2007 NORTHEASTERN NATURALIST 14(3):323–342 Impacts of Land-use Management on Small Mammals in the Adirondack Park, New York Michale J. Glennon1,2,* and William F. Porter1 Abstract - We examined the response of small-mammal communities to human disturbance along a gradient from wilderness to managed forest to rural residential development in the Adirondack Park of northern New York. Our objectives were to determine if small-mammal community composition and structure differed among sites along a gradient of human impact and to relate changes in small-mammal community composition to habitat changes along the gradient. We sampled small mammals with track tubes on 10 replicates each of old growth, managed forest, and areas of residential development in the central Adirondacks. We estimated differences in species composition, abundance, diversity, evenness, and community structure, and identified habitat variables associated with differences in these measures. We used canonical correspondence analysis (CCA) to examine the relationship between patterns in small-mammal community composition and sampled habitat variables. Some small-mammal species demonstrated a numerical response to increasing human impact on the landscape, with abundance of total small mammals, eastern chipmunk, and Sorex species highest in old growth and declining in managed forests and areas of residential development. Differences among management types were reflected more strongly in total abundance and community structure than through indices of richness and diversity, which did not differ among management types. Canonical correspondence analysis revealed that 32% of the variability in the small-mammal community could be explained by habitat characteristics and that variables describing availability of coarse woody debris, as well as presence of shrubs in the understory, may create conditions favorable to many smallmammal species in these habitats. Introduction Biological diversity encompasses all forms of life and habitat, making biodiversity conservation inherently complex and often controversial. Ecological communities do not conform to political boundaries. Consequently, maintaining and protecting them often involves compromise among conflicting values and desired land uses. The Adirondack Park, in northern New York, is a prime example of such compromise. The park is a 2.3 million-ha state park, consisting of a mix of publicly owned wilderness areas interspersed among a myriad of private land uses. The roughly 1.1 million ha (48%) of the park in public ownership is part of the New York State Forest Preserve. It is protected by the New York State constitution as “forever wild forest land” and is a critical component of the Northern Forest ecosystem 1State University of New York, College of Environmental Science and Forestry, 1 Forestry Drive, Syracuse, NY 13210. 2Present address - Wildlife Conservation Society, 7 Brandy Brook Avenue, Suite 204, Saranac Lake, NY 12983. *Corresponding author - mglennon@wcs.org. 324 Northeastern Naturalist Vol. 14, No. 3 and the Champlain Adirondack Biosphere Reserve (New York State Constitution, Article XIV, Section 1, 1938; Harper et al. 1990). The remainder of the Adirondack Park is in private ownership and, while land use is regulated, is subject to a wide variety of uses and impacts to the landscape. This mix has brought much attention to the Adirondack region as an example of an application of sustainable development principles (Erickson 1998, Hohmeyer 1997, Thorndike 1997). The Adirondack Park provides an opportunity to examine how increasing human impact affects biodiversity in the context of an interspersed landscape. Much of the private land has long been owned by the forest industry. Ongoing harvest and regeneration activities have generated debate about the impacts of forest management on biodiversity. In the past 30 years, this debate has broadened to include residential development because of the growing demand for space for new and second homes (Curran et al. 1990, Harper et al. 1990, Miller 1990). These concerns have been expressed in other areas where wilderness is in transition to or interspersed with managed forests and where residential development is occurring at unprecedented rates, with unknown ecological consequences (e.g., Odell and Knight 2001, Theobald et al. 1997). Understanding the effects of increasing human development in such wilderness settings is critical to the maintenance of the biological diversity associated with these temperate forest regions. We sought to understand the impact of human activity in the Adirondack Park through examination of patterns in diversity and abundance of small mammals. Other studies have shown that small mammals are good indicators of biodiversity because of their sensitivity to environmental change, wide array of life-history strategies, and ecological importance to ecosystem function (Carey and Harrington 2001, Steele et al. 1984, Sullivan et al. 2000). We sampled small mammals along a gradient of intensity of human use, comparing abundance and diversity on 3 land uses: old-growth forest, managed forest, and rural residential development. Our objectives were to determine if small-mammal community composition and structure differed among sites along this gradient of human impact and to relate changes in small-mammal community composition to changing habitat conditions along that gradient. Study Area The Adirondack Park encompasses an area of 19,700 km2 in northern New York State (Fig. 1). The region is mountainous with elevations ranging from 30 to 1600 m. Average annual summer temperatures in the park are between 18 and 21 °C and winter temperatures between -1 and 5 °C. Mean annual precipitation is 101–122 cm, and average snowfall is 152–356 cm. Dominant vegetation is a mixture of boreal and northern hardwood forest containing Picea rubens Sarg. (red spruce), Picea mariana P. Mill. (black spruce), Abies balsamea (Linn.) Mill. (balsam fir), Tsuga canadensis (Linn.) Carr. (eastern hemlock), Acer saccharum Marsh (sugar maple), Acer rubrum L. (red maple), 2007 M.J. Glennon and W.F. Porter 325 Betula alleghaniensis Britton (yellow birch), Betula paperyfera Marsh (white birch), and Fagus grandifolia Ehrh. (beech). Old-growth forest areas included in our study consisted of northern hardwood forest stands unharvested during the past 100 years or more. Oldgrowth stands contained canopy-dominant trees with ages at least one-half the maximum life expectancy for respective species, consistent with previous definitions of old-growth conditions in the Adirondack Park (McGee et al. 1999). Records for 4 of the old-growth sites indicated no past cutting. Research has shown that approximately 2000 km2 of Adirondack forests were so minimally disturbed that they appear today as they did to the earliest white visitors to the region (McMartin 1994). The remaining 6 old-growth stands were acquired by the state between 1871 and 1891 (McMartin 1994) and have not been harvested since. All old-growth stands were located on mesic, midslope sites dominated by mixed northern hardwood overstory and were larger than 2 ha. Managed forests were partially cut, uneven-aged northern hardwood stands, also larger than 2 ha in size and located on mesic, mid-slope sites. Detailed stand-level records for these sites were not available, but all sites have received a variety of partial cuttings for at least the past 100 years (see McGee et al. [1999] for a description of 6 of the 12 sites). Most managed stands had at least 1 old skid trail adjacent to or partially within the sampling grid. Figure 1. Adirondack Park, showing state and private lands, with locations of study sites (■) shown. 326 Northeastern Naturalist Vol. 14, No. 3 Developed sites consisted of areas of residential use within the towns of Newcomb, Ray Brook, Long Lake, Lake Placid, Blue Ridge, Tupper Lake, and Saranac Lake in the central part of the Adirondack Park. Developed sites were in single ownerships on mesic, midslope sites, at least 2 ha in size, and characterized by an interspersion of northern hardwood or mixed northern hardwood forest and openings in the form of lawn surrounding or directly adjacent to at least 1 year-round residence. Residential structures ranged in age from 27–142 years. Most residential owners had pets (domestic cats or dogs). These sites were characterized by an average of approximately 40% lawn and 60% forest, and the average distance to the next nearest house was 170 m. All sites were within 100 m of contiguous forest. Our sample locations can all be classified as rural residential development, and share many similarities with the exurban development pattern characterized by single units interspersed in a wilderness setting that has begun to appear in the Rocky Mountain West and the Sierra Nevada (Crump 2003, Esparza and Carruthers 2000, Hansen et al. 2002, Maestas et al. 2001). Methods We characterized small-mammal species richness and abundance by sampling 10 replicates in each of old-growth forest, managed forest, and rural residential development (Fig. 1). Locations in old-growth stands and residential development were sampled in each of 2 years. We replaced stands in 2 managed forest sites in the second year because logging operations blocked access to the original sites; the remaining 8 managed stands were sampled in both years. Small mammals were sampled during May–August of 2001 and 2002. Small mammals were sampled with track tubes (Glennon et al. 2002) on 7 x 7 grids with 20-m spacing between track-tube locations. Small-mammal data consisted of numbers of detections of small-mammal species within track tubes summed over the total number of track-tube nights at each grid. Each grid was sampled over two 48-hr sampling periods, resulting in a total of 98 tube nights for each location. For example, if chipmunks were detected in 36 of 49 tubes in the first sampling period and 40 of 49 tubes in the second sampling period, the total detection rate for chipmunk at that location would be 36 + 40 = 76 observations in 98 tube nights, or 76/98. A complete description of the sampling technique is provided in Glennon et al. (2002). In short, track tubes are constructed from 30-cm sections of plastic rain gutter which provides protection for a sticky track surface on the inside. The track surface is made from contact paper and small squares with ink on them are placed at either end. The center of each tube is baited and animals can then enter tubes from either end, transferring ink onto the contact paper and leaving footprints. Unlined white paper is then adhered to the contact paper to preserve a permanent record of the tracks from each track tube (Fig. 2). We tested this technique against a more traditional live-trapping approach in a separate study (Glennon et al. 2002) and found 2007 M.J. Glennon and W.F. Porter 327 that relative-abundance data obtained from track tubes correlated highly with those obtained through live trapping. Track tubes do not allow individual animals to be distinguished, but a number of species are easily identified. We were able to distinguish footprints of Blarina brevicauda (Say) (short-tailed shrew), Glaucomys sabrinus (Shaw) (northern flying squirrel), Tamiasciurus hudsonicus (Erxleben) (red squirrel), Sciurus carolinensis Gmelin (grey squirrel), and Tamias striatus L. (Richardson) (eastern chipmunk) (Glennon et al. 2002). Several other species could not be distinguished from one another due to similarity of print characteristics. These included: 1) Napaeozapus insignis Preble (woodland mouse) and Zapus hudsonius (Zimmermann) (meadow jumping mouse); 2) Peromyscus maniculatus Wagner (deer mouse) or Peromyscus leucopus (Rafinesque) (white-footed mouse) and Myodes gapperi (Vigors) (red-backed vole); and 3) Sorex cinereus Kerr (masked shrew) and Sorex fumeus (Miller) (smoky shrew). We combined these species into groups representing jumping mice, mouse/vole, and Sorex, respectively. Figure 2. Small-mammal track tube constructed from plastic rain gutter, with track surface and eastern chipmunk tracks shown. 328 Northeastern Naturalist Vol. 14, No. 3 We collected habitat data on all sample sites at 10 points chosen randomly from among the 49 points per grid and pooled data across the 10 sample points for each grid. Habitat data consisted of live-tree data, coarse woody debris, and microhabitat variables (Table 1). Plot sizes and methods were comparable to habitat sampling techniques used on long-term smallmammal monitoring sites at Huntington Wildlife Forest (Newcomb, NY), the location of our prior study (Glennon et al. 2002). On each of the 10 randomly chosen points per grid, all trees larger than 5 cm dbh were tallied and measured for diameter, and all species were noted within 19.6 m2 circular plots. We separated understory trees into 2 groups, those <1 cm in size and those between 1 and 5 cm, and counted the total number of trees in each group, noting species of each. Coarse woody debris (CWD) was also measured on all plots. Dead woody material was tallied within each of 3 different size (diameter) classes (<20 cm, 20–40 cm, and >40 cm) and placed into one of 3 different decay classes (1, 2, or 3) (McGee 1998). Species of woody debris was noted if it could be determined, and status was noted as log, stump, or snag. We noted the presence of the most significant components of the habitat at ground level at each of the 10 points sampled per grid, Table 1. Means and standard errors for habitat data collected on small mammal sampling grids in old-growth forest, managed forest, and areas of residential development in Adirondack Park, 2000. The acronyms CWD and VSD refer to coarse woody debris and vertical structural diversity, respectively. Different superscripts denote statistical differences in pairwise tests; means with different superscripts are statistically different. See methods section for a full description of variables. Old growth Managed Developed ANOVA Habitat variable Mean SE Mean SE Mean SE P Trees < 1 cm (#) 293.3 98.7 481.0 104.7 255.3 93.7 0.261 Trees 1-5 cm (#) 50.7 10.8 39.4 12.1 22.0 10.8 0.188 Trees > 5 cm (#) 11.3 2.9 12.5 3.3 17.1 2.9 0.353 Mean DBH (cm) 18.2 1.5 17.0 1.7 13.8 1.5 0.118 Total CWD (# items) 23.2 3.2 25.1 3.6 14.9 3.2 0.084 Mean size class CWD 1.6A 0.04 1.3B 0.05 1.1C 0.04 0.001 Mean decay class CWD 2.4A 0.1 2.1AB 0.1 1.9B 0.1 0.003 Logs (#) 18.6A 2.4 18.9A 2.6 7.3B 2.4 0.001 Snags (#) 3.8 1.5 4.3 1.7 5.0 1.5 0.848 Stumps (#) 0.8 0.6 2.0 0.7 2.6 0.6 0.155 Litter depth (cm) 3.1A 0.3 2.9 0.3 1.7B 0.3 0.005 Canopy cover (%) 97.4A 5.2 91.6A 5.8 62.9B 5.2 0.001 VSD 1.0 1.2 1.8 1.3 4.1 1.2 0.167 Mean slope 12.8 2.8 12.3 3.2 5.2 2.8 0.135 Leaf litter 8.8A 0.7 8.3A 0.8 5.3B 0.7 0.003 CWD 7.6A 0.8 7.6A 0.9 5.0B 0.8 0.049 Shrub 5.1A 0.7 1.4B 0.8 1.1B 0.7 0.001 Herbaceous 1.8A 1.0 6.6B 1.1 5.2B 1.0 0.008 Rock 2.3 0.7 1.3 0.8 1.6 0.7 0.568 Fern 2.6 0.8 3.3 0.9 0.9 0.8 0.138 Moss 0.5 0.5 0.4 0.6 1.8 0.5 0.155 Lawn NA NA NA NA 2.5 0.6 0.006 Softwood (%) 9.3A 6.6 4.7A 7.3 50.5B 6.6 0.001 2007 M.J. Glennon and W.F. Porter 329 recording the characteristics of the forest floor at each sampling point as made up primarily of 3 of the following: rock, moss, fern, herbaceous vegetation, coarse woody debris, shrub, leaf litter, and lawn. We measured 4 other variables at each location: canopy cover (taken from 4 measurements in each of the cardinal directions from the plot center), slope, leaf-litter depth (taken from 4 measurements at the plot boundary in each of the cardinal directions), and vertical structural diversity (VSD) below 1 m. We calculated vertical structural diversity as the number of individual plant parts touching a 1-m PVC pipe placed at the plot boundary in each cardinal direction. Vertical structural diversity has been shown to be important in determining characteristics of bird and other faunal communities (MacArthur and MacArthur 1961, Raizer and Amaral 2001). We examined differences among the 3 land-use types in individual species abundance, richness, diversity, and evenness of small mammals. We calculated diversity using Fisher’s alpha statistic and the Simpson evenness index (Magurran 1988). We used analysis of variance by land-use type over years with type-year interaction to determine if there were significant differences in small-mammal abundance between years (Carey and Johnson 1995). We then used analysis of variance and Bonferroni post hoc (Miller 1985) tests to determine which response variables differed among management types. Finally, we used CCA (Attayde and Bozelli 1998, Bellows et al. 2001, du Bus de Warnaffe and Dufrêne 2004, Miller et al. 2003, Ter Braak 1986) to explore relationships between small-mammal community characteristics and habitat variables. Canonical correspondence analysis is a direct-gradient analysis method which allows for the simultaneous exploration of a large number of habitat variables and is not overly sensitive to species with skewed distributions (Palmer 1993, Ter Braak 1986). We examined habitat variables for multicollinearity, and highly correlated variables were combined or omitted from analysis. We measured a total of 24 habitat variables at each of the study sites. This number of environmental variables is too large to effectively relate their independent effects on small-mammal community structure and composition (Ter Braak 1986). Therefore, after removing highly correlated variables, we reduced the remaining list to 7 variables, which were hypothesized to capture key features in the habitat and to be of importance to small-mammal communities. These variables included the total amount of coarse woody debris (Carey and Johnson 1995), the average size (diameter class) of coarse woody debris (Bellows et al. 2001), canopy cover (Bellows et al. 2001, Carey and Johnson 1995), prevalence of shrub (Bellows et al. 2001, Carey and Johnson 1995, Racey and Euler 1982), herbaceous cover, and rock in the understory, and the percentage of conifer cover in the overstory (Racey and Euler 1982). Canonical correspondence analysis is performed by initially summarizing the variation in the species data by ordination, and then relating the ordination axis to the environmental variables (Ter Braak 1986). We used correspondence analysis (CA) to explore the initial 330 Northeastern Naturalist Vol. 14, No. 3 structure of the small-mammal community data, and followed with CCA to relate patterns in small-mammal community composition to the environmental variables. All analyses were performed in ADE-4 (Thioulouse et al. 1997). The importance of ordination axes in the CCA was interpreted by their eigenvalues and the intraset correlations, which are the correlations between the environmental variables and the ordination axes themselves (Ter Braak 1986). We standardized environmental variables to zero mean and unit variance in order to remove the effects of different units of measurement and make the canonical variables comparable to each other (Ter Braak 1986). The species-environment correlations, or the correlations between the sample scores derived for an axis from the species data and the sample scores that are linear combinations of the environmental data (Attayde and Bozelli 1998), were used to determine the degree to which the measured environmental variables are sufficient to explain the major variation in the species data (Ter Braak 1986). Lastly, the CCA ordination diagram presented an overall picture of the relationship between the species and environmental variables. In an ordination diagram, species and sites are represented by points, and the environmental variables are represented by arrows, with species points and arrows jointly representing the species’ distributions along each of the environmental variables (Ter Braak 1986). A projected inertia decomposition following CCA was used to determine the proportion of the variance in the abundance of each species that can be explained by the multiple regression of the CCA and the total variability in the species data that can be accounted for by the environmental data (Thioulouse et al. 1997). Results Small-mammal community We corrected data prior to statistical analysis to account for track tubes disturbed by Ursus americanus Pallas (black bears) and mustelids (Mustela spp. or Martes spp.) by subtracting one half of 1 tube night for each track tube affected, in a manner analogous to corrections often made for sprung traps in small-mammal live-trapping data (Nelson and Clark 1973). Corrected data were logarithmically transformed to meet assumptions of normality (Carey and Harrington 2001, Caughley 1977). Detections of northern flying squirrel and grey squirrel were too few for statistical analysis. Analysis of variance did not detect any significant year-by-type interactions and, therefore, small-mammal abundances were averaged over 2000–2001. We found differences among land-use types for total abundance and evenness (Table 2). Total abundance was highest in old-growth forest and declined along the gradient to its lowest point in areas of residential development (Table 2). Old growth and residential development were similar and higher in evenness than managed forest. Individual species that showed greatest changes were squirrels and shrews. Sorex was lowest in residential development, and eastern chipmunk was lowest in 2007 M.J. Glennon and W.F. Porter 331 managed forest; both reached peak abundance in old growth. Red squirrel was lowest in old growth and highest in residential development. In general, small-mammal communities in developed areas consisted of a larger proportion of sciurids than the other 2 types, whereas old-growth sites had a larger proportion of muroid and dipodid species. Habitat relationships Old growth, managed forest, and residential development differed in terms of measured habitat variables (Table 1). Old-growth sites were characterized by a live-tree structure consisting of fewer but larger trees than the other two types, in addition to a well-developed layer of second growth (trees 1–5 cm dbh). Old-growth sites also generally possessed a closed canopy, deeper litter layer, a greater presence of litter, coarse woody debris, shrub, and rock as microhabitat features, and were found on steeper slopes. Old-growth sites had more logs and fewer stumps and snags than the other two types. Debris in the old growth was significantly larger and more decayed than in the other two types. There was an overall greater diversity of size and decay of coarse woody debris items in old-growth stands. Managed forest sites were characterized by the highest abundance of seedlings and a large amount of small, less-decayed coarse woody debris. Managed sites had more open canopies and shallower litter layers than old-growth sites, and were located on relatively flat sites with the largest occurrence of ferns and other herbaceous vegetation as major components of the microhabitat. Lastly, developed sites contrasted highly with the other two types and were more variable in terms of most habitat components. Residential development had a high abundance of trees >5 cm dbh, but few trees in smaller size classes. Developed sites had the largest number of snags and stumps relative Table 2. Abundance (raw, uncorrected data; denotes total counts of tracks of species detected in track tubes), diversity, and evenness of small mammals in old-growth forest, managed forest, and areas of residential development in the Adirondack Park, 2000–2001, and analysis of variance. Corrected, transformed data were used in statistical analyses. Different superscripts denote statistical differences in pairwise tests; means with different superscripts are statistically different. Old Growth Managed Developed ANOVA Abund. SE Abund. SE Abund. SE F P Blarina brevicauda 52 1.477 50 1.154 43 0.847 0.222 0.802 Glaucomys sabrinus 14 0.411 3 0.187 0 0 NA NA Napaeozapus insignis/ 228 3.991 93 2.379 130 2.820 0.817 0.453 Zapus hudsonius Peromyscus/ 873 5.389 625 6.442 660 4.532 1.218 0.313 Myodes gapperi Sorex 93A 1.168 28B 0.674 10B 0.212 7.962 0.002 Sciurus carolinensis 0 0 0 0 29 0.942 NA NA Tamiasciurus hudsonicus 9A 0.400 19a 0.941 162b 2.557 4.788 0.017 Tamias striatus 1119A 6.440 406B 6.811 669B 4.938 8.194 0.002 Total abundance 2388A 10.932 1224AB 11.659 1703B 8.403 3.594 0.032 Fisher’s alpha diversity 0.790 0.069 0.837 0.140 0.954 0.082 0.826 0.449 Simpson’s evenness 2.384AB 0.134 1.965B 0.189 2.598A 0.152 3.950 0.032 332 Northeastern Naturalist Vol. 14, No. 3 to the other 2 types, but coarse woody debris in these sites was much smaller and much less decayed than in managed or old-growth sites. Both litter depth and canopy cover were low in developed areas, and the microhabitat was composed mainly of lawn, herbaceous vegetation, and rock. Comparison of the ordination results (Table 3) showed that the first eigenvalue of the CCA was lower but of similar magnitude to the first eigenvalue of the CA, and the species-environment correlations were moderately high. This indicated that the appropriate environmental variables had been measured and can account for some proportion of the variability in the species data. Examination of the canonical correlations and the intraset correlations (Table 4) revealed that axis 1 was most closely associated with softwood cover and contrasted sites of high conifer composition with those that had an understory characterized by shrubs and the presence of large diameter coarse woody debris. Axis 2 was also related to softwood cover, but contrasted high softwood cover with sites with large amounts of coarse woody debris. The CCA ordination diagram (Fig. 3) demonstrates the differences in habitat structure among the three land-use types, with developed sites generally lying on the positive side of axis 1 and corresponding with high softwood cover and lower values of total woody debris, canopy, and shrub cover. Managed sites clustered primarily below axis 1 and were more closely related to higher canopy, shrub, and herbaceous cover and higher amounts of coarse woody debris. Finally, old-growth sites were distributed primarily to the right of axis 2, and related to higher shrub and canopy cover and large diameter coarse woody debris. Table 3. Eigenvalues and species-environment correlations from a canonical correspondence analysis (CCA) of small-mammal data from old-growth, managed forest, and residential development sites in the Adirondack Park, NY. Axis 1 Axis 2 Eigenvalues CA 0.20 0.17 CCA 0.12 0.02 Correlation coefficients CCA 0.73 0.59 Table 4. Canonical coefficients and the intraset correlations of habitat variables with the first two axes of canonical correspondence analysis (CCA) of small-mammal data from old-growth, managed, and residential development sites in the Adirondack Park, NY. Canonical coefficients Correlation coefficients Axis variable 1 2 1 2 Total coarse woody debris -0.41 -0.54 -0.16 -0.46 Average-size coarse woody debris 0.07 0.64 0.48 0.13 Canopy closure 0.34 0.24 0.33 -0.20 Shrub 0.38 -0.46 0.58 -0.37 Herb 0.09 -0.55 0.003 -0.02 Rock -0.40 -0.78 -0.36 -0.37 % softwood -0.46 0.71 -0.78 0.46 2007 M.J. Glennon and W.F. Porter 333 The distributions of species points in the ordination diagram demonstrated strong relationships to the environmental gradients for some species and a weaker association with environmental variables for other species. A few species, or species groups, namely eastern chipmunk, mouse/vole, and the short-tailed shrew, clustered somewhat toward the center of the ordination diagram, which may indicate that they are found across most of the measured gradients. This pattern is similar to that revealed by the community structure (Table 2), which shows that both mouse/vole and the short-tailed shrew did not differ statistically across land-use types, while the eastern chipmunk did have a higher relative abundance in old-growth forest. Other species were distributed more closely to the ends of gradients in the CCA diagram. The Sorex species showed a strong association with the positive end of axis 1, indicating an affiliation for sites with large woody debris and high shrub and canopy cover. The northern flying squirrel was detected only in managed and old-growth sites and, on our diagram, showed a preference for high shrub and canopy cover with large woody debris. The red squirrel was found at the high end of the gradient Figure 3. The distribution of 8 small-mammal species or species groups detected on old-growth, managed forest, and residential development sites in the Adirondack Park, NY. Canonical correspondence analysis (CCA) ordination diagram with species (•), study sites (■, 􀂨,º), and environmental variables (arrows). The small mammals are: Blbr = Blarina brevicauda, Glsa = Glaucomys sabrinus, Nain/Zahu = Napaeozapus insignis/Zapus hudsonius, Pema/Myga = Peromyscus/Myodes gapperi, Scca = Sciurus carolinensis, Sorex, Tahu = Tamiasciurus husdonicus, and Tast = Tamias striatus. The environmental variables are: AVERAGE SIZE CWD = average size (diameter class) of coarse woody debris items; SHRUB = frequency of shrub as major forest floor habitat feature; CANOPY = percentage canopy cover; HERB = frequency of herbaceous vegetation as major forest floor habitat feature; ROCK = frequency of rock as major forest floor habitat feature; TOTAL CWD = total number of logs, stumps, and snags; and % SOFTWOOD = percentage conifer species in overstory. 334 Northeastern Naturalist Vol. 14, No. 3 for softwood cover, while the grey squirrel, which was found only in developed sites, also seemed to show a preference for sites with high softwood cover and rock as a microhabitat feature. A projected inertia decomposition (Table 5) revealed that the percentage of variation explained by the CCA ranged from 5% (jumping mice) to 59% (red squirrel) and that, overall, 32% of the variability in species distributions could be accounted for by the environmental variables. Discussion Our study examined the relationship between biodiversity and land-use management in the context of a highly forested landscape characterized by broad interspersion of managed and unmanaged land-use types. Our first objective was to determine how the structures of small-mammal communities changed along a gradient of increasing human impact from old growth to managed forest to rural residential development. Differences among management types were reflected more readily in total overall abundance and individual species’ abundances than through indices reflecting richness and diversity. Examination of community structures revealed shifting patterns in response to the land-use gradient. Increasing human impact generally resulted in increasing representation of sciurids and decreasing representation of soricids, while muroids and dipodids remained a significant component of all habitats. Further, increasing human impact resulted in an increase in representation of more generalist species such as red squirrel and eastern chipmunk, and a decline in more specialized species such as the northern flying squirrel. We also found that, while small mammals did not differ in terms of richness, diversity, or evenness across land-use types, overall abundance was highest in old-growth forests and lower in areas of managed forest and residential development. These results agree with other investigations (Carey 1995, 2000; Carey and Johnson 1995; Carey et al. 1999) of small-mammal abundance and diversity in varying management types. Small-mammal populations in managed, naturally young, and old- Table 5. Projected inertia decomposition from a canonical correspondence analysis (CCA) of small-mammal data from old-growth, managed, and residential development sites in the Adirondack Park, NY. Variance (%) explained Species by multiple regression Unexplained variance (%) Blarina brevicauda 36.73 63.26 Glaucomys sabrinus 38.70 61.29 Napaeozapus insignis/Zapus hudsonius 4.95 95.04 Peromyscus/Myodes gapperi 17.42 82.57 Sorex 39.01 60.98 Sciurus carolinensis 56.48 43.51 Tamiasciurus hudsonicus 58.89 41.10 Tamias striatus 19.05 80.94 Total 32.40 67.59 2007 M.J. Glennon and W.F. Porter 335 growth forests on the Olympic Peninsula did not differ in evenness, richness, or rank order, but old growth supported 1.5 times more individuals and biomass than managed forest (Carey and Johnson 1995). Our second objective was to determine how habitat structure related to changes in small-mammal community structure and to elucidate overall relationships between environmental variables and species abundance. Several structural characteristics of the habitat were different among old growth, managed forest, and developed areas, and demonstrated relationships to small-mammal community structure. The amount of variability we found in the small-mammal data attributable to environmental variables (32%; Table 5) is comparable to other studies of habitat affinities of small mammals (Bellows et al. 2001). Ter Braak (1986) points out that, in interpreting percentages of variance accounted for by CCA, 100% is not the goal, as some portion of the total variance is due to noise in the data, and that an ordination diagram that explains only a small portion of the variability in the species data may still be quite informative. Our CCA revealed a number of changes in small-mammal community structure that appear to be related to the management types investigated. The least disturbed land-use type on our gradient was that of oldgrowth forest. Canonical correspondence analysis revealed that northern flying squirrels and shrews of the genus Sorex were associated with sites characterized by high canopy and shrub cover, and large coarse woody debris items (Fig. 3). The northern flying squirrel has been found in higher abundance in old growth relative to managed forest in a number of studies (Carey 1989, 1991; Carey et al. 1992; Witt 1992), and higher abundances have been attributed both to the availability of cavities (Carey et al. 1992) and to the frequency of hypogeous sporocarps of mycorrhizal fungi, a primary food source (Waters and Zabel 1995). Its mycophagous behavior and dependence on cavities make the northern flying squirrel relatively specialized in its habitat preference for areas with a high proportion of snags and favorable growing conditions for fungi. Abundance of hypogeous fungi has been found to be positively correlated with the presence of downed logs (Amaranthus et al. 1994, Clarkson and Mills 1994) and was found to be lower in managed young Tsuga heterophylla (Ref.) Sarg. (western hemlock) stands than in naturally mature or old-growth stands (North et al. 1997). Soricid shrews also have specialized feeding habits, concentrating primarily on insect and invertebrate prey. Downed logs provide refugia for invertebrates (Harmon et al. 1986) and also trap moisture and increase local humidity (Harmon et al. 1986, Tallmon and Mills 1994, Yahner 1986). Shrews are sensitive to water loss due to their high metabolic requirements (Getz 1961, Miller and Getz 1977). The greater representation of northern flying squirrel and Sorex species in old-growth forests in this study, and overall higher total small-mammal abundance, may be related to the structural complexity provided by shrub and herb cover in the understory and an abundance of large, well-decayed coarse 336 Northeastern Naturalist Vol. 14, No. 3 woody debris, which may provide critical microclimates, cover, and food resources in these stands. Timber harvest was the major form of human impact in managed forest sites, representing the intermediate point on the disturbance gradient. No small-mammal species associated strongly with the environmental variables most closely related to managed sites. Other researchers have also detected ambiguous responses of small-mammal populations to varying forest management practices including clearcutting, shelterwood, and stripcutting (Brooks and Healy 1988, Kirkland 1990, Scott et al. 1982, Swan et al. 1984, Von Trebra et al. 1998). In contrast to other regions in the Northern Forest, clearcutting is not permitted in the Adirondack Park on areas larger than 25 acres (10 ha), and it is more common to find selective or partial harvesting methods used as a means of regeneration (K. Didier, Wildlife Conservaion Society, Bronx, NY, pers. comm.). The authors of this paper have also investigated large-scale responses of bird communities to forest management and residential development in the Adirondacks and found few differences in bird-community integrity between wilderness areas and areas of resource management that have commonly been associated with forestry (Glennon and Porter 2005). It may be that, for both bird and small-mammal communities, forest management, with the restrictions placed upon it in the Adirondack Park and in the context of a wilderness setting, does not have large impacts on faunal communities (Glennon and Porter 2005). The endpoint of our gradient and the other major human-disturbance factor assessed in our study was rural residential development. The centers of species distributions on the CCA diagram did not overlap particularly well with the locations of developed sites (Fig. 3), but red squirrel and grey squirrel were located at the ends of environmental gradients describing characteristics more closely related to residential areas than to managed or old-growth forest. Increases in these more generalized sciurid species around areas of residential development may indicate the importance of a wider variety of available food resources. Both species, as well as the eastern chipmunk, are seed consumers and respond to supplemental food sources, such as bird feeders, resulting from human inhabitance (Boutin 1990, Bowers and Breland 1996). Such additional food sources in developed areas probably contributed to increased representation of species capable of exploiting numerous resources. There was significant conifer cover surrounding the residential development sites, which represents an important food source for red squirrels (Bayne and Hobson 2000), and probably best explains their presence in these areas. Bayne and Hobson (2000) also note that food resources may be enhanced for the generalist red squirrel in fragmented areas because of increased cone production by conifers along forest edges. Oak, a significant food source for grey squirrels (Korschgen 1981), was not a large component of the forest at any of our study site locations and is found more commonly in the southern part of the Adirondacks and the 2007 M.J. Glennon and W.F. Porter 337 Lake Champlain valley. Presence of grey squirrels in residential areas and not in old-growth or managed sites may indicate that these habitats provide suitable resources for grey squirrels in parts of the Adirondacks where they are otherwise somewhat rare. Saunders (1988) suggests that grey squirrels in more central locations in the Adirondack Park would probably not exist without the presence of bird feeding stations. Our results correspond with other studies investigating the relationship between small mammals and development. Though such studies are far less common than those investigating forest-management effects, effects of development are somewhat species specific. Development in central Ontario had a significant effect on the small-mammal community. In two separate studies, eastern chipmunk, short-tailed shrew, and red squirrel were tolerant or positively related to development, whereas the masked shrew, red-backed vole, and woodland jumping mouse were negatively related to development (McDonnell et al. 1984, Racey and Euler 1982). Small mammals in Oxford, UK had similar richness in most or all habitat patches and responded to the urban environment only indirectly as it modified the growth or structure of vegetation (Dickman and Doncaster 1987). Though no single habitat factor was of primary importance to all smallmammal species in all habitat types, variables relating to cover and food resources for small mammals helped to explain changing small-mammal community structures amongst the three land-use types. Amount and size of woody debris, as well as the presence of high canopy cover and abundant shrubs, showed positive associations with the centers of the distributions for northern flying squirrel, short-tailed shrew, shrews of the genus Sorex, and the mouse/vole group. Carey and Johnson (1995) also found that abundance of coarse woody debris, in addition to prevalence of shrub cover, was a primary determinant of small-mammal abundance. Bellows et al. (2001) found that mean diameter of woody debris, canopy closure, and shrub cover strongly influenced the distributions of habitat generalists such as the shorttailed shrew and white-footed mouse. Carey and Harrington (2001) found that most small-mammal species were significantly related to a habitat factor describing a gradient of coarse woody debris. In this study, coarse woody debris was least available, smallest, and least decayed in residential areas, which may reduce its utility for small mammals by reducing its usefulness as nesting or escape cover or by decreasing its ability to serve as a substrate for invertebrate or fungal food resources. In areas of residential development, rock may have provided an alternative source of nesting and escape cover. Our research infers that maintaining coarse woody debris, as well as a prevalent understory of trees or shrubs, can create conditions favorable to many small-mammal species in forested environments. Bellows et al. (2001) drew similar conclusions for the maintenance of small-mammal communities in the upper coastal plain of Virginia. Permitting large logs and stumps to remain and reach significant stages of decay can aid in providing cover, nesting, and food resources for a large number of species. 338 Northeastern Naturalist Vol. 14, No. 3 Supplemental food in the form of bird feeders may enable or enhance persistence of some small mammals in residential areas. The gradient encompassed in the transition between old-growth forest, managed forest, and low-density rural development represents a trajectory representative of actual land-use change for open-space lands in the Adirondacks and elsewhere, both in the past and possibly in the future. This trajectory is especially relevant with respect to the transition to residential development. Though much of the known remaining old-growth forest is protected on state-owned lands, most of the virgin Adirondack forest of 1800 was logged in some manner by 1885 (Jenkins and Keal 2004). The forest has recovered significantly from the large-scale clear-cutting that took place in some regions, and forest management has historically played a large role in the economy of the park. Today, however, management and timber companies are experiencing diminishing returns in the Adirondacks and finding that the value of their lands for development often exceeds its value for timber management or other extractive uses (Harper et al. 1990). This pattern is characteristic of much of the Northern Forest region, and a trend of selling of major holdings of timber companies in the area is occurring (Harper et al. 1990, Jenkins and Keal 2004). The State of New York, together with land trusts such as The Nature Conservancy and the Open Space Institute, make what efforts they can to protect lands in the Adirondacks through acquisition or easement, but funds are limited. Understanding the potential changes associated with a transition from old forest to managed forest to residential development provides key insights into the ecological effects of such changes and leads to suggestions for how ecological integrity may be maintained. Acknowledgments Funding and logistical support for this research were provided by: the New York State Department of Environmental Conservation, the Roosevelt Wildlife Station at the State University of New York, College of Environmental Science and Forestry; and the Adirondack Ecological Center in Newcomb, NY. Data were collected by S. Duerr, D. Copney, Z. Hart, J. Murphy, J. Oldroyd, A. Peck, and E. Speith. Literature Cited Amaranthus, M., J.M. Trappe, L. Bednar, and D. Arthur. 1994. Hypogeous fungal production in mature Douglas-fir forest fragments and surrounding plantations and its relation to coarse woody debris and animal mycophagy. Canadian Journal of Forest Research 24:2157–2165. Attadye, J.L., and R.L. Bozelli. 1998. Assessing the indicator properties of zooplankton assemblages to disturbance gradients by canonical correspondence analysis. Canadian Journal of Fisheries and Aquatic Sciences 55:1789–1797. Bayne, E., and K. Hobson. 2000. Relative use of contiguous and fragmented boreal forest by red squirrels (Tamiasciurus hudsonicus). Canadian Journal of Zoology 78:359–365. 2007 M.J. Glennon and W.F. Porter 339 Bellows, A.S., J.F. Pagels, and J.C. Mitchell. 2001. Macrohabitat and microhabitat affinities of small mammals in a fragmented landscape on the upper coastal plain of Virginia. American Midland Naturalist 146:345–360. Boutin, S. 1990. Food supplementation experiments with terrestrial vertebrates: Patterns, problems, and the future. Canadian Journal of Zoology 68:203–220. Bowers, M.A., and B. Breland. 1996. Foraging of gray squirrels on an urban-rural gradient: Use of the GUD to assess anthropogenic impact. Ecological Applications 6:1135–1142. Brooks, R.T., and W.M. Healy. 1988. Response of small-mammal communities to silvicultural treatments in eastern hardwood forests of West Virginia and Massachusetts. Pp. 313–318, In R.C. Szaro, K.E. Severson, and D.R. Patton (Eds.). Management of Amphibians, Reptiles, and Small Mammals in North America. United States Department of Agriculture Forest Service, Fort Collins, CO. General Technical Report RM-166. Carey, A.B. 1989. Wildlife associated with old-growth forests of the Pacific Northwest. Natural Areas Journal 9:151–162. Carey, A.B. 1991. The biology of arboreal rodents in douglas fir forests. United States Department of Agriculture Forest Service, Portland, OR. General Technical Report 276. Carey, A.B. 1995. Sciurids in pacific northwest managed and old-growth forests. Ecological Applications 5:648–661. Carey, A.B. 2000. Effects of new forest management strategies on squirrel populations. Ecological Applications 10:248–257. Carey, A.B., and C.A. Harrington. 2001. Small mammals in young forests: Implications for management for sustainability. Forest Ecology and Management 154:289–309. Carey, A.B., and M.L. Johnson. 1995. Small mammals in managed, naturally young, and old-growth forests. Ecological Applications 5:336–352. Carey, A.B., S.P. Horton, and B.L. Biswell. 1992. Northern Spotted Owls: Influence of prey base and landscape character. Ecological Monographs 62:223–250. Carey, A.B., J. Kershner, B. Biswell, and L. Dominguez de Toledo. 1999. Ecological scale and forest development: Squirrels, dietary fungi, and vascular plants in managed and unmanaged forests. Wildlife Monographs 142:1–71. Caughley, G. 1977. Analysis of Vertebrate Populations. John Wiley and Sons, New York, NY. Clarkson, D.A., and L.S. Mills. 1994. Hypogeous sporocarps in forest remnants and clearcuts in southwest Oregon. Northwest Science 68:259–265. Curran, R.P., W.S. O’Neill, and D.S. Smith. 1990. The environmental implications of development. Pp. 346–405, In Bauer, P., and G. Cox, (Eds.). The Adirondack Park in the 21st Century, Volume 2, State of New York, Albany, NY. Crump, J.R. 2003. Finding a place in the country–exurban and suburban development in Sonoma County, California. Environment and Behavior 35:187–202. Dickman, C.R., and D.P. Doncaster 1987. The ecology of small mammals in urban habitats I: Populations in a patchy environment. Journal of Animal Ecology 56:629–640. du Bus de Warnaffe, G., and M. Dufrêne. 2004. To what extent can management variables explain species assemblages? A case study of carabid beetles in forests. Ecography 27:701–714. 340 Northeastern Naturalist Vol. 14, No. 3 Erickson, J.D. 1998. Sustainable development and the Adirondack experience. Adirondack Journal of Environmental Studies 5(2):24–32. Esparza, A.X., and J.I. Carruthers. 2000. Land-use planning and exurbanization in the rural Mountain West–evidence from Arizona. Journal of Planning Education and Research 20:23–36. Getz, L.L. 1961. Factors influencing the local distribution of shrews. American Midland Naturalist 65:67–88. Glennon, M.J., and W.F. Porter. 2005. Effects of land-use management on biotic integrity: An investigation of bird communities. Biological Conservation 126:499–511. Glennon, M.J., W.F. Porter, and C.L. Demers. 2002. An alternative technique for estimating diversity of small mammal populations. Journal of Mammalogy 83:734–742. Hansen, A.J., R. Rasker, B. Maxwell, J.J. Rotella, J.D. Johnson, A. Wright Parmenter, U. Langner, W.B. Cohen, R.L. Lawrence, and M.W. Kraska. 2002. Ecological causes and consequences of demographic change in the new west. Bioscience 52:151–162. Harmon, M.E., J.F. Franklin, F.J. Swanson, P. Sollins, S.V. Gregory, J.D. Lattin, N.H. Anderson, S.P. Cline, N.G. Aumen, J.R. Sedell, G.W. Leinkamper, K. Cromack, Jr., and K.W. Cummins. 1986. Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15:133–302. Harper, S.C., L.L. Falk, and E.W. Rankin. 1990. The Northern Forest Lands Study of New England and New York. United States Department of Agriculture, Forest Service, Rutland, VT. Hohmeyer, E. 1997. Sustaining economic development in the Adirondack Park. Adirondack Journal of Environmental Studies 4(2):26–29. Jenkins, J.C., and A.J. Keal. 2004. The Adirondack Atlas: A Geographic Portrait of the Adirondack Park. Syracuse University Press, Syracuse, NY. Kirkland, G., Jr. 1990. Patterns of initial small-mammal community change after clearcutting of temperate North American forests. Oikos:313–320. Korschgen, L.J. 1981. Foods of fox and grey squirrels in Missouri. Journal of Wildlife Management 45:260–266. MacArthur, R.H., and J.W. MacArthur. 1961. On bird species diversity. Ecology 42:594–598. Maestas, J.D., R.L. Knight, and W.C. Gilgert. 2001. Biodiversity and land-use change in the American mountain west. The Geographical Review 91(3):509–524. Magurran, A.E. 1988. Ecological Diversity and its Measurement. Princeton University, Princeton, NJ. McDonnell, J.A., D.L. Euler, and T.P. Clark. 1984. A test of predicted impacts of cottage development on small mammals. Wildlife Society Bulletin 12:156–161. McGee, G.G. 1998. Structural characteristics of Adirondack northern hardwood forests: Implications for ecosystem management. Ph.D. Dissertation. State University of New York, College of Environmental Science and Forestry, Syracuse, NY. McGee, G.G., D.J. Leopold, and R.D. Nyland. 1999. Structural characteristics of old-growth, maturing, and partially cut northern hardwood forests. Ecological Applications 9:1316–1339. 2007 M.J. Glennon and W.F. Porter 341 McMartin, B. 1994. The Great Forest of the Adirondacks. North Country Books, Utica, NY. Miller, R. 1985. Multiple comparisons. Pp. 575–591, In S. Kotz, N.L. Johnson, and C.B. Read (Eds.). Encyclopedia of Statistical Sciences, Volume 5. John Wiley and Sons, New York, NY. Miller, P.N. 1990. An analysis of the use and ownership of private and non-Forest Preserve public land. Pp. 274–295, In P. Bauer and G. Cox (Eds.). The Adirondack Park in the 21st Century, Volume 2, State of New York, Albany, NY. Miller, D.H., and L.L. Getz. 1977. Factors influencing local distribution and species diversity of forest small mammals in New England. Canadian Journal of Zoology 55:806–814. Miller, J.R., J.A. Wiens, N.T. Hobbs, and D.M. Theobald. 2003. Effects of human settlement on bird communities in lowland riparian areas of Colorado (USA). Ecological Applications 13:1041–1059. Nelson,L., Jr., and F.W. Clark. 1973. Corrections for sprung traps in catch/effort calculations of trapping results. Journal of Mammalogy 54:295–298. North, M., J. Trappe, and J. Franklin. 1997. Standing crop and animal consumption of fungal sporocarps in Pacific Northwest forests. Ecology 78:1543–1554. Odell, E.A., and R.L. Knight. 2001. Songbird and medium-sized mammal communities associated with exurban development in Pitkin County, Colorado. Conservation Biology 15:1143–1150. Palmer, W. 1993. Putting things in even better order: The advantages of canonical correspondence analysis. Ecology 74:2215–2230. Racey, G.D., and D.L. Euler. 1982. Small-mammal and habitat response to shoreline cottage development in central Ontario. Canadian Journal of Zoology 60:865–880. Raizer, J., and M.E. Amaral. 2001. Does the structural complexity of aquatic macrophytes explain the diversity of associated spider assemblages? Journal of Arachnology 29:227–237. Saunders, D.A. 1988. Adirondack Mammals. State University of New York, College of Environmental Science and Forestry, Syracuse, NY. Scott, V.E., G.L. Crouch, and J.A. Whelan. 1982. Responses of birds and small mammals to clearcutting in a subalpine forest in central Colorado. United States Department of Agriculture Forest Service, Research Note RM-422, Fort Collins, CO. Steele, B.B., R.L. Bayn, Jr., and C. Val Grant. 1984. Environmental monitoring using populations of birds and small mammals: Analyses of sampling effort. Biological Conservation 30:157–172. Sullivan, T.P., D.S. Sullivan, and P.M.F. Lindgren. 2000. Small mammals and stand structure in young pine, seed-tree, and old-growth forest, southwest Canada. Ecological Applications 10:1367–1383. Swan, D., B. Freedman, and T. Dilworth. 1984. Effects of various hardwood forest management practices on small mammals in central Nova Scotia. Canadian Field Naturalist 98:362–364. Tallmon, D., and L.S. Mills. 1994. Use of logs within home ranges of California redbacked voles on a remnant forest. Journal of Mammalogy 75:97–101. Ter Braak, C.J.F. 1986. Canonical correspondence analysis: A new eigenvector technique for multivariate direct gradient analysis. Ecology 67:1167–1179. 342 Northeastern Naturalist Vol. 14, No. 3 Theobald, D.M., J.R. Miller, and N.T. Hobbs. 1997. Estimating the cumulative effects of development on wildlife habitat. Landscape and Urban Planning 39:25–36. Thioulouse, J., D. Chessel, S. Dolédec, and J.M. Olivier. 1997. ADE-4: A multivariate analysis and graphical display software. Statistics and Computing 7:75–83. Thorndike, E. 1997. The Adirondack Park in the 21st Century. Part I: Proposals for a long-range policy and research agenda. Adirondack Journal of Environmental Studies 4(2):14–20. Von Trebra, C., D.P. Lavender, and T.P. Sullivan. 1998. Relations of small-mammal populations to even-aged shelterwood systems in sub-boreal spruce forest. Journal of Wildlife Management 62:630–642. Waters, J.R., and C.J. Zabel. 1995. Northern flying squirrel densities in fir forests of northeastern California. Journal of Wildlife Management 59:858–866. Witt, J.W. 1992. Home ranges and density estimates for northern flying squirrels, Glaucomys sabrinus, in western Oregon. Journal of Mammalogy 73:921–929. Yahner, R.H. 1986. Microhabitat use by small mammals in even-aged forest stands. American Midland Naturalist 115:174–180.