Printer Friendly

Attributes of rock crevices selected by Allegheny and eastern woodrats in the zone of contact in the Appalachian Mountains of North Carolina.


Two species of woodrats inhabit the eastern United States. The Allegheny woodrat (Neotoma magister) is generally associated with rocky habitats and ranges throughout the Appalachian Mountains to Pennsylvania, the Cumberland Plateau and the Ohio River Valley (Castleberry et al., 2006). The eastern woodrat (N. floridana), which is comprised of eight subspecies, occupies a variety of habitats and occurs throughout much of the south-central and southeastern United States (Monty and Emerson, 2003). Neotoma magister share distributional boundaries with two subspecies of N. floridana: N. f. haematoreia in North Carolina and Tennessee, and N. f. illinoensis in Alabama, Tennessee, Kentucky and Illinois (Ray et al., 2002). Both N. magisterand N. f. haematoreia are listed as federal species of concern (LeGrand et al., 2006), because of declining populations throughout their ranges (Monty et al., 2003; Monty and Emerson, 2003; LoGiudice, 2006). Although reasons for their decline are unknown, LoGiudice (2006) suggests a suite of factors including habitat fragmentation, changes in forest composition, fatal exposure as a secondary host to a parasite and proliferation of human-adapted predators.

In the Appalachian Mountains of western North Carolina and eastern Tennessee, where both Neotoma magister and N. floridana haematoreia are found (Ray et al., 2002), there is a lack of information on the habitats these species use. Recent surveys in North Carolina indicate that both N. magister and N. f haematoreia are sporadically distributed and found in rocky habitats above 300 m, although N. f haematoreia occasionally occur in old barns and abandoned buildings (Ray, 2000).

All members of the genus Neotoma use dens for protection from predators and adverse weather conditions, and as places to rest, store food, nest and rear young (Cameron and Rainey, 1972; Nowak and Paradiso, 1983). Neotoma dens are elaborate (Nowak and Paradiso, 1983), contain multiple entrances and exits, are large enough to accommodate nests and food cache sites and have peripheral areas for latrine sites (Rainey, 1956; Cameron and Rainey, 1972; Monty and Emerson, 2003). In areas where Neotoma are associated with rocky habitats, the availability of suitable rock crevices for den sites is considered a primary factor limiting the size and distribution of populations (e.g., Cameron and Rainey, 1972; Monty et al., 2003; Castleberry et al., 2006; LoGiudice, 2006).

Few quantitative data are available on the microhabitat attributes of rock crevices selected by woodrats. Several studies have reported general characteristics of rock crevices used by Neotoma magister and N. floridana (e.g., Newcombe, 1930; Poole, 1940; Rainey, 1956; Cudmore, 1985). However, only limited inferences on the relative importance of selected attributes can be drawn from these types of studies, because no comparisons have been made to an attribute's availability in the surrounding environment. Two studies of N. magister have examined rock characteristics in terms of their use versus availability (Balcom and Yahner, 1996; Chamblin et al., 2004). However, both are limited in scope with respect to microhabitat associations. Balcom and Yahner (1996) compared only general surface characteristics (e.g., percent rock cover, slope) of sites in Pennsylvania, and Chamblin et al. (2004) investigated characteristics of constructed rock drainage channels on two reclaimed surface mines in West Virginia. To our knowledge, no published studies have quantified the crevice attributes selected by N. floridana.

In this study, we investigate and compare the microhabitat attributes of rock crevices used by Neotoma magister and N. floridana haematoreia in their zone of contact in the Appalachian Mountains of North Carolina, and determine whether the attributes selected differed from those available in the surrounding environments.


We conducted our study of Neotoma magister and N. floridana haematoreia in the Blue Ridge Physiographic Province of western North Carolina. We selected 14 active woodrat sites based on their accessibility and successful trapping efforts by Ray (2000). Study sites were located throughout 11 counties, with a minimum distance of 15 km between sites (Table 1). At each site, between one and six woodrats were captured 1 y prior to our study and identified to species using mitochondrial DNA Ddoop analysis (Ray, 2000). Based on these results, study sites were designated as either N. magister (n = 9) or N. f. haematoreia (n = 5). The MtDNA D-loop tests indicated that all animals were distinct species, with no evidence of hybridization (Ray, 2000).

All 14 study sites were rocky habitats, characterized by abundant outcrops and/or loose talus slopes and composed of gneisses and schists (J.W. Miller, pers. comm.). Sites were located at higher elevations (Neotoma magister: range 610-1402 m, N. floridana haematoreia: range 671-975 m; Table 1) in the Greenschist and Amphibolite Metamorphic Facies Zones of the Blue Ridge Belt (North Carolina Geologic Survey, 1985). Vegetation at the sites generally consisted of scattered trees and shrubs located on dry to mesic slopes often surrounded by intact stands of second-growth hardwoods. Dominant trees varied depending on slope and elevation, but usually included a variety of oaks (Quercus spp.) and hickories (Carya spp.), as well as a mixture of other species including tulip poplar (Liriodendron tulipifera), red maple (Acer rubrum) and white pine (Pinus strobus). Dominant shrubs generally included rhododendron (Rhododendron maximum and R. catawbiense), mountain laurel (Kalmia latifolia) and blueberry ( Vaccinium spp.).

We measured microhabitat characteristics of rock crevices from Dec. 2000 to Apr. 2001. At each of the 14 study sites, we located 10 crevices occupied by woodrats. Woodrat presence was determined by characteristic signs such as latrine sites, and piles of sticks, leaves and other debris stuffed in crevices (Rainey, 1956; Barbour and Davis, 1974; Hoffmeister, 1989). At each of the 10 crevices, we measured the maximum width, height and depth of the opening, determined the aspect, counted the number of internal fissures with openings > 5 cm in diameter that could be used as passageways (Rainey, 1956), and noted whether the floor was wet or dry.

For each crevice used by woodrats, a corresponding crevice without woodrats was located to determine whether the attributes used differed from those available in the surrounding environment. We used a random numbers table to generate a compass bearing and a distance of < 15 m from used crevices, then selected the closest crevice to the random point that appeared suitable for woodrats but contained no evidence of woodrat activity. Data were collected at random crevices as described above.

We examined the normality of the used and corresponding random measurements of each attribute, and their paired differences by species using the Shapiro-Wilk test (Shapiro and Wilk, 1965). Data were not normally distributed, so nonparametric statistics were used for all analyses. We transformed crevice aspects using the equation: 1180[degrees]--crevice aspect I = [theta], where [theta] equals the degree a crevice deviates from south. For each species, we analyzed the paired differences of measurements taken at used and random crevices to determine whether an attribute's use differed from its availability in the surrounding environment. We used Wilcoxon Sign Rank tests to determine whether the mean difference of an attribute differed significantly from zero. To examine whether an attribute's use differed between species, we used Wilcoxon Rank Sum tests to determine whether the means of each attribute differed significandy. For each comparison, a Bonferroni type adjustment of the alpha level was used (Tabachnik and Fidell, 1989). The experiment wise error rate was set at 0.05, and the comparison wise error rate (alpha level) was set at 0.0025. An attribute was considered selected if its use differed from its random availability.

We also performed a discriminant analysis to determine which attributes best distinguished the species. This technique identifies linear combinations of variables (canonical variates) that differentiate among groups (Williams, 1983). Statistical Analysis System (SAS) programs were used for all analyses (SAS Institute Inc., 1990).


We examined a total of 140 rock crevices used by woodrats (Neotoma magister n = 90, N. floridana haematoreia: n = 50). Both species used crevices with larger dimensions (height, width and depth) and more internal fissures than random crevices (all P < 0.0001; Table 2). The aspect of used crevices differed from that of random crevices for N. magister (P < 0.0001), but not for N. f. haematoreia (P = 0.0065). Crevices used by N. magister had more of a southeastern exposure than random crevices (Table 2). All crevices used by N. magister (n = 90) and N. f haematoreia (n = 50) had dry floors, compared to 51% (n = 46) and 60% (n = 30) of random crevices.

The dimensions (height, width, depth) of crevices used by Neotoma magister did not differ from those used by N. floridana haematoreia (all P > 0.07). Dimensions of random crevices also did not differ between species (all P > 0.18), with the exception that the height of random crevices was greater at N.f. haematoreia sites (P = 0.002; Table 2). Crevices used by N. magister had fewer internal fissures than crevices used by N. f. haematoreia (P = 0.002). However, there was no difference in the number of internal fissures between species for random crevices (P = 0.93). The aspect of crevices differed between species (P < 0.0001), with N. magister occupying crevices that were more south-facing, and N. f haematoreia occupying crevices that were more east-facing (Table 2). The aspect of random crevices also differed between species (P < 0.0001), in the same way as used crevices (Table 2). Elevations of study sites overlapped between species (Table 1); however, the mean elevation of N. magistersites was 1192 m, while the mean elevation of N. f. haematoreia sites was 829 m.

Two canonical variables resulted from the discriminant analysis. The first variable had a canonical correlation of 0.7970 and accounted for 63.5% of the variability. The second canonical variable had a correlation of 0.5607 and accounted for 31.4% of the variability. The first canonical variable (Canl) had high positive loadings for number of internal fissures and crevice depth and a high negative loading for crevice aspect (Table 3). The second canonical variable (Can2) had high positive loadings for crevice aspect and number of internal fissures. In a plot of the discriminant analysis, Can1 separated the used crevices from the random crevices along the x-axis, and Can2 separated the crevices of Neotoma magister and N. floridana haematoreia (both used and random) along the y-axis (Fig. 1).


Adequate cover has been suggested as a primary factor determining the suitability of woodrat habitat (Fitch and Rainey, 1956; Rainey, 1956; Cudmore, 1985; Balcom and Yahner, 1996; Casdeberry et al., 2001; Chamblin et al., 2004). In our study, both species selected rock crevices with larger dimensions (height, width and depth) than were available in the surrounding environments, confirming the importance of cover to woodrats in the southern Appalachians. Woodrats are reportedly vulnerable to extreme winter conditions (Fitch and Rainey, 1956; Rainey, 1956; Nawrot and Klimstra, 1976). In the southern Appalachians, larger crevices likely enhance the ability of woodrats to survive severe winters by providing greater protective cover, especially from wind and precipitation. Larger crevices also have the added advantage of greater storage capacity for food caches and stick piles. Woodrats are thought to use stick piles in rock crevices to deter predators (Rhoads, 1903; Poole, 1940; Rainey, 1956) and to increase their protection against adverse weather (Rainey, 1956).

Our findings that both species inhabited only dry crevices indicate that the crevices selected were effective at protecting woodrats from precipitation and providing a dry area for caching food. Cudmore (1985) suggested that moisture was an important factor in the suitability of rocky sites to Neotoma magister. He noted that all occupied cliff sites along the Ohio River in Indiana were dry, and that wet sites never harbored woodrats. Our findings support Cudmore's (1985) assertion, and suggest that woodrats might avoid wet crevices to maintain the insulative effectiveness of their pelage (McFarland et al., 1985). Fitch and Rainey (1956) and Rainey (1956) also reported that dry crevices help preserve food caches and maintain nests for longer periods.


In rocky habitats, woodrats are thought to minimize the risk of predation by inhabiting rock formations containing multiple entrances and exits around dens (Rhoads, 1903; Newcombe, 1930; Poole, 1940; Rainey, 1956; Cameron and Rainey, 1972; Chamblin et al., 2004). In our study, Neotoma magister and N. floridana haematoreia both selected crevices with more internal fissures than were available in the surrounding environments, supporting this supposition. The importance of internal fissures to both species was also reflected in our discriminant analysis, as this attribute had the highest positive loading in the first canonical variable (Table 3).

Neotoma are thought to have a highly-developed tactile sense as a result of prominent vibrissae (Rhoads, 1903; Poole, 1940; Rainey, 1956; Wiley, 1980). Ray (2000) suggested that Neotoma magister may be better suited for internal fissures than N. floridana haematoreia, because of their longer vibrissae. However, the finding in our study that N. magister selected crevices with fewer internal fissures than N. f haematoreia does not support Ray's (2000) suggestion. The importance of vibrissae for Neotoma to navigate internal fissures is uncertain. Results of laboratory experiments conducted by Dunning and Payne (1979) indicated that N. magister relied on hearing rather than vibrissae for navigating a radial maze in the dark. The reason why N. f. haematoriea would select crevices with more internal fissures than N. magister is unknown and requires further investigation.

Our study suggests that crevice aspect may be an important factor that distinguishes Neotoma magister from N. floridana haematoreia in the southern Appalachians. Crevices selected by N. magister were more south-facing than those available in the surrounding environments, while N. f. haematoreia used crevices with aspects similar to the surrounding environments. This selection by N. magister for crevices with a southern exposure is also highlighted by the smaller standard deviation of aspect of used versus random crevices (Table 2). This finding is supported by other studies that also have reported that N. magister prefer rocky sites with south-facing aspects (Cudmore, 1985; Balcom and Yahner, 1996).

In the mountains of North Carolina, Neotoma magister generally occur at higher elevations than N. floridana haematoreia (Ray, 2000). The distribution in elevations of our study sites generally reflects this trend. As a result, N. magister are exposed to relatively colder climatic conditions (Holdridge, 1947). This may be the reason that N. magister select sites with southern exposures that provide warmer and drier microclimatic conditions (Chapin et al., 2002). This adaptation likely allows N. magister to survive more extreme winter conditions than N. f haematoreia.

Our study indicates that both Neotoma magister and N. floridana haematoreia are habitat specialists in the southern Appalachian Mountains. Both species selected large, dry crevices containing numerous internal fissures; attributes that are considered important to protect woodrats from extreme winter conditions and predators. Our study also found that N. magister prefer crevices with a southern exposure, which likely allow N. magister to survive colder climatic conditions, and thus inhabit higher elevations in the mountains. The small differences in attributes of rock crevices selected by both species in our study suggest that habitat is not a factor that will prevent hybridization between these species where they co-occur in the mountains of North Carolina. A possible zone of hybridization has been identified in the Linville Gorge area of Burke County, North Carolina, as both species were captured at one trap site (Ray et al., 2000). Additional studies are needed to further elucidate and compare habitat preferences of these species where they are known to co-occur and in other parts of their ranges that differ in elevation, climate and zones of contact with other species of Neotoma.

Acknowledgments.--We thank D. Ray for providing data on the study areas and for reviewing a draft of the manuscript, S. Patch for his review of the statistical methods and B. Miller for his insight on the geology of the southern Appalachians.


BALCOM, B. J. AND R. H. YAHNER. 1996. Microhabitat and landscape characteristics associated with the threatened Allegheny woodrat. Conserv. Biol., 10:515-525.

BARBOUR, R. W. AND W. H. DAVIS. 1974. Mammals of Kentucky. The University Press of Kentucky, Lexington, Kentucky.

CAMERON, G. N. ANO D. G. RAINEY. 1972. Habitat utilization by Neotoma lepida in the Mojave Desert. J. Mature., 53:251-266.

CASTLEBERRY, N. L., S. B. CASTLEBERRY, W. M. FORD, P. B. WOOD AND M. T. MENGAK. 2002. Allegheny woodrat (Neotoma magister) food habits in the central Appalachians. Am. Midl. Nat., 147:80-92.

CASTLEBERRY, S. B., P. B. WOOD, N. L. CASTLEBERRY AND M. T. MENGAK. 2001. Movements of Allegheny woodrats in relation to timber harvesting. J. Wildl. Manage., 65:148-156.

--, M. T. MENGAK AND W. M. FORD. 2006. Neotoma magister. Mammalian Species, 789:1-5.

CHAMBLIN, H. D., P. B. WOOD ANDJ. W. EDWARDS. 2004. Allegheny woodrat (Neotoma magister) use of rock drainage channels on reclaimed mines in southern West Virginia. Am. Midl. Nat., 151:346-354.

CHAPIN, F. S., III, P. A. MATSON AND H. A. MOONEY. 2002. Principles of Terrestrial Ecosystem Ecology. Springer-Verlag, New York, New York.

CUDMORE, W. W. 1985. The present distribution and status of the eastern woodrat, Neotoma floridana, in Indiana. Indiana Acad. Science, 94:621-627.

DUNNING, D. C. AND L. N. PAYNE. 1979. Orientation in cave-dwelling wood rats. Behav. Ecol. Sociobiol., 6:1-9.

FITCH, H. S. AND D. G. RAINEY. 1956. Ecological observations on the woodrat, Neotoma floridana. Univ. Kansas Pub. Mus. Nat. Hist., 8:499-533.

HOFFMEISTER, D. F. 1989. Mammals of Illinois, University of Illinois Press, Urbana, Illinois.

HOLDRIDGE, L. R. 1947. Determination of world plant formations from simple climatic data. Science, 105:367-368.

LEGRAND, H. E., S. P. HALL, S. E. MCRAE AND J. T. FINNEGAN. 2006. Natural Heritage Program list of the rare animal species of North Carolina. North Carolina Natural Heritage Program, Department of Environment and Natural Resources, Raleigh, North Carolina.

LOGIUDICE, K. 2006. Toward a synthetic view of extinction: a history lesson from a North American rodent. BioScience, 56:687-693.

MCFARLAND, W. N., F. H. POUGH, T. J. CADE AND J. B. HEISER. 1985. Vertebrate Life, 2nd ed. MacMillan Publishing Company, New York, New York.

MONY, A. M. AND R. E. EMERSON. 2003. Eastern Woodrat (Neotoma floridana and Allies), p. 381-393. In: G. A. Feldhammer, B. C. Thompson and J. A. Chapman (eds.). Wild Mammals of North America, 2nd ed. Johns Hopkins University Press, Baltimore, Maryland.

--, E. J. HEIST, E. R. WAOLE, R. E. EMERSON, E. H. NICHOLSON AND G. A. FELDHAMER. 2003. Genetic variation and population assessment of eastern woodrats in southern Illinois. Southeast. Nat., 2:243-260.

NAWROT, J. R. AND W. D. KLIMISTRA. 1976. Present and past distribution of the endangered southern Illinois woodrat ( Neotoma floridana illinoensis). Chicago Acad. Sci. Nat. Hist. Misc., 196:1-12.

NEWCOMBRE, C. L. 1930. An ecological study of the Allegheny cliff rat (Neotoma pennsylvanica Stone). J. Mature., 11:204-211.

NORTH CAROLINA GEOLOGIC SURVEY. 1985. Geologic map of North Carolina. Department of Natural Resources and Community Development, Division of Land Resources, Raleigh, North Carolina.

NOWAK, R. M. AND J. L. PARADISO. 1983. Walker's Mammals of the World, 4th ed. Volume II. Johns Hopkins University Press, Baltimore, Maryland.

POOLE, E. L. 1940. A life history sketch of the Allegheny woodrat. J. Mamm., 21:249-270.

RAINEY, D. G. 1956. Eastern woodrat, Neotoma floridana: life history and ecology. Univ. Kansas Pub. Mus. Nat. Hist., 8:535-646.

RAY, D. K. 2000. Phylogenetics and evolution of woodrats (genus Neotoma) in the southern Appalachian Mountains. M. S. thesis. University of North Carolina at Wilmington, Wilmington, North Carolina.

--, D. W. FRESHWATER AND W. D. WEBSTER. 2002. Multiplex PCR identification of eastern and Allegheny woodrats in zones of contact. J. Wildl. Manage., 66:86-90.

RHOADS, S. N. 1903. The Mammals of Pennsylvania and New Jersey. Philadelphia, Pennsylvania.

SAS INSTITUTE INC. 1990. SAS Procedures Guide, version 6th ed. SAS Institute, Inc., Cary, North Carolina.

SHAPIRO, S. S. AND M. B. WILK. 1965. An analysis of variance test for normality complete samples). Biometrika, 52:591-611.

TABACHNIK, B. G. AND L. S. FIDELL. 1989. Using Multivariate Statistics, 2nd ed. Harper & Row Publishers, Inc., New York, New York.

WILEY, R. W. 1980l Neotoma floridana. Mammalian Species, 139:1-7.

WILLIAMS, B. K. 1983. Some observations on the use of discriminant analysis in ecology. Ecology, 64:1283-1291.

C. REED ROSSELL, JR., (1) STACEY H. ROACH, (2) AND IRENE M. ROSSELL, Environmental Studies Department, University of North Carolina at Asheville, Asheville 28804 and CHRIS MCGRATH, North Carolina Wildlife Resources Commission, 315 Morgan Branch Road, Leicester 28748. Submitted 29 January 2008; accepted 22 April 2008.

(1) Corresponding author: e-mail:

(2) Present address: Department of Forestry and Environmental Resources, North Carolina State University, Raleigh 27695
TABLE 1.--General locations and approximate elevations of Neotoma
magister and N floridana haematoreia study sites in the mountains
of North Carolina

  Site name                     County            Elevation (m)

Neotoma magister
  Short-0ff Mountain            Burke County           610
  Little Lost Cove Cliff        Avery County           975
  Blue Ridge Parkway            Buncombe County       1036
  Sitting Bear Mountain         Burke County          1128
  Stackrock Creek               Avery County          1128
  Walker Falls                  Buncombe County       1158
  Woods Mountain                McDowell County       1158
  Raven Rock                    Watauga County        1158
  Mitchell Falls                Yancey County         1402
Neotoma floridana haematoreia
  Nantahala Gorge               Swain County           671
  Linville Falls                McDowell County        792
  Harmon Den                    Haywood County         792
  Granite City                  Jackson County         914
  Highlands                     Macon County           975

TABLE 2.--Means (SD) of attributes measured at used and random rock
crevices of Neotoma magister and N. floridana haematoreia in the
mountains of North Carolina. Values followed by the same letter are
not significantly different across rows (Bonferroni-adjusted alpha
level = 0.0025)

                     N. magister (n = 90)

Attribute                 Used           Random

Height (cm)           96.1 (79.3)a     38.4 (32.4)b
Width (cm)           129.0 (11.0)a     47.8 (49.7)b
Depth (cm)           165.0 (106.3)a    51.8 (26.7)b
Number of internal
  fissures             4.2 (1.9)a       1.0 (1.2)b
Aspect ([degrees])   168.3 (44.1)a    194.2 (81.6)b

                     N f. haematoreia (n = 50)

Attribute                 Used           Random

Height (cm)          104.3 (68.8)a    49.7 (27.9)c
Width (cm)           135.6 (96.0)a    50.7 (39.2)b
Depth (cm)           193.0 (128.4)a   63.8 (54.1)b
Number of internal
  fissures             5.4 (2.4)c      0.9 (1.2)b
Aspect ([degrees])   106.3 (61.3)c    92.3 (66.0)c

TABLE 3.--Canonical coefficients for attributes of used and random
rock crevices of Neotoma magister and N. floridana haematoreia in
the mountains of North Carolina

Attribute                     Can1    Can2

Height                         0.37   0.16
Width                          0.38   0.11
Depth                          0.52   0.27
Number of internal fissures    0.79   0.46
Aspect                        -0.41   0.83
COPYRIGHT 2009 University of Notre Dame, Department of Biological Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Notes and Discussion
Author:Rossell, C. Reed, Jr.; Roach, Stacey H.; Rossell, Irene M.
Publication:The American Midland Naturalist
Article Type:Report
Geographic Code:1USA
Date:Jul 1, 2009
Previous Article:Intraspecific variation in life history traits among two forms of Ambystoma barbouri larvae.
Next Article:Selection of seeds of common native and non-native plants by granivorous rodents in the Northeastern United States.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters