University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Barbara J. Keller, Natural Resources Douglas F. Smith, Ph.D. Department of Biology Northland College, Ashland, WI

ABSTRACT

According to Jackson (1961), southern (Glaucomys volans) and northern (G. sabrinus) flying squirrels occupied the southern and northern parts of Wisconsin, respectively, and did not overlap in their distributions except in the “tension zone”. However, recent studies identify tremendous overlap in their distributions. What is not known is whether the two species are sympatric at the scale of the patch, and in what patches they co-occur. Therefore, the objective in this study was to identify the range of habitats occupied by G. volans in northern Wisconsin. Five northern hardwood stands that ranged in age and composition were studied to determine what habitat and microhabitat features were most important to G. volans. Pearson’s correlation coefficients showed significant correlations in several variables including: shrub density and snag dbh, shrub density and number of snags, average dbh and number of trees, number of shrubs and number of oaks (Quercus spp.), number of trees and number of oaks, and number of oaks and oak dbh. Heterogeneity exists within stands in number of oaks per station, diameter of oaks, shrub density per station, course woody debris (logs) volume per station, and diameter of snags per station. These results will be used to predict habitat use by G. volans with data obtained from live-trapping (beginning September 2001).

Introduction

Southern flying squirrels (Glaucomys volans) are common, small nocturnal, non-hibernating sciurids that reside in deciduous forests throughout the eastern and central United States. The range of G. volans has been recorded by several authors (Braun 1988; Hazard 1982; Muul 1974; Weigl 1977) with specific reference to Wisconsin by Jackson (1961). Jackson (1961) describes the “tension zone” in Wisconsin as the point at which the northern distribution of G .volans and the southern distribution of the Northern flying squirrel (Glaucomys sabrinus) meet. This should be the only area within Wisconsin where the ranges of the two species overlap.

33 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

However, recent investigations contradict Jackson’s (1961) report. Scientists have found G. sabrinus occurring south of the tension zone (Anderson, unpubl. data) and G .volans occurring north of the tension zone (Smith, unpubl. data). G. sabrinus is larger than its counterpart, averaging a weight of 89.6 g and 273 mm in length (Hazard, 1982). Their range is throughout Canada and the upper Midwest, in the west in northern California and Colorado, and in small populations where there is suitable habitat in the south. G. sabrinus occupies forests dominated by second or old growth conifers, with a well developed understory. They feed primarily on hypogeous fungi, but also supplement their diet with various nuts, tree lichens, buds, staminate cones, fruits, insects and sap (McIntire & Carey 1989). The squirrels nest in cavities located in snags and trees. G. volans occupy forests dominated by oak (Quercus spp.) and hickory (Carya spp.), or other mast producing hardwoods. The species ranges throughout eastern North America from southern Ontario to the Gulf Coast, and in small populations south in Mexico and South America. G. volans is smaller than the northern species, weighing an average of 61.6 g and 228 mm in length, allowing the two species to be easily distinguished (Hazard 1982). The basic diet of G. volans consists of hickory nuts and acorns, and in the spring may also consist of blossoms of sugar maple and juneberries, as well as various insects. G. volans nest in cavities in trees and snags, and have more potential to build outside nests in the southern limits of the range. G. volans may be able to survive the harsh winters of northern Wisconsin because the genus has several behavioral adaptations to combat cold temperatures (Stapp et al. 1991). First, flying squirrels build food caches throughout the year when food is abundant. They can then access these caches during the winter when food is scarce. Second, during winter squirrels aggregate together in highly insulated nest cavities, with as many as 19 squirrels per nest to retain heat (Muul 1974; Stapp et al. 1991). They remain inactive much of the winter and leave the nest only to forage (Muul 1974). The objective of this study was to provide a clearer definition between the habitat of G. volans and G. sabrinus by investigating the habitat requirements of G. volans. Additionally, an attempt was made to determine if and in what habitats the two species are sympatric. It is important to study the habitat of G. volans in northern Wisconsin because it is likely a newly established resident to the northern hardwood forest. Also, because G. volans and G. sabrinus use snags for nests and for winter, scientists have suggested the two may be ecological competitors. [The time parameter for this research will encompass at least one year. The first installment of this research, focuses on the habitat parameters and the

34 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001 experimental design. Live-trapping for the second stage will begin in late August, after the first phase is complete.]

Methods

The study area was located in the Washburn District of the Chequamegon-Nicolet National Forest in Bayfield County, Wisconsin. Squirrels will be trapped in six stands beginning in August, 2001 to investigate their habitat use north of the tension zone. An attempt was made to stratify stands by age, but it was found that all oak stands originated from the same decade (CNNF, Washburn District unpublished data). Stands were stratified by oak composition as well as by shrub density. Traps were placed 50 m apart along 200 m transects. Each plot (4 ha) contained a matrix of 5 trap stations per transect along 5 transects, for a total of 25 traps per plot. Sherman live traps were placed 1.5 m high in the largest tree within a 5 m radius of each station. Traps will be baited with a mixture of peanut butter, molasses, and oats and checked daily at dusk. Squirrels captured will be identified (to species), weighed, sexed, aged, ear-tagged (numbered Monel 101 tag) and released. Sexing and aging will be determined by the presence and size of the mammaries or the presence and size of testes (Bookhout 1996). At each trapping station habitat characteristics were measured within a 10 m diameter circle around the trap tree, including species composition, canopy closure (with an ocular tube), shrub cover, dbh of capture tree, and dbh and condition of snags present (condition measured by number of limbs present; [Fogel et al.1973]). For each plot, species, abundance, and dbh of trees were measured, as well as dbh and number of snags. Shrub coverage was recorded by density and richness per plot. Habitat use will be characterized at 2 spatial scales: stand and microhabitat. Habitat characteristics will be compared among all 6 stands to determine the range of habitat occupied by G. volans. Microhabitat at capture sites will be compared among stands to determine where G. volans concentrate their activity. Pearson’s correlation coefficients were obtained for stand variables to determine which variables should be included in comparisons, and which variables were correlated enough so that both need not be included in later analyses. When trapping is complete, logistic regression will be used to determine the habitat preferences of the G. volans.

35 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Results

The average diameter of trees, number of stems, number and dbh of oaks, percent of oaks, and basal area was calculated for 5 stands (Figs. 1-4; Table 1). Shrub density and richness, number and dbh of snags, and log volume were also recorded for each of the stands (Table 2;Figs. 5-8). Observations have not yet been collected from the sixth stand due to time constraints and difficulty in locating appropriate stand characteristics. Stand 1 has the largest average diameter of trees of all stands, and a relatively high basal area (Table 1). Stand 1 also has the highest number of oaks per trap station, and a relatively high percentage of oak throughout the stand, although the diameter of the oaks is average in comparison with the other stands (Table 1). The species distribution of stand 1 (Fig. 9) indicates an oak-red maple (Acer rubrum) dominated forest with several other tree species. The average shrub density and richness for stand 1 (Table 2) is the lowest of all the stands, indicative of a closed canopy with few gaps. Coarse woody debris volume appears average in comparison with other stands (Table 3). Stand 2 is a red maple-oak-paper birch (Betula papryifera) dominated forest with two other species in low frequency (Fig. 10). It has relatively high basal area and average diameter size of trees (Table 1). With one of the smallest density of trees per trap station (10.28, Table 1) and a low density of shrubs (26.92/plot; Table 2), it is a relatively open forest with a closed canopy. Stand 2 has a small percentage of oak, but a large average oak dbh of 22.56. The coarse woody debris volume for stand 2 is low for logs and stumps, but average for number of snags per station (Table 3). Stand 3 is a red maple-trembling aspen (Populus tremmuloides) - oak stand with low occurrence of other species (Fig. 11). It has the second highest dbh of all the stands (Table 1), as well as the highest average number of trees per trap station (Table 1). Stand 3 has one of the smallest average number of oaks per station, and a small percentage of oak and small oak diameter. Stand 3 appears to be a particularly even-aged stand as the average dbh of oaks compared to the whole stand is virtually the same. With the largest density of shrubs per trap station (Table 2), it appears that while stand 3 has large trees, which might indicate a more open forest, it also has the most trees per plot (Table 1) and the highest shrub density (Table 2). Coarse woody debris volume in stand 3 is lowest for logs, low for stumps, and the least for snags relative to the other stands (Table 3). This could be the youngest of all the forests due to low coarse woody debris values and high shrub content, or it could be a richer site with higher decomposition rates.

36 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Stand 4 is a red maple-oak-sugar maple (Acer saccarhum)-paper birch (Betula papryifera) forest with small numbers of other species (Fig. 12). It has the smallest average dbh and basal area values of all the stands, as well as a small number of stems, least number of oaks per stand, and smallest percentage of oaks (Table 1). However, stand 4 has the largest oak dbh of all the stands (Table 1), and a relatively average shrub density (Table 2). Stand 4 has the largest volume of coarse woody debris for logs, stumps, and snags (Table 3). This stand shows all the typical indicators (small stem numbers, high tree dbh, low shrub density, high coarse woody debris volume) of an older, open forest. The species composition of stand 5 is red maple-oak with small numbers of other species (Fig. 13). Stand 5 has small dbh values, as well as the least number of trees per station (Table 1; Fig. 6), but a large number of oaks per station and the highest percentage of oak (Table 1). However, the oaks are the smallest (dbh) of all the stands (Table 1). Stand 5 has a large density of shrubs (Table 2), and large values for coarse woody debris volumes in all categories (Table 3). This stand has characteristics typical of a younger forest (low dbh, low basal area, high shrub density) but a large volume of coarse woody debris, which would be contradictory of a young forest. Pearson’s correlation coefficients were obtained with two-tailed significance levels on all the variables recorded within and among the stands. Number of shrubs and the total dbh of the snags were positively correlated (r = 0.217, P= 0.015,). With each increase in snag dbh, the shrub density of the stand also increases. In accordance with these results, the number of snags at a trapping site and the number of shrubs at a site were also positively correlated (r = 0.202, P = 0.024). The average dbh of trees was negatively correlated with the number of trees in a stand (r = -0.593, P < 0.001). A bivariate linear regression model with a best fit line plot (Fig. 14) illustrates this correlation. This could mean as the number of trees in a stand decreases (natural mortality associated with age), the average dbh of the remaining trees increases. The number of shrubs at a stand was negatively correlated with the number of oaks in a stand (r = -0.303, P = 0.001). According to these results, stands with a high density of oaks should have low shrub density, and visa versa. However, the number of oaks in a stand was positively correlated with the number of trees in a stand (r = 0.228, P = 0.011). Also positively correlated with the number of oaks was the dbh of the oaks in a stand (r = 0.231, P = 0.009,).

37 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Discussion

Northern hardwood forest stands differ in species composition, oak distribution and abundance, number of trees, coarse woody debris volume, and shrub density. In addition, any site within a stand differs greatly from any other randomly chosen location within a forest stand. Thus, when G. volans capture sites are compared to habitat features at the site, it can be inferred what habitat features are most important to G. volans, and the likelihood of finding G. volans in other stands can be predicted. Analyses in this study revealed several informative principles concerning forest growth. For example, as trees in a forest gain in diameter, many die naturally (self-thinning) as the larger trees fill the canopy and prevent sunlight from reaching those that are slower growing. Perhaps this will also explain the positive correlation between snag numbers and shrub density. As the older, larger trees die, they create gaps in the canopy that allow shrubs and small trees to flourish. The negative correlation between shrub numbers and oak numbers may also be explained. In Northern Wisconsin, most of the oak population was harvested during the widespread logging at the turn of the century, referred to as the “great cutover”. Due to the long growing lifespan oaks, and a long cutting rotation, they have not yet been harvested, and now typify older hardwood forests. Many qualities of this older forest could explain the loss in shrub density. However, as shrub density and oak density are very important to the habitat of G. volans, both variables will nonetheless be included in capture/habitat comparisons. This will, in fact, be a particularly interesting comparison to consider, as shrub density likely provides important vertical cover from overhead predators; and oak mast is a staple of the G. volans diet. G. volans may compensate for this by nesting/roosting in older parts of a stand that contain more snags of suitable size for cavities, and forage where shrub density is greatest but with oak basal area also high enough to produce mast. It should be noted that older forests usually contain a larger amount of coarse woody debris, and a positive correlation between snags and shrubs apparently exists (refer to results section). Theoretically then, within these stands pockets of dense shrub cover should exist where older trees have died or fallen, opening up the canopy. In fact, high within stand variability exists (refer to standard deviation bars in Figs. 4-8) with the average number of oaks per station, the diameter of the oaks, average shrub density per station, average log volume per station, and average diameter of snags per station. These stands are therefore very heterogeneous in these areas, meeting the needs of G. volans. Sufficient variability does exist among the shrub density and coarse woody debris (particularly snags and logs) values between the stands to

38 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001 warrant capture success comparisons. The shrub density of stands 1 and 2 is small, while that of stands 3, 4, and 5 is significantly larger. The log volume of stand 3 is small, while at stands 1, 2, and 5 it is medium, and very large at stand 4. The snag volume of stands 1 and 2 are small, and medium for stands 3, 4, and 5. Stand 6 will preferably have little or no oak composition, a very high shrub density, and a high volume of coarse woody debris. National forest logging data will be examined to locate a suitable stand. This will allow for sufficient comparisons of trapping success between the stands. Further research will include the actual trapping of the squirrels in the fall of 2001, and statistical analyses of the features at successful capture sites within and among stands. Results will be compared with ongoing research at Northland College of G. sabrinus habitat use.

39 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Works Cited

Anderson, E. and J. S. Borovansky. Habitat selection by northern and southern flying squirrels in central Wisconsin. unpublished manuscript. UWNSP Wildlife Department.

Fogel, R., M. Ogawa, and J. M. Trappe. 1973. Terrestrial decomposition: a synopsis. Rep. 135, 12 p. Coniferous Forest Biome, Seattle, Washington.

Goertz, J. W. and R. M. Dawson. 1975. Response to nest boxes and reproduction by Glaucomys volans in Northern Louisanna. Journal of Mammology 156:933-939.

Hall, D. 1991. Diet of the northern flying squirrel at Sagehen Creek, California. Journal of Mammology 72:615-617.

Hayward, G. and R. Rosentreter. 1994.Lichens as nesting material for northern flying squirrels in the Northern Rocky Mountains. Journal of Mammology 75:663-673.

Hazard, E. B. 1982. The Mammals of Minnesota. University of Minnesota Press, Minneapolis, Minnesota. p71-74.

Hooper, E. T. 1951. Records of the southern flying squirrel in Mexico. Journal of Mammology 33:109-110.

Jackson, H. H. 1961. Mammals of Wisconsin. The University of Wisconsin Press, Madison, Wisconsin. Pp175-184.

Jordan, J. S. 1948. A midsummer study of the southern flying squirrel. Journal of Mammology 29:44-48.

Maser, C.; Maser, Z.; Witt, J. and Hunt, G. 1986. The northern flying squirrel: a mycophagist in southwestern Oregon. Canadian Journal of Zoology 64:2086-2089.

McIntire, P and A. Carey.1989. A microhistological technique of food habits of mycophagous rodents. U.S.D.A. Forest Service. Pacific Northwest Research Station. Research Paper PNW-RP-404.

Muul, I. 1969. Photoperiod and reproduction in flying squirrels- Glaucomys volans.

40 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Muul, I. 1974. Geographic variation in nesting habits of Glaucomys volans. Journal of Mammology 55:840-844.

Muul, I. Behavioral and physiological influences on the distribution of the flying squirrel, Glaucomys volans. Flying Squirrel pgs 1-66.

Payne, L.; D. R. Young and J. Pagels. Plant community characteristics associated with the endangered northern flying squirrel, Glaucomys sabrinus, in the southern Appalachains. American Midland Naturalist 121:285-292.

Ransome, D. B. and T. P. Sullivan. 1997. Food limitation and habitat preference of Glaucomys sabrinus and Tamiascuirus hudsonicus. Journal of Mammology 78:538-549.

Rosenburg, D. K. and R. G. Anthony. 1992. Characteristics of northern flying squirrel populations in second and old growth forests in Western Oregon. Canadian Journal of Zoology 70:161-166.

Sollenberger, D. E. 1943. Notes on the breeding habits of the eastern flying squirrel (Glaucomys volans volans). Journal of Mammology 163-173.

Sollenberger, D. E. Notes on the life history of the small eastern flying squirrel. Journal of Mammology 282-293.

Sonenshine, D. E., Gerretani, D.G. and G. Enlow. 1973. Improved methods for capturing wild flying squirrels. Journal of Wildlife Management 37:588-590.

Weigl, P. D. and D. W. Osgood. 1973. Study of northern flying squirrels, Glaucomys sabrinus, by Temperature Telemetry. American Midland Naturalist 92:482-486.

41 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

List of Tables:

Table 1. Average diameter of trees, number of stems, number of oaks, and percent of oaks per trap station in Northern Wisconsin plots.

Table 2. Average shrub density and richness per trap station in Northern Wisconsin plots.

Table 3. Coarse woody debris volume totals and averages per trap station in Northern Wisconsin plots.

42 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Table 1: Average diameter of trees, number of stems, number of oaks, and percent of oaks per trap station in Northern Wisconsin plots:

Basal Oak Stand Dbh1 area2 # stems # oaks % oaks dbh1 1 19.50 0.08 11.72 3.56 27.58 21.02 2 17.36 0.07 10.28 2.00 21.07 22.56 3 19.42 0.10 13.16 2.56 23.55 19.67 4 13.92 0.04 11.20 1.84 16.25 24.22 5 15.30 0.05 10.00 2.80 29.12 19.03 1cm 2cm2

Table 2: Average shrub density and richness per trap station in Northern Wisconsin plots:

Stand Average density Richness 1 11.28 3 2 26.92 4 3 55.72 4 4 38.32 5 5 49.56 4

Table 3: Coarse woody debris volume totals and averages per trap station in Northern Wisconsin plots:

Stand Log Volume Stump Volume Snag Volume TL Avg TL Avg TL Avg 1 3191.10 58.02 188.12 20.90 157.6 7.50 2 4110.89 42.38 162.77 10.17 572.8 7.54 3 1657.85 36.04 273.66 17.10 277.4 6.60 4 5987.08 61.72 3864.81 241.55 636.6 11.99 5 3710.07 68.71 1040.61 570.81 504.5 10.73

43 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

List of Figures:

Fig. 1. Mean number of trees in each trap site in Northern Wisconsin plots (+/- std.). Fig. 2. Mean diameter (cm) of trees in each trap site in northern Wisconsin plots(+/- std.).

Fig. 3. Mean number of oak trees in each trap site in Northern Wisconsin plots (+/- std.).

Fig. 4. Mean dbh of oak trees in each trap site in Northern Wisconsin plots (+/- std.). Fig. 5. Mean shrub density in each trap site in Northern Wisconsin plots (+/- std.).

Fig. 6. Mean number of snags in each trap site in Northern Wisconsin plots (+/- std.).

Fig. 7. Mean total snag dbh in each trap site in Northern Wisconsin plots (+/- std.).

Fig. 8. Mean log volume in each trap site in Northern Wisconsin plots (+/- std.).

Fig. 9. Percentage distribution of tree species in stand 1.

Fig. 10. Percentage distribution of tree species in stand 2.

Fig. 11. Percentage distribution of tree species in stand 3.

Fig. 12. Percentage distribution of tree species in stand 4.

Fig. 13. Percentage distribution of tree species in stand 5.

Fig. 14. Bivariate linear regression of stem dbh versus number of stems.

44 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Fig. 1

25 20 15 10 5 number of trees of number 0 1 2 3 4 5 stand id

Fig. 2

30.0 25.0 20.0 15.0 10.0

tree dbh (cm) treedbh 5.0 0.0 1 2 3 4 5 stand id

45 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Fig. 3

7 6 5 4 3

2 # oaks# 1 0 -1 1 2 3 4 5 -2 stand id

Fig. 4

35.0 30.0 25.0 20.0 15.0

10.0 oak dbh (cm) dbh oak 5.0 0.0 1 2 3 4 5 stand id

46 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Fig. 5

100

80

60

40

shrubdensity 20

0 1 2 3 4 5 stand id

Fig. 6

6 5 4 3 2 1

0 number of snags of number -1 1 2 3 4 5 stand id

47 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Fig. 7

50

40

30

20

10

snag volume (cm) volume snag 0 1 2 3 4 5 -10 stand id

Fig. 8

500.00

400.00

300.00

200.00

100.00

log volume (cm) volume log 0.00

-100.00 1 2 3 4 5

stand id

48 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Fig. 9

Acer rubrum

Acer saccharum

Betual 30% allegheniensis Betula papyrifera 52% 0% Pinus resinosa 8% Quercus rubra 1% 9%

Fig. 10

Acer rubrum

Acer 19% saccarhum

0% Betula papyrifera 14% 60% Pinus resinosa

7% Quercus rubra

49 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

Fig. 11

Abies balsamea 1% 19% Acer rubrum

3% Acer saccharum

57% Populus 18% tremmuloides Betula papyrifera 2% Quercus rubra

Fig. 12

Acer rubrum

17% Acer 1% 29% saccharum 3% Betula papyrifera Ostraya 24% virginiana Populus 26% tremmuloides Quercus rubra

50 University of Wisconsin-Superior McNair Scholars Journal, volume 2, 2001

Fig. 13

Acer rubrum

Acer saccharum

Betula papryifera 28% 32% Ostraya virginiana Pinus banksiana 9% Pinus resinosa 0% 14% 0% 16% Populus 1% tremmuloides Quercus rubra

Fig. 14

50.0 "stem_dbh" 40.0 Predicted "stem_dbh" 30.0

20.0 stem dbh stem 10.0

0.0 0 10 20 30 number of stems

51 Habitat Use Of Southern Flying Squirrels (Glaucomys volans) In Northern Wisconsin

52