# The role of demographic and environmental variables on the presence of snow tracks by river otters Lontra canadensis

> If a river otter is right there, will you find their tracks in the snow? Following five radio-tagged otters on frozen lakes in British Columbia, researchers found tracks far more often for males than females, and more often the farther an otter had moved. One young female left tracks at fewer than one in five of her locations, a warning that snow-track surveys can miss part of a population.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2012-crowley-johnson-role-demographic-environmental-variables/

## Citation and access

- **Authors:** Shannon Crowley, Chris J. Johnson and Dexter Hodder
- **Year published:** 2012
- **Journal or source:** Wildlife Biology (Short communication)
- **Full citation:** Crowley, S., Johnson, C. J. and Hodder, D. 2012. The role of demographic and environmental variables on the presence of snow tracks by river otters Lontra canadensis. Wildlife Biology 18(1): 105–112.
- **DOI:** 10.2981/10-069
- **Access status:** Hosted (freely available elsewhere)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2012_Crowley_and_Johnson_etal_the-role-of-demographic-and-environmental-variables-on-the-presence-of-snow-tracks-by-river-otters.pdf

## Tags

- **Species:** Mammals > North American river otter
- **Region:** North America > Canada
- **Tracking:** Snow tracks; Spoor age estimation; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Daily travel distance; Movement
- **Methods and technology:** Akaike Information Criterion (AIC); Detectability; Logistic regression; Radio telemetry
- **Data type:** New field data
- **Observer qualification:** Not reported
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 2012

## Detailed keywords

- **Species:** Studied: North American river otter (Lontra canadensis), five radio-marked individuals (one juvenile male, two adult males, one juvenile female, one adult female). Mentioned for comparison only: Eurasian otter (Lutra lutra), wolverine (Gulo gulo).
- **Location and habitat:** Tezzeron and Pinchi Lakes, in and next to the John Prince Research Forest (13,000 ha, co-managed by the University of Northern British Columbia and Tl’azt’en Nation), 45 km northwest of Fort St. James, central British Columbia, Canada. Subboreal spruce zone; rolling, low mountains (700 to 1,267 m above sea level) with many lakes, rivers and streams. Frozen lakeshores in winter, with only one stretch of open water; otters mostly reached water through near-shore burrows and small holes under the snow. Long history of logging; a closed mercury mine and some homes at Pinchi Lake.
- **Topics:** snow-track surveys; bias in track surveys and population indices; detectability of tracks; sex differences in track presence; daily movement distance; effect of weather (temperature, days since snowfall) on tracks; telling old from new tracks; repeated use of trails; heterogeneity in detection and capture between individuals; occupancy and mark-recapture survey design; observer disturbance.
- **Statistics or models:** Mixed-effects logistic (logit) regression with a random effect for each otter (Stata 9.2, GLLAMM procedure with adaptive quadrature); eight candidate models compared with Akaike’s Information Criterion for small samples (AICc), delta AICc and Akaike weights; variance inflation factors (VIF) for multicollinearity; receiver operating characteristic (ROC) curve and area under the curve (AUC) for predictive accuracy; Pearson’s standardized residuals for outliers. Possible inconsistency: the paper gives a mean VIF above 1 as the sign of high multicollinearity, a threshold almost any model would exceed, yet says no model had high multicollinearity. Note also that the abstract and methods call the otters “radio-collared”, but the methods describe radio-transmitters implanted in the body cavity. Movement distance is reported in m/day, but the units behind its coefficient in Table 2 are not stated.
- **Data type:** New field data: otters captured in late autumn 2007; tracks recorded at 125 otter locations from January to March 2008. Weather data from Environment Canada’s National Climate Data and Information Archive. Data were not collected in late winter, when hard snow and ice meant otters left no detectable tracks.

## Synopsis

### 1. What they did

Surveys of tracks and other signs are widely used to judge where animals live and how many there are, and snow-track surveys suit river otters well because their tracks are easy to identify and they travel across open frozen lakes and rivers. But if some animals leave more tracks than others, or weather changes how many tracks are left, survey results could be biased. The authors set out to test which factors affect whether otter tracks are present in the snow.

- Marking otters. In late autumn 2007, they live-trapped river otters along the shores of Tezzeron and Pinchi Lakes, implanted radio-transmitters, and classed each otter’s age from weight, skull size, signs of nursing and tooth wear.

- Finding the otters. From January to March 2008, they located five otters (one juvenile male, two adult males, one juvenile female and one adult female) by radio telemetry, homing in on each one until they knew where the otter was to within about 5 m.

- Searching for tracks. Staying 30 to 50 m from the otter, they searched on snowshoes in a concentric pattern, moving outwards in steps of 15 to 20 m, out to 100 m around the otter’s location, using binoculars. They recorded whether there were otter tracks made since their previous visit.

- Modelling. They tested which of five factors best explained whether tracks were present: mean daily temperature, days since the last snowfall of more than 1 cm, sex, age, and the distance the otter had moved since its previous location.

### 2. What they found

Track presence

- They recorded the presence or absence of tracks at 125 otter locations.

- Tracks were found at 88.9% of locations for the juvenile male, 64.3% and 58.8% for the two adult males, 39.1% for the adult female and 16.7% for the juvenile female.

- Daily movement varied a lot between otters, from about 391 m a day (juvenile female) to about 1,374 m a day (adult female).

Best model

- The best model contained only sex and daily movement distance. It was far stronger than the others: its Akaike weight was 0.89, 11.5 times that of the next best model, and it predicted well (area under the curve 0.801).

- Tracks were more likely to be found for males than females, and more likely the farther an otter had moved that day.

- The three best models all included sex. Weather (temperature and days since snowfall) did not explain track presence as well as sex and movement.

- Movement distance was not linked to sex, so the difference between males and females was not simply because males moved farther. The adult female actually moved the farthest of all five otters but left tracks at only 39.1% of her locations.

How the otters moved

- Females did not make long-distance movements over open snow. Males often left tracks on top of the snow and ice on both short and long journeys.

- Only one marked otter, the adult female, had access to open water; most otters reached the water through burrows and small holes under the snow near shore.

- The otters reacted only slightly to the researchers and kept using the same areas, suggesting they were not displaced.

### 3. What they concluded

- This was the first study to examine how demographic, behavioural and environmental factors affect the interpretation of otter track surveys.

- Sex and daily movement both affect whether otter tracks are left in the snow. A track survey of a population with many less-detectable females would underestimate numbers or distribution.

- Weather probably still plays a role in finding tracks, even though it did not explain track presence as well as sex and movement here.

- With only five otters, it is hard to separate the effect of sex from the habits of individual otters. Either way, track surveys that do not correct for detection differences linked to behaviour should be used with caution.

- Survey design (how long, how often, and the length, shape and pattern of routes) may need adjusting for differences in detection between groups of animals. Track surveys of river otters that do not account for these factors should be interpreted with caution until further research can measure or correct the bias.

### 4. How tracking was or was not used

Tracking was the core of the study. The researchers searched for and identified river otter tracks in the snow within 100 m of known otter locations. They sketched tracks in a field notebook to tell old from new tracks, estimated track age from known weather events, and checked previously found tracks on following days to see whether trails were reused. Snowfall or melt between visits usually made old and new tracks easy to tell apart. Shallow snow left a clear body imprint that was easy to detect and identify, without forcing otters onto fixed trails. They counted only tracks made since the previous visit, and recorded tracks as unknown when sun angle or blowing snow made visibility very poor. Tracks that could not be linked to the marked otter’s own movements were rarely seen (fewer than 5 times). No trailing was reported.

### 5. Tracker or observer certification

Not reported. The paper does not say who searched for and identified the tracks, or what their experience was.

### 6. Other technologies, skills or innovations

- Implanted radio-transmitters (Advanced Telemetry Systems M1250B, about 100 g) and passive integrated transponder (PIT) tags under the skin.

- Radio telemetry by triangulation and homing, with a laser range finder to estimate the distance to the otter’s location.

- Snowshoes and 10 x 50 binoculars for the track search.

- Mixed-effects logistic models comparing eight candidate explanations.

### 7. Implications for conservation (from the paper)

- Track surveys often inform management decisions such as harvest quotas and conservation status, so hidden biases matter.

- Estimates of otter distribution, relative abundance and habitat selection from track surveys may be misread if behavioural differences between individuals are ignored. Occupancy models should consider specific sources of variation in detection, such as behaviour.

- Mark-recapture estimates may also be affected. Studies that combine track detection with known individuals can help explain why some animals are more detectable than others, and help design surveys that reduce this bias.

### 8. Implications for tracking (Original Wisdom's observations)

_These are Original Wisdom's observations, not the authors'._

- The study is a reminder that the absence of tracks is not the absence of the animal: even with a radio-marked otter known to be within 100 m, tracks were found at under 17% of one female’s locations.

- How an animal moves (over the snow or below it) may matter more than how far it moves. The adult female travelled farthest but left comparatively few tracks on the surface.

## Cite this summary

Lawrence, K. 2026. Summary of: The role of demographic and environmental variables on the presence of snow tracks by river otters Lontra canadensis (Crowley and Johnson et al., 2012). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2012-crowley-johnson-role-demographic-environmental-variables/