# The function of strategic tree selectivity in the chemical signalling of brown bears

> Brown bears do not rub, bite and claw just any tree. On the coast of British Columbia, researchers found that bears chose large, less common trees along busy trails near salmon and sedge, especially resin-rich firs and spruces. They describe how to recognize a marking tree, how to tell fresh marks from old, and the stomp trails bears leave as they approach.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/2013-clapham-nevin-function-strategic-tree-selectivity/

## Citation and access

- **Authors:** Melanie Clapham, Owen T. Nevin, Andrew D. Ramsey and Frank Rosell
- **Year published:** 2013
- **Journal or source:** Animal Behaviour
- **Full citation:** Clapham, M., Nevin, O. T., Ramsey, A. D. and Rosell, F. 2013. The function of strategic tree selectivity in the chemical signalling of brown bears. Animal Behaviour 85(6): 1351–1357.
- **DOI:** 10.1016/j.anbehav.2013.03.026
- **Access status:** Hosted (freely available elsewhere)
- **PDF:** https://www.originalwisdom.com/wp-content/uploads/2026/11/2013_Clapham_and_Nevin_etal_the-function-of-strategic-tree-selectivity-in-the-chemical-signalling-of-brown-bears.pdf

## Tags

- **Species:** Mammals > Brown bear
- **Region:** North America > Canada
- **Tracking:** Claw marks; Hair; Paths; Sign surveys; Spoor age estimation; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Communication
- **Methods and technology:** Analysis of variance (ANOVA); Camera traps; Chi-square test; Logistic regression; Mann–Whitney U test
- **Data type:** New field data
- **Observer qualification:** Not reported
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 2013

## Detailed keywords

- **Species:** Studied: brown bear (Ursus arctos). Marking trees and available trees: amabilis fir (Abies amabilis), Sitka spruce (Picea sitchensis), Western hemlock (Tsuga heterophylla), Western red cedar (Thuja plicata), red alder (Alnus rubra) and Pacific crabapple (Malus fusca). Food resources in the area: salmon (Oncorhynchus spp., five species) and sedge (Carex spp.). Camera traps were used to rule out American black bear (Ursus americanus) marks. Compared with tree marking in other species from the literature: white-tailed deer, tiger, Alaskan moose, American bison, Iberian wolf, giant panda, European wildcat and black-backed jackal.
- **Location and habitat:** Glendale Cove, an estuarine intertidal zone of Knight Inlet, British Columbia, Canada (Pacific mid-coast; about 15 km² including water). Hypermaritime temperate coastal forest at forest edges bordering the estuary, sedge meadows and a salmon spawning river and channel; game trails and a disused logging road.
- **Topics:** scent marking; chemical signalling and communication; marking trees (rub trees); bite marks, claw marks, rubbing and hair on trees; aging marks as fresh or old; stomp trails (pad depressions leading to marking trees); tree species, size (diameter at breast height) and lean; game trails; conspicuous objects; energetic cost of signalling; breeding season; behavioural plasticity.
- **Statistics or models:** Chi-square goodness-of-fit tests with subdivided post hoc tests; multiple logistic regression (backwards stepwise, conditional); two-tailed independent samples t test; Mann–Whitney U test; one-way analysis of variance (ANOVA). Possible inconsistencies in the paper: (1) for tree size in the logistic regression the paper gives B = 0.21 with SE = 0.009 and Wald χ² = 5.729; these numbers do not agree with each other (B = 0.021 would fit), so 0.21 may be a typo; (2) trail transects total 2.16 km in the Methods but 2.2 km in the Results (rounding); (3) the Results say all marking trees were on game trails, yet two of the 50 were found on the fixed transects; (4) the sample sizes differ slightly between tests (t test 302 degrees of freedom, Mann–Whitney 47 and 247 trees), without explanation.
- **Data type:** New field data: May to October, 2009 to 2011. Camera traps ran 1 June to 31 July in 2009 and 2010 and 15 April to 31 July in 2011 (breeding season), and 1 August to 5 October in 2010 and 2011 (nonbreeding season).

## Synopsis

### 1. What they did

Many large mammals leave scent marks on conspicuous objects such as trees. Making strong-smelling scent costs time and energy, so the authors asked whether brown bears choose their marking trees strategically, to make each mark work harder. They expected bears to pick trees in places they visit often, where a resource (such as food or mates) is worth defending, and to pick trees that stand out: rarer species, conifers over broadleaved trees, and larger trees.

The study ran from May to October, 2009 to 2011, at Glendale Cove in Knight Inlet, British Columbia, Canada. About 40 to 50 brown bears feed on spawning salmon there in autumn, and about 20 feed on sedge in the tidal marshes in spring, during the breeding season.

- Finding marking trees. The team searched 13 fixed transects (2.6 km in total, 200 m each, running straight up the hillside from the shore), 16 transects along wildlife trails (2.16 km in total), and also searched the forest unsystematically. A tree counted as a traditional marking tree only if it showed rubbing (such as hair and bark worn smooth) and visible claw and bite wounds at about the height of a standing bear.

- Measuring the trees. For each marking tree they recorded species, trunk diameter at breast height, lean, the height of the tallest mark, a description of the marks, and signs of overmarking (marks of different ages).

- Measuring what was available. They recorded every tree within 5 m of each marking tree, every tree over 2 m tall along the fixed transects (the wider landscape) and along the trail transects.

- Confirming with cameras. Seventeen camera traps watched 22 marking trees, confirming that the marks were made by brown bears and not black bears.

### 2. What they found

Where the trees were

- All marking trees were at forest edges bordering the estuary and river, clustered near rich food: sedge in spring and the salmon river in autumn.

- Marking trees were found on game trails; only 2 of the 50 were found on the fixed transects. Heavily marked trees were usually on more heavily used trails.

- Trees higher on the hillside showed some marking, but they were widely spaced and not marked enough to count as traditional marking trees.

Which trees bears chose

- Of the 50 marking trees, 23 were amabilis fir, 11 Sitka spruce, 8 Western hemlock, 7 Western red cedar and 1 red alder.

- Bears did not mark trees in proportion to what was available. Across the landscape, amabilis fir, Sitka spruce and Western red cedar were marked more than expected, and Western hemlock, the most common tree, much less. Close to the marking trees, amabilis fir and Western red cedar were again chosen more than expected and Western hemlock less.

- Marking trees were larger than unmarked trees nearby (average diameter 42.5 cm versus 31.7 cm).

- The lean of the tree made no difference.

- Tree species and size together predicted whether a tree was marked; the model classified 87.4% of trees correctly.

- The tallest marks were very similar in height, about 2 m (average 2.02 m on 30 trees).

What the marks looked like

- Marks from a bear rubbing with a wet coat were usually visible for less than 2 hours, while bite and claw marks stayed fresh-looking for the rest of the year.

- Fresh marks were recognized by resin oozing from wounds, torn bark with exposed fragments, the colour of exposed wood, and loose hair.

- Amabilis fir and Sitka spruce often lost large sections of bark to biting and clawing, leaving scars. Western red cedar and Western hemlock usually showed rubbing instead: worn bark with a changed texture. Rubbing darkened the smooth bark of the one red alder.

- Several marking trees had a line of depressions in the ground leading to them, made by bears grinding and stomping their feet as they walked. Some lasted the whole study, some only one season. When the authors watched stomping directly, it often left no lasting mark, depending on the ground, so trees with no visible trail may still have had stomp trails.

- Dead trees that had been marking trees when alive were sometimes still marked.

### 3. What they concluded

- Brown bears are selective about the trees they mark. Location comes first: trees along well-used trails near food and mates, where other bears are most likely to find the scent. Then they favour rarer species and larger trees that stand out.

- Amabilis fir and Sitka spruce, the most marked species, give off pungent resin when torn. The authors suggest this may help the scent last longer or draw other bears to the tree. Because strong long-lasting scents are costly to make, the authors speculate that bears may save energy by using aromatic trees.

- Lean does not seem to matter: bears seem to mark the side of the tree facing the trail, whatever the lean.

- Mark height is unlikely to be a main signal of size for brown bears, since mark heights varied little, cubs have been seen climbing and marking higher, and not all marking involves standing back rubs.

- Across mammals, whether a tree is chosen depends on context. The main function is to limit the energy cost of scent marking by placing marks where they are most likely to be found.

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

The study rests on reading signs on trees and on the ground:

- Identifying marking trees: bite and claw wounds, rubbed bark with a changed texture, and hair left on the trunk.

- Aging marks: resin oozing from wounds, torn bark, the colour of the exposed wood and loose hair showed fresh marks; marks of different ages on one tree showed overmarking.

- Trails: the team followed wildlife trails until they faded or met a road, river, shore or meadow, and found most marking trees along them. They also recorded the stomp trails of foot depressions leading up to marking trees.

Camera traps and daily checks of marking trees in the first week confirmed that the signs were made by brown bears.

### 5. Tracker or observer certification

Not reported. The paper does not say who identified the marking trees or how their skill was checked, although the definition of a marking tree depends on reading bite, claw and rub signs. Camera traps were used to confirm that the marks were made by brown bears.

### 6. Other technologies, skills or innovations

- Infrared camera traps (Reconyx RC55 and PC85) at 22 marking trees.

- A hipchain counter (Walktax, Haglöf) to measure transect length, a diameter tape and a clinometer to measure tree lean.

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

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

- The paper gives a clear field description of a brown bear marking tree and how to tell fresh marks from old ones, and shows how quickly some signs fade: a wet-coat rub may vanish within 2 hours, while bites and claw marks last the year.

- Stomp trails leading to marking trees, and the fact that marking trees cluster on heavily used trails near food, are useful clues for finding marking trees and bear travel routes.

- Tree species changes how a mark looks (bark scars on fir and spruce, worn bark on cedar and hemlock), which matters when reading bear signs in other forests.

## Cite this summary

Lawrence, K. 2026. Summary of: The function of strategic tree selectivity in the chemical signalling of brown bears (Clapham and Nevin et al., 2013). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/2013-clapham-nevin-function-strategic-tree-selectivity/