# Selection of Sites for Winter Night Beds by White-tailed Deer in a Hemlock-Northern Hardwood Forest

> Where does a white-tailed deer choose to sleep on a freezing winter night? By wiping out old beds every evening and searching the snow at sunrise, researchers in Maryland found 140 fresh night beds and compared them with random spots. Deer favoured dense hemlock overhead and on the windward side, a tree close by, and an open view to the southeast.

Summary by Kersey Lawrence, Original Wisdom. Updated 9 November 2026. Page: https://www.originalwisdom.com/tracking-literature/1985-lang-gates-selection-sites-winter-night/

## Citation and access

- **Authors:** Brian K. Lang and J. Edward Gates
- **Year published:** 1985
- **Journal or source:** The American Midland Naturalist
- **Full citation:** Lang, B. K. and Gates, J. E. 1985. Selection of sites for winter night beds by white-tailed deer in a hemlock-northern hardwood forest. The American Midland Naturalist 113(2): 245–254.
- **DOI:** 10.2307/2425570
- **Access status:** Publisher link only
- **Where to find it:** https://doi.org/10.2307/2425570

## Tags

- **Species:** Mammals > White-tailed deer
- **Region:** North America > United States
- **Tracking:** Beds; Snow tracks; Tracking; Tracks; Tracks and signs identification
- **Behaviour and ecology:** Behaviour; Habitat selection
- **Methods and technology:** Analysis of variance (ANOVA); Chi-square test; Mann–Whitney U test
- **Data type:** New field data
- **Observer qualification:** Not reported
- **Discipline:** Behaviour and ecology; Wildlife tracking
- **Year published:** 1985

## Detailed keywords

- **Species:** Studied: white-tailed deer (Odocoileus virginianus). Trees recorded around bed and random sites: eastern hemlock (Tsuga canadensis), red maple (Acer rubrum), sugar maple (A. saccharum), American beech (Fagus grandifolia; the paper gives the scientific name only), black cherry (Prunus serotina), yellow birch (Betula lutea), B. lenta, white oak (Quercus alba; scientific name only in the table), red oak (Q. rubra), Q. prinus, Ostrya virginiana and Carpinus caroliniana (scientific names only), plus snags.
- **Location and habitat:** Deer wintering area along Blandy Run Creek, which flows into the Frostburg reservoir, on the Allegheny Plateau in northeastern Garrett County, western Maryland, United States (part of the City of Frostburg watershed). Rolling tableland with 10-35% slopes and nearly level alluvial bottomland, 716 to 823 m above sea level. Three habitat types: a northeast-facing slope of eastern hemlock and sugar maple, a southwest-facing slope of sugar maple and red oak, and bottomland of eastern hemlock, red maple and yellow birch. Beds and random sites were in the eastern hemlock, red maple and yellow birch cover type. Humid continental climate; prevailing winter winds from the west and northwest; about 180 cm of snow a year.
- **Topics:** winter night beds; bed-site selection; deer wintering areas (deer yards); microhabitat selection; thermal cover and energy conservation; coniferous (hemlock) cover; wind chill and shelter-seeking; open southeastern exposure (morning sun, view or escape route); distance to nearest tree; spatial heterogeneity of trees; canopy cover above the bed; overnight temperature and snow depth.
- **Statistics or models:** One-way analysis of variance (ANOVA); Bartlett’s test for homogeneity of variance; Mann-Whitney U test (densiometer values); chi-square tests corrected for continuity (tree species in each quadrant); t-test (overnight temperatures); principal components analysis (PCA) on 11 variables from the random sites, with bed sites plotted in that space and tested with a one-sample chi-square test; logarithmic and square-root transformations. Cover density scored after Nudds (1977); spatial heterogeneity index after Roth (1976). Inconsistencies: the Study Area describes northeast-facing and southwest-facing slopes, but the Results give snow depth for a “SE-facing slope”. The text reports 25 surveys (20 in 1981, 5 in January 1982), but the temperature comparison uses 23 visits (14 with beds, 9 without).
- **Data type:** New field data: bed-site surveys January to mid-March 1981 and January 1982; microhabitat measurements in late March and April, before leaf-out.

## Synopsis

### 1. What they did

White-tailed deer gather in sheltered wintering areas, often in conifer stands on level bottomland, to save energy and escape harsh weather. Where they choose to bed down within such an area had not been studied for night beds. The authors set out to find which features of the immediate surroundings (the microhabitat) deer favour for night beds in an eastern hemlock and northern hardwood forest in western Maryland.

- Finding beds. They laid out nine transects, each 35 m by 700 m (22.05 ha in total), across the wintering area. In the late afternoon they brushed snow into all old beds. At sunrise the next morning they searched the transects for fresh night beds in the snow and marked each one with an aluminium tag. Surveys ran one to three times a week from January to mid-March 1981 and in January 1982, while snow lay on the ground.

- Weather. At three weather stations they recorded overnight lowest and highest temperatures and five random snow depths before each survey.

- Measuring the sites. In late March and April, before the leaves came out, they measured 62 features at 140 night-bed sites and 100 random sites in the same cover type. These included tree basal area, canopy cover directly overhead, how dense the cover was in four directions at different heights and distances, and the distance, direction, size and species of the nearest tree in each quarter around the site.

- Comparing. They compared bed sites with random sites one feature at a time, then used principal components analysis (PCA), a way of combining many measurements into a few main patterns, to see where bed sites fell among the sites available.

### 2. What they found

When deer bedded there

- The surveys found 0 to 28 night beds per visit (average 5.6).

- Nights when beds were found were much colder (average low of -17.3 °C) than nights with no beds (average low of -8.9 °C).

- No beds were found on five visits in March, when the weather was milder.

- Snow was shallowest on the bottomland (average 11.2 cm) and deepest on the northeast-facing slope (average 21.7 cm, up to 42.0 cm).

What the bed sites were like

Compared with random sites, night-bed sites had:

- More cover directly overhead: canopy cover averaged about 94% above beds against 84% above random sites.

- More open views to the southeast: less low cover (from the ground to 1 to 1.5 m) to the southeast at 5 and 10 m. At 5 m there was more cover to the southeast between 2 and 2.5 m high.

- A tree close by: the nearest tree was closer (average 1.3 m against 1.7 m).

- Patchier tree spacing: a higher spatial heterogeneity index (52.1% against 42.5%).

- More hemlock to the north and west: eastern hemlock was the nearest tree to the north at 56.4% of beds (34.0% of random sites) and to the west at 52.9% (37.0%).

- Less red maple to the north and less black cherry to the south than expected.

There were no differences in tree basal area or in cover to the northwest. Overall, 8 of the 62 features differed between bed and random sites, besides the overhead canopy cover.

The overall pattern

Four main patterns from the PCA explained 72.5% of the variation among random sites. Deer beds clustered where low cover to the southeast was sparse and trees were close and patchily spaced: 68 of the 140 beds fell in that quarter of the available habitat, far more than expected. Even so, deer used a wide range of sites.

From tracks and trails in the snow

Following snow trails and tracks, the authors found that deer moved into dense hemlock cover in severe weather (very low temperatures with gusty, strong winds) and left once the weather eased, apparently to browse elsewhere.

### 3. What they concluded

- Although many sites seemed suitable, deer chose night beds with: (1) an open southeastern exposure; (2) hemlock trees to the west and north; (3) patchy tree spacing; (4) a tree close by; and (5) dense cover directly above.

- An open southeast side may let in morning sun, give an open view or offer an escape route. An open view to the southeast may matter most when the wind, carrying scent, blows from the west and northwest.

- Hemlock to the west and north would shelter deer from the prevailing west-northwest winds. Dense cover overhead would cut heat lost to a cold, clear night sky.

- The chosen sites apparently gave better conditions for survival than other sites available.

- The night beds were similar to those in a study in Ontario, Canada, though there most beds were high on slopes, especially north-facing ones, while here most were on nearly level bottomland. In both places, dense hemlock cover was a key factor.

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

The whole study rests on finding and identifying fresh deer night beds in the snow. Old beds were erased the evening before, so beds found at sunrise could be dated to the previous night. The authors also report that their observations of snow trails and tracks showed deer entering dense hemlock cover in severe weather and leaving when it eased. No trailing method is described.

### 5. Tracker or observer certification

Not reported. The paper does not say who found and identified the beds or what experience they had. Five people are thanked for help in the field.

### 6. Other technologies, skills or innovations

- A simple way to date beds: brushing snow into old beds the afternoon before each survey, so that every bed found the next morning was fresh from that night.

- Spherical densiometer (Model C) for canopy cover overhead; cover density scores after Nudds (1977) in four directions, at five heights and two distances; angle gauge for basal area; compass, steel tape or rangefinder for nearest trees.

- Maximum-minimum thermometers at three weather stations.

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

The paper does not discuss conservation or management directly. It links dense coniferous cover, especially hemlock, with shelter for deer in severe winter weather.

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

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

- Clearing old beds the evening before is a simple, low-cost way to know exactly when a bed was made, and it could be used in other bed or sign studies in snow.

- For trackers looking for winter deer beds, this study points to where to look in this kind of forest: under dense hemlock overhead, close to a tree, with hemlock on the windward side and a more open view to the southeast, especially on the coldest, windiest nights.

## Cite this summary

Lawrence, K. 2026. Summary of: Selection of Sites for Winter Night Beds by White-tailed Deer in a Hemlock-Northern Hardwood Forest (Lang and Gates, 1985). Original Wisdom, Tracking Literature. https://www.originalwisdom.com/tracking-literature/1985-lang-gates-selection-sites-winter-night/