
Abstract The ability to track animals with Global Positioning System (GPS) collars opened an enormous potential for studying animal movements and behaviour in their natural environment. One such endeavour is to identify clusters of GPS locations as a way to estimate predator kill rate. Clapp et al. (2021) developed an R package ( GPSeqClus ) to assess a location dataset based on user‐defined parameters to identify clusters and their characteristics. These characteristics can then help to distinguish resting‐site clusters from kill sites of their large (>50 kg) prey. We identified location clusters of an adult male wolf Canis lupus on Ellesmere Island, Nunavut, Canada in July 2009 and tracked him until he died in April 2010. Identifying location clusters was challenging because the collar only obtained two GPS locations per day (12 h apart). In July 2010, we searched 30 of 52 location‐clusters we identified as kill/scavenge sites and found 17 of them as such, given they had muskox Ovibos moschatus or caribou Rangifer tarandus pearyi remains nearby. We also documented five wolf rendezvous sites, two den sites, and the wolf's death site to total 60 location‐clusters in all. We used a two‐step process in testing the R Package GPSeqClus (hereafter GPSeqClus ): (1) compare the number of clusters our method discerned with the number identified by the new algorithm, and (2) compare the number of biologically significant clusters (e.g. den sites, kill/feeding sites) we found with the number the new algorithm located. We made these tests with GPSeqClus by varying the search radius, number of days at a site, and minimum number of locations required for a cluster. GPSeqClus compared well to our technique, with the best sub‐algorithm among the 25 we tested only missing three of our identified clusters and yielding six additional clusters. GPSeqClus identified 16 of the 17 confirmed sites of remains, all wolf home sites, and the wolf's carcass site. Identifying clusters using a 500‐m search radius, a 1.5‐day window, and a minimum of two GPS locations per cluster was suitable for a coarse GPS acquisition rate of two locations per day when prey are large, such as muskox or caribou. Given that GPSeqClus performed well with our coarse location dataset, we expect it will also perform even better with a collar acquiring more than two locations per day. Having a field‐tested utility such as GPSeqClus will enhance carnivore predation studies elsewhere.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.070 | 0.237 |
| Meta-epidemiology (narrow) | 0.003 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.004 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.003 | 0.005 |
| Open science | 0.005 | 0.004 |
| Research integrity | 0.002 | 0.005 |
| Insufficient payload (model declined to judge) | 0.011 | 0.009 |
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 source (direct Gemma or distilled Codex), 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".