2012 NORTHEASTERN NATURALIST 19(3):391–410 Potential Effects of Beech Bark Disease on Small Mammals and Invertebrates in Northeastern US Forests Danielle E. Garneau1,*, Meghan E. Lawler2, Andrew S. Rumpf 2, Emily S. Weyburne2, Thomas M. Cuppernull2, and Adam G. Boe3 Abstract - The forests of the northeastern United States have become less contiguous and vigorous over the last century due to threats including acid rain, ice storm damage, and forest diseases. Often, trees have become the targets of widescale disease and pest outbreaks. Beech bark disease has been successful because of the effectiveness of the scale insect Cryptococcus fagisuga and the opportunistic Nectria coccinea var. faginata fungal vector. Since the severity of beech bark disease negatively affects mast production and canopy turnover, the abundance of small-mammal and insect populations can be limited. We explored the effects of beech bark disease, as well as other abiotic factors, on the diversity of small-mammal and invertebrate populations. We expected that biodiversity would vary according to disease severity in Fagus grandifolia (American Beech) stands, such that higher biodiversity and more seed predators would be noted in healthier forests. At sites in New York and Vermont, Sherman and pitfall traps were used to capture mammals and invertebrates, respectively. Correlations between tree size and disease severity levels were quantifi ed by noting the diameter at breast height (dbh) and by ranking according to disease intensity levels. Although biodiversity indices were not signifi cantly different among sites, there were signifi cant differences in dbh (F = 3.48, P = 0.0154, d.f. = 3) and disease intensity levels (F = 21.13, P < 0.0001, d.f. = 3). Surprisingly in 2008, beechnut seed production was greatest in the Catskill Mountains of New York, the site with the greatest disease manifestation. Mammal richness was highest at the Champlain Valley site where there were fewer Napaeozapus insignius (Woodland Jumping Mice). Patterns of small-mammal abundance at the stand level, elucidated in canonical correspondence analyses, were explained in part by land-use history, soil characteristics, elevation, recent cutting, temperature, and precipitation in 2007. Invertebrate family richness was greatest in the Adirondacks of New York as compared to other sites. At the site level, beechnut density, land-use history, and soil order were the most important variables explaining variation in invertebrate assemblages. Results from this study show that patterns of biodiversity cannot be directly explained by disease and beech mast alone in the short-term. Rather, multi-year community dynamics must be measured. Introduction Forests in the northeastern United States have grown less contiguous and vigorous over the last century as a result of threats such as acid deposition, ice and windstorm damage, and pathogens (Jenkins et al. 1999, Lewis et al. 2008, Takahashi and Lechowicz 2007). Additionally, widescale habitat fragmentation has led to a decline in biodiversity that can negatively affect vital ecosystem ser- 1Center for Earth and Environmental Science, State University of New York Plattsburgh, Plattsburgh, NY 12901. 2Department of Biology, Colgate University, Hamilton, NY 13346. 3Department of Biology, Colby College, Waterville, ME 04901. *Corresponding author - dgarn001@plattsburgh.edu. 392 Northeastern Naturalist Vol. 19, No. 3 vices such as moderating climate change, providing for an abundance of potable water, cycling of nutrients, and creating a resource base for higher trophic levels (Balvanera et al. 2006). In communities, foundation species (e.g., masting trees) support numerous wildlife taxa (Ellison et al. 2005). Masting trees are a predictable, regionally synchronous source of protein-rich food for many forest-dwelling species. Researchers have noted that the abundance of Peromyscus leucopus Rafi nesque (White-Footed Mice) peaks following high previous fall masting events (Schmidt and Ostfeld 2008, Wang et al. 2009, Wolff 1996). Causes of fluctuations in masting are numerous and have often been associated with both biotic and abiotic factors. Biotic factors hindering masting are often disease-related, but the proposed ultimate cause is considered to be climatic factors such as temperature (Clotfelter et al. 2007, Wang et al. 2009), precipitation (Clotfelder et al. 2007, Holmsgaard and Olsen 1966, Jensen 1982), winter severity (Wang et al. 2009), and late spring frosts of the previous year (Clotfelter et al. 2007, Jensen 1982, Kelly et al. 2008). Sork et al. (1993) surveyed three Quercus (oak) species and determined that not all experienced high seed rain during the same year. This asynchronicity suggests that climate is not the only factor contributing to the masting phenomenon. In particular, rodent communities in more northerly regions are expected to respond more rapidly to masting fluctuations, as the diversity of their food base is lesser and the need to store Quercus rubra L. (Red Oak) acorns, which require dormancy and are ideal for caching, is essential (Wang et al. 2009). Increased generalist rodent densities have been positively correlated with high masting events, in part resulting from the pulses of their invertebrate prey which also consume mast (Ostfeld et al. 1996, Schmidt and Ostfeld 2008, Wolff 1996). Kelly et al. (2008) noted that during the spring of an abundant mast year, seeds are not the only predictor of increased rodent abundance. In particular, flowers are abscised and consumed by invertebrates, thereby affording increased growth and fecundity which compensate for the irregularity of beechnut crops. In addition to insectivores, generalist seed predators (e.g., mice and squirrels) also experience reduced survival and fecundity as a result of the decline and functional loss of Fagus grandifolia Ehrh. (American Beech) (Storer et al. 2004). Wolff (1996) suggests that community dynamics are influenced by pulses in masting, such that the decline in seed resources, as well as the lagged pulse of predators, result in overall density reduction of rodents. Since the severity of beech bark disease negatively affects mast production, small-mammal caches may be limited, potentially reducing overwintering survival, reproductive vigor, and lifespan (Suchomel and Heroldova 2008). A depletion of any of these beech-derived resources would put a serious strain on the health of many small-mammal and invertebrate populations. Along with other notable masting forest dominants, such as Oak and Acer spp. (maples), American Beech has been targeted by pests and pathogens. About the time of maturation, American Beech trees are pierced by Cryptococcus fagisuga Lind., a phloem-feeding scale insect, thereby facilitating a Nectria coccinea var. faginata Lohman, Watson, and Ayers fungal infection (Griffin et al. 2003). Over time, the bark of the infected American Beech trees cracks, 2012 D.E. Garneau et al. 393 cankers form, and growth of small branches and foliage decrease, all of which results in increased tree mortality. The decline of this important northern hardwood species has initiated a shift in forest composition favoring maples and other northern hardwoods. At many sites, the once prominent American Beech overstory has been reduced to a shrub layer in the form of a dense understory thicket (Griffin et al. 2003). Similar shifts in forest community dynamics have also occurred as a result of Cryphonectria parasitica Murrill (Chestnut Blight). Specifically, the functional role of Castanea dentata Marsh. Borkh. (American Chestnut), a predictable yearly source of seed, was replaced by oaks, which experience a more pulsed life history. As a result of reduced reliability of mast and alteration of forest structure, higher trophic levels in the forest community were altered (Diamond and Giles 2000). Land-use history (Hane 2005), stand age (McNulty and Masters 2005), topographical variables (McShea 2000), and soils can affect the distribution, structure, and abundance of forest dominants, which in turn influences higher trophic levels (Le Blanc et al. 2010, Rodrigo et al. 2008). Thinning of stands to enhance marketable species, such as maples, serves to reduce the prevalence of beech bark disease by limiting the abundance of infected trees (Hane 2005). Additionally, shifts in leaf-litter quality, increases in coarse woody debris (CWD), and enhanced light levels and soil temperatures resulting from canopy gaps attract differing assemblages of phytophagous insects and higher-order organisms (Storer et al. 2004). Smock and McGregor (1988) postulate that loss of American Chestnut triggered an opportunity for Tsuga canadensis Engelm. (Eastern Hemlock) recruitment, which led to declines in invertebrate functional responses and a resultant reduction in growth rates and overall size. American Beech leaves are considerably more lignifi ed and slower to decompose than those of maples (Lovett et al. 2005). Koivula et al. (1999) determined that variable leaf-litter quality affected the distribution of Carabidae (ground beetles) in a boreal forest. Petrillo and Witter (2004) noted that increases in downed woody debris, resulting from disease-related beech mortality, increased arthropod abundance on the forest floor. Furthermore, a decline in leaf-litter quality affects numerous trophic levels such as phytophagous insects (e.g., Hemiptera and Hymenoptera) and Myodes gapperi Vigors (Red-backed Vole) because of the added diffi culty of meeting energetic demands (Bocock 1964, Kaminski et al. 2007). Researchers have noted that shifts in canopy dominance in diseased forest stands have direct implications for nutrient cycling. In the case of Eastern Hemlock, management practices (e.g., timber harvest) to mitigate Adelges tsugae Annand (Hemlock Woolly Adelgid) damage, has resulted in both increased nitrifi cation rates and negatively impacted riparian systems (Ellison et al. 2005). This canopy turnover could ultimately influence insectivores and other higher-order predators in terrestrial as well as aquatic systems. As forest communities respond to the increase in introduced pests, abiotic factors alter canopy and understory composition in aftermath forests. Climatic shifts, resulting in higher temperatures, have resulted in the loss of many canopy species, as forest pests expand their range. Low summer precipitation and April 394 Northeastern Naturalist Vol. 19, No. 3 frosts have been associated with declines in seed set in American Beech trees (Holmsgaard and Olsen 1966). Kasson and Livingston (2011) noted that mild winters and dry summers from 2000–2004 resulted in conditions that favored spread of beech bark disease agents in Maine. With the decline of American Chestnut and Eastern Hemlock, Betula lenta L. (Black Birch) has opportunistically fi lled their niche, which may have higher-trophic-level consequences (Stadler et al. 2005, USDA Forest Service 2004a). In infested forest stands, increases in soil cations have been noted, as throughfall delivers these nutrients to the soil (Stadler et al. 2005). As a result, these same researchers suggest that soil respiration rates increased and microbes responded with increased rates of decomposition, leading to bottom-up trophic responses. The longer the duration of infestation, the more inherent will be these nutrient cycling processes, which likely influence small-mammal and invertebrate communities. Disturbances, either biotic (e.g., forest disease or pests) or anthropogenic (e.g., forest management practices), have been found to alter small-mammal communities. Increasing levels of disturbance, from selective cutting to clearcutting, resulted in increased weights among Blarina brevicauda Say (Northern Short-tailed Shrew), Peromyscus maniculatus Wagner (Deer Mice), and White-footed Mice (Kaminski et al. 2007). Additionally, these same researchers noted that intensive harvest increased trapping success of Red-backed Vole and Napeozapus insignius Miller (Woodland Jumping Mice) and that declines in leaf litter positively affected the majority of captured small mammals. Forest structural changes, which ensue in disease-aftermath forests, result in complex trophic dynamics in the newly formed landscape mosaic. In this study, we investigated the potential effects of beech bark disease and other abiotic factors on the diversity of small mammals and invertebrates in the northern hardwood forest. We expected that biodiversity trends would vary among mast- and insect-eating small mammals and phytophagous and detritivorous insects according to the severity of disease in American Beech co-dominant stands. We predicted that higher diversity levels would be observed in lessdiseased forests. We also predicted that temperature and precipitation patterns from the prior year would have the potential to influence biotic factors such as mast crops, thereby affecting higher trophic levels. Older stands were anticipated to experience intense beech bark disease, and therefore reduced seed crops, increases in CWD, and alterations in leaf litter, resulting in a lowered abundance of seed predators. Small mammals, such as Woodland Jumping Mice and Glaucomys sabrinus Shaw (Northern Flying Squirrel), with alternative food sources such as endogone and glomalean fungi (Brannon 2005, Orrock et al. 2003, Whittaker 1963) and graminoids, were predicted to exhibit little change in abundance in relation to beech bark disease infestation levels. field-Site Description Comprehensive trapping was conducted from early June to July 2008 in two forest stands within each of four sites in New York and Vermont (fig. 1). Sites were located in A) the Catskill Mountains in Claryville, NY (41°58'2.676"N, 2012 D.E. Garneau et al. 395 74°30'48.5562"W), managed by the Frost Valley YMCA camp; B) the Adirondack Mountains (Huntington Forest) in Newcomb, NY (44°2'21.4074"N, 74°15'37.0728"W), managed by the SUNY College of Environmental Science and Forestry; C) the Champlain Valley in Altona, NY (44°50'1.6074"N, 73°33'8.7762"W), managed by the W.H. Miner Agricultural Research Institute; and D) the Green Mountains (Coolidge State Forest) (43°34'7.4634"N, 72°47'2.7672"W) in North Shrewsbury, VT. The Catskill forest site has been unmanaged since the early 1900s and has not been harvested (B. Snyder, Frost Valley YMCA, Claryville, NY, pers. comm.). In the Huntington Forest of the Adirondacks, a series of selective harvests (1950–52, 1971–74) and cutting for a right of way to a Odocoileus virginianus Zimmermann (White-Tailed Deer) exclosure (1961), in addition to herbicide treatments (Silvisar 51) in 1971, occurred in proximity to the area sampled (C. Demers, SUNY ESF Ecology Center, Newcomb, NY, pers. comm.). Non-commercial timber harvest occurred on the Champlain Valley site at least 15 years prior to this study (H. Boyce, Northwoods Forest Consultants, Jay, NY, pers. comm.). Lastly, Coolidge State Forest, in the Green Mountains, was logged for commercial and non-commercial timber harvest in the 1960s (L. Thornton, Vermont fish and Wildlife, Rutland, VT, pers. comm.). Dominant overstory tree species at the locations were a mixture of American Beech, Acer rubrum L. (Red Maple), Acer pennsylvanicum L. (Striped Maple), Acer saccharum Marsh. (Sugar Maple), Betula alleghaniensis Britton. (Yellow Birch), Eastern Hemlock, Populus tremuloides Michx. (Quaking Aspen), Pinus strobus L. (White Pine), and Quercus rubra L. (Red Oak). Understory species figure 1. Map of the northeastern United States and American Beech monitoring sites: A) Catskill, B) Adirondack, C) Champlain Valley, D) Green Mountain. 396 Northeastern Naturalist Vol. 19, No. 3 included American Beech and maple saplings, Mitchella repens L. (Partridgeberry), Viburnum lantanoides Michx. (Hobblebush), Gaultheria procumbens L. (Wintergreen), Dryopteris intermedia L. (Fancy Fern), lycopods, and graminoids. Authority for all plant species was derived from the USDA NRCS Database. Stands within these locations were selected based on American Beech tree co-dominance and abundance, in addition to road access. Authority for all mammal species was derived from Whitaker and Hamilton (1998). Methods Preliminary small-mammal feasibility surveys, at one-third the trapping effort, were conducted in August 2007 at the Adirondack and Catskill sites. During May–June 2008, forty-fi ve Sherman live traps (33 small [6 ¼ x 2 x 2 ½ in] and 12 large [10 x 3 x 3 in]), were placed at each stand, approximately 15 m apart, for a total of n = 90 traps per site. Conard et al. (2008) noted that at least three nights are needed to provide an adequate estimate of species richness within a site, when at minimum 9 and at maximum 144 traps per hectare are sampled. Traps were baited with rolled oats and crushed sunflower seeds at dusk for four consecutive nights (effort = 360 trap nights per site). Wilson and Mabry (2010) suggest that disinfecting Sherman traps to reduce odor, which has the potential to bias captures, does not influence capture success as compared to traps with residual odor. The traps were checked daily at dawn, animals processed, and released at point of capture. Animal processing entailed identifi cation to species level when possible, as well as assessment of sex, body weight (g), body length (cm), and tail length (cm). At each stand, two pitfall traps intended for capturing smaller vertebrates and invertebrates were recessed into the ground within 1–5 m of each Sherman trap, depending on ground conditions. The two types of traps were simultaneously employed to maximize sampled species and reduce bias for those with smaller bodies. Pitfall traps consisted of 150-ml plastic cups fi lled with a mixture of approximately 30 ml of 70 % isopropyl alcohol solution and dishwashing liquid to break the surface tension. Pitfall traps remained in the fi eld continuously for four nights, after which specimens were retrieved and stored in alcohol for processing. Specimens were later identifi ed to family level, according to fi eld guides (Arnett and Jacques 1981, Eaton and Kaufman 2007, Evans 2007, Milne and Milne 2000, White 1983). American Beech trees were quantifi ed by noting the diameter at breast height (dbh) of the nearest American Beech tree in each of four cardinal directions surrounding each Sherman trap. Only American Beech trees with a dbh greater than 2.5 cm were measured. Tree disease levels were ranked according to beech bark disease intensity levels using a previously described ranking scale (Griffi n et al. 2003): 1 = very little or no sign of either causal agent (Cryptococcus fagisuga or Nectria coccinea var. faginata); 2 = Cryptococcus fagisuga present, bark beginning to crack, tree still shows vigor, canopy at least 75% intact; 3 = bark heavily cracked, signifi cant cankering from Nectria coccinea var. faginata 2012 D.E. Garneau et al. 397 colonies, some crown damage or limb loss, canopy 25–75% intact; 4 = bark severely cracked, large girdling cankers, signifi cant crown loss or snag, canopy <25% intact; and 5 = tree dead because of beech bark disease. At each stand, ten 5-gal buckets were strung among three reproductive American Beech trees to estimate seed production in the fall (McNulty and Masters 2005). Mast collection from seed traps occurred at the sites in the Catskills and the Adirondacks, as a preliminary site survey in mid-November of 2007, and between 7 November–27 November 2008 at each site. Shannon-Wiener diversity indices H' = -Σpi * ln(pi), where pi is the decimal fraction of relative importance value of the ith species, evenness (J) = H'/H'max * 100, and species richness (S) = total number of species for small mammal assemblages were calculated for all sites (Margurran 2004). A repeated measures analysis of variance (ANOVA) was performed to determine if biodiversity varied according to site and disease state (SPSS, Inc. Chicago, IL). Data were transformed when normality was not satisfied (i.e., ln transformation of dbh). For each site, edaphic factors were extracted using GIS from the Soil Survey Geographic Database (SSURGO) and included soil order, subgroup, mineralogy, particle size, soil temperature regime, pH, and slope. Factors such as elevation (m), year of published beech bark disease infestation (Morin et al. 2007), land-use history, and most recent cutting (per site manager) were included in the analysis. Additionally, other abiotic factors such as mean temperature (°F) and mean precipitation (mm) for meteorological seasons (winter: Dec.–Feb., spring: March–May, summer: June–Aug., fall: Sept.–Nov.), and late spring frost were derived from regional weather station data available online (http://www. weatherunderground.com/; National Climate Data Center at http://www.ncdc. noaa.gov/cdo-web/). A principle components analysis (PCA) was performed on edaphic and abiotic factors to derive a smaller number of uncorrelated variables using PCOrd (McCune and Mefford 2011). The significant PCA axes were then used as explanatory variables describing edaphic and abiotic factors in further analyses. Canonical correspondence analysis (CCA) was run to determine whether differences in small-mammal assemblages at the stand level were explained by edaphic and other stand factors. The main matrix contained small-mammal relative abundance data, and the secondary matrix contained year of beech bark disease infestation, year of recent cut, and a PCA axis defined by elevation and winter/fall/spring precipitation from the previous year, as well as land-use history, soils, and temperature from the previous year (2007). Matrix relationships were tested using Monte Carlo randomization tests (200 permutations), and biplots were overlain for significant variables on the ordination graph by means of linear combination scores (Kovalenko et al. 2010). No predictor variables were inter-correlated. Invertebrate assemblage patterns were explored according to beechnut density, soil order, and land-use history at the stand-level using a CCA. 398 Northeastern Naturalist Vol. 19, No. 3 Results There were significant differences in beech bark disease intensity levels among trees among sites (fig. 2). Specifically, only ten percent of trees sampled at the Champlain Valley site expressed signs of being in the killing or aftermath phase (rank 3–5). Contrastingly, the highest disease infestation level was noted in the Catskill site (29%), followed by the Adirondack and Green Mountain sites (20%) (F = 21.13, P < 0.00001, d.f. = 3). Stand age varied by site, which is reflected in the size of the trees sampled. The dbh of American Beech trees was significantly different among the study sites (F = 3.48, P = 0.0154, d.f. = 3), with largest differences observed when comparing the Catskill to the Green Mountain site. There was a significant correlation between diameter at breast height and disease state (ϱ = 0.37452, P = < 0.0001), such that trees with an average dbh of 15.3 cm had the most intense disease phenotype. The trees with a disease state of 3, characterized by bark with intense cracking, cankering, and crown and limb damage, had the largest average dbh at 20.9 cm, while the trees with a disease level of 1, characterized by low to no sign of scale or fungus, had the smallest average dbh at 8.4 cm. figure 2. Relative abundance of American Beech trees in each disease intensity level at each site in 2008: A) Catskill, B) Adirondack, C) Champlain Valley, D) Green Mountain. Disease level is noted as 1 = very little or no sign of either causal agent (Cryptococcus or Nectria); 2 = Cryptococcus present, bark beginning to crack, tree still shows vigor, canopy at least 75% intact; 3 = bark heavily cracked, signifi cant cankering from Nectria colonies, some crown damage or limb loss, canopy 25–75% intact; 4 = bark severely cracked, large girdling cankers, signifi cant crown loss or snag, canopy <25% intact; and 5 = tree dead because of beech bark disease (Griffi n et al. 2003). 2012 D.E. Garneau et al. 399 Seed production per seed trap was not signifi cantly different among sites in the fall of 2008 (F = 2.06, P = 0.117, d.f. = 3), as most buckets yielded moderate mast abundance, and only a few collected in the Catskill site captured abundant beechnut yield. Upon standardizing for site area, the Catskill site beechnut production was 5.7, 4.1, and 2.2 times that of the Champlain Valley, Adirondack, and Green Mountain sites, respectively (fig. 3). These 2008 mast data were contrasted to the unproductive fall 2007, where 12.8 and 43.7 times fewer seeds were collected in seed traps (n = 20/site) in the Catskill and Adirondack sites, respectively (fig. 3). The highest small-mammal species richness and evenness in 2008, as estimated by the Shannon-Wiener index, occurred at the Champlain Valley site (Table 1). Small-mammal richness declined from the Catskill, to the Green Mountain, and the Adirondack sites, respectively (Table 1, fig. 4). In 2008, Woodland Jumping Mice represented the largest percentage of small mammals captured in all sites except for the Champlain Valley (fig. 4). The Green Mountain site contained the greatest abundance of insectivores, such as Northern Short-tailed Shrews and Sorex spp. (shrews), while the Adirondack site contained none. Deer Mice, ubiquitous seed predators, were not captured at the Catskill site in 2008 (fig. 4). These data can be contrasted to the preliminary trapping survey performed in figure 3. Beechnut production (seeds/ha) in 2007 at A) Catskill and B) Adirondack sites and at each site in 2008. 400 Northeastern Naturalist Vol. 19, No. 3 2007, where 70% (n = 55) and 75% (n = 68) of all captures at the Catskill and Adirondack sites were Deer Mice (fig. 4). Both biotic and abiotic variables were correlated with the structure of smallmammal and invertebrate communities. Approximately 58% of cumulative variance in our ordination of small-mammal species data was explained by standspecifi c land-use history, soil factors, elevation, winter and fall precipitation in 2007, and winter, fall, and spring mean temperatures in 2007 (Monte-Carlo P = 0.1800; fig. 5). Less impacted historical land use occurred in sites where soil type consisted of inceptisols with mixed mineralogy and mesic temperature regimes. Likewise at higher elevations, mammals experienced lesser precipitation figure 4. Relative abundance of small mammals captured at all American Beech trapping sites, excluding all recaptured individuals. Trapping occurred June–July 2008 at four-day intervals per site. A) Catskill (n2007 = 55, n 2008 = 16), B) Adirondack (n 2007 = 68, n 2008 = 44), C) Champlain Valley (n = 32), D) Green Mountain (n = 61). Table 1. Shannon-Wiener diversity indices (H'), richness, and evenness (J) for species of small mammals and families of invertebrates captured in June–July 2008 in four beech co-dominant sites. Biodiversity index Catskill Adirondack Champlain Valley Green Mountain H'mammals 1.332 0.680 1.800 1.319 Species richness 5 4 7 5 Jmammals 0.827 0.491 0.925 0.678 H'invertebrates 2.046 2.421 2.727 1.876 Family richness 30 46 36 37 Jinvertebrates 0.068 0.053 0.076 0.051 2012 D.E. Garneau et al. 401 in 2007. Additionally, approximately 85% of cumulative variance in our sitespecifi c ordination of invertebrate family data was explained by land-use history, beechnuts, and soil order (fig. 6). In particular, higher beechnut densities were noted in the Catskill site where Carabidae abundance was high, as compared to the greater importance of land-use history in the Adirondacks and soil order in the Green Mountains and Champlain Valley. Invertebrate family richness declined respectively from the Champlain Valley, to the Adirondack, Catskill, and the Green Mountain sites, according to Shannon- Wiener diversity indices (Table 1). Invertebrate family evenness values decreased from the Champlain Valley, to the Catskill, Adirondack, and the Green Mountain sites, respectively (Table 1). Among all sites, the most abundant invertebrate taxa were Carabidae (ground beetle), Pulmonata (snail and slug), Staphylinidae (rove beetle), Agelenidae (funnel-web spider), and Gryllidae (cricket). The Champlain Valley site possessed the fewest representative Carabids, but did contain other families such as Simuliidae (black fly), Geotrupidae (earth-boring dung beetle), Chrysomelidae (leaf beetle), and Dolichopodidae (long-legged fly) that were much less abundant at other sites. In terms of abundance, invertebrates were most frequently captured at the Green Mountain site. Discussion Numerous factors have been known to influence the distribution and abundance of small mammals and invertebrates in northern forests. In particular, figure 5. Stand-specifi c canonical correspondence analysis (CCA) ordination of smallmammal assemblages. 402 Northeastern Naturalist Vol. 19, No. 3 biotic factors such as mast abundance and forest disease might inhibit the survival and reproduction of small mammals and ground-dwelling invertebrates. Our research, as well as those of others, reveals that American Beech, at least in the Adirondack Park (McNulty and Masters 2005), experience a bumper crop every other year. Based on preliminary surveys of the Catskill and Adirondack forest sites in 2007, we expected a lower capture rate of Peromyscus spp. the following year (Jensen et al. 2012). Low Peromyscus spp. yields might result from less available mast for these seed predators, as was noted in several northeastern US locations in the fall 2007 (Jensen et al. 2012). Summer 2008 captures revealed a higher percentage of Woodland Jumping Mice as compared to Peromyscus spp. in all sites surveyed, except for the Champlain Valley. We did not anticipate the figure 6. Site-specifi c invertebrate canonical correspondence analysis (CCA) ordination (10 most abundant taxa) according to soil order, beechnuts, and land-use history. 2012 D.E. Garneau et al. 403 high capture rate of Woodland Jumping Mice, nor did we expect the number of captures to exceed the number of captured Peromyscus spp. Kirkland and Griffi n (1974) observed that Woodland Jumping Mice and Peromyscus spp. appear to be complementary in population abundance trends, which might explain the low numbers in alternating mast years. Brower and Cade (1966) explained this complementary relationship as a function of population size and density of ground cover, which might account for the trends in abundance from 2007–2008 at the Catskill and Adirondack sites. It is possible that changes in forest structure to favor thickets of American Beech might promote this fi nding. Seed rain was also an important factor explaining variance of invertebrate communities at the Catskill site, where Carabidae, Agelenidae, and Calliphoridae (Blow Fly) responded to higher levels of American Beech mast. Kelly et al. (2008) noted that following a high mast year, greater abundances of senesced flowers in leaf litter attracts predatory spiders, potentially explaining the increase of Agelenidae at the Catskill site. These same researchers noted an increase in rodents and caterpillars during this part of the American Beech cycle, which might explain the increases in fecal/carrion-feeding species of the family Calliphoridae and larvae/pupaefeeding Carabidae, respectively. Forest disease can influence the northern forest community as has been noted in many historic cases where loss of a foundation tree species (e.g., Ulmus americana L. [American Elm], American Chestnut, and Eastern Hemlock) has left behind wildlife assemblages that reflect adaptation to a different biotic and edaphic regime (Ellison et al. 2005). In particular, studies have noted that time since disease infestation, stand age, and prior land-use history can shape the response of the forest community post-infestation (Feldhammer 1979, Hane 2005, McNulty and Masters 2005, McShea 2000). Morin et al. (2007) noted that beech bark disease arrived in the Catskill and Green Mountain site in the 1960s, while the Adirondack and Champlain Valley were spared for another 15 years, perhaps a result of their topography. Recent cutting (i.e., surrogate for stand age) did appear to explain the trends in mammals but not invertebrates; however, average dbh and beech bark disease intensity level varied significantly among sites in New York and Vermont (USDA Forest Service 2004b). Because the moderate- and larger-size trees typically produce the most robust seed mast, and are the trees targeted by disease, the food supply for beechnut consumers was affected. Surveys of tree size at all sites yielded another unexpected result: American Beech trees with the largest dbh did not exhibit the most intense disease phenotype as predicted. Only the Catskill site displayed a positive trend for enhanced disease in larger trees, which might be explained because this site was continually forested and had not experienced the cutting regime as at other sites. Early beech bark disease research suggested that this disease typically strikes trees in more mature cohorts; however, more recent research suggests that this correlation cannot be made at the stand level (Griffin et al. 2003), as was similarly observed in this study. McNulty and Masters (2005) noted that from 1989–2002, the majority of mature American Beech trees >38 cm dbh in the central Adirondacks were no longer alive and that 404 Northeastern Naturalist Vol. 19, No. 3 the majority of smaller trees showed signs of disease. In their study, although beech bark disease infection was greatest in larger trees, beechnut production continued to rise. This observation suggests the potential that seeds recruited into a population with disease might be better able to withstand the pest and fungal attack. Additional explanations might be that American Beech thickets have been known to increase the abundance of nut-producing tissues, thereby increasing nut yield in a diseased forest (Lucas et al. 2005, McNulty and Masters 2005) or that stressed trees increase their reproductive investment (Hagen et al. 2003). The Catskill site had the greatest yield of mast in 2008 and contained the highest density of trees in the aftermath disease stage (29%), which indicates that this site might be exhibiting enhanced mast production as a regenerating thicket and its inhabitants might be better maintained than at other sites. Thicket structure has been found to create unique microhabitat topography which provides escape cover, food, and burrow or nest structures (Iiada 2006).We predicted capture rates of beechnut seed predators to be highest at this site; however, we found a predominance of Woodland Jumping Mice (i.e., fungal and seed consumers) and Northern Short-tailed Shrews (i.e., insectivores) along with fewer members of the seed predator guild. Of additional importance, land-use history was a strong predictor of community composition at these sites. Orwig and Abrams (1994) noted that oak-pine forest communities were linked closely to post-settlement land-use histories (e.g., logging, farming, charcoal and iron industry demands, fire). In particular, land-use history explained the majority of the variance of mammal species and invertebrate families at the Adirondack site. The Huntington forest site in the Adirondacks was not only the target of a selective cut on several occasions, but also was treated with pesticides, resulting in a more disturbed habitat than the other sites surveyed. The Adirondack site also possessed the lowest species richness of mammals and highest family richness of invertebrates. In the Adirondacks, higher densities of Woodland Jumping Mice, a species known to eat fungus, might be a response to the presence of fungalphilic invertebrates such as crickets and Opiliones (harvestmen). Given that leaf fall is the source of nutrients for a variety of phytophagous insects, changes in invertebrate composition are a good indicator of flux in overstory forest composition, as in the case of disease. All sites were dominated by Carabidae, except for the Champlain Valley. Carabidae are easy to catch in pitfall traps, and sampling bias may have limited the capture of other species (Seldon and Beggs 2010). The relatively minimal catch of Carabidae in the Champlain Valley site may be the result of a logging history within the W.H. Miner Experimental Forest. Carabidae prefer more complexly managed forest microhabitats (Werner and Raffa 2000); the Champlain Valley site has historically experienced timber harvest and the forest is situated in an agricultural region, resulting in heterogeneous structural complexity. The prevalence of slugs in the Green Mountain site might result from the dense, mature understory vegetation. Much research has centered on the role of edaphic and site factors in influencing forest community assemblages (Feldhammer 1979, Nowacki et al. 1990). 2012 D.E. Garneau et al. 405 Soil order, pH, slope, aspect, and climatic factors are often important in plant establishment and growth patterns that direct the ecology of phytophagous and predatory species of mammals and invertebrates. Although abiotic factors (e.g., temperature, precipitation, elevation, and soil characteristics, date of beech bark disease arrival, recent cutting, and land-use history) helped explain wildlife patterns at our sites, the relationship between these variables and the communities was not statistically signifi cant. Warmer seasonal temperatures experienced at the Castkill site perhaps contributed to the highest mast yield of any site. Higher elevation and lower fall and winter precipitation in 2007 in the Adirondack and Catskill sites were important predictors for small-mammal assemblages. Wetter soils increase fungal abundance and might explain the presence of Northern Flying Squirrels in the Champlain Valley alone. Additionally, seed predators such as Deer Mice and Tamias striatus L. (Eastern Chipmunk) may have responded to increases in seed production following a rainy year, thus contributing to the highest small-mammal richness realized at this site. Additionally, the lag in onset of the disease at this site might have facilitated increased richness in the mammal assemblage. In the future, trapping during periods of high mast, in addition to non-mast years, should increase small-mammal and invertebrate abundance and allow for greater strength when comparing communities using multivariate statistical approaches. Soil order, which was predictive of invertebrate species, separated the Catskill site from all others. Soils of the Catskills are less developed inceptisols, as compared to the spodosols in the Adirondack, Champlain Valley, and Green Mountains. Additionally, the Catskill site experienced a late spring frost on Julian day 150, as compared to day 142 at all other sites. Late frost can affect plant phenology and overwintering success of small mammals (Goodrum et al. 1971, Sork et al. 1993), perhaps explaining our lowest mammal abundances in the Catskills. Shrew species, in particular Northern Short-tailed Shrew are associated with the previous year’s mean winter precipitation, especially at the Catskill site. Both Sorex spp. and Northern Short-tailed Shrew had dramatically different abundances across the four sites. There were no shrews captured in 2008 at the Adirondack site, as compared to their comprising 35% of the yield in the Catskill site (30% Sorex spp., 5% Northern Short-tailed Shrew). Contrastingly, in 2008 at the Green Mountain site, 10% of the insectivore yield was Sorex spp. as compared to 30% Northern Shorttailed Shrew. The Champlain Valley site had a Northern Short-tailed Shrew capture rate of 30%, but lacked Sorex spp. A study performed in a south-central Pennsylvanian forest found a relatively high capture of Soricids in comparison to other small mammals (McCay and Storm 1997), providing support that our disease-disturbed forests exhibit similar trends to those that are anthropogenically disturbed (e.g., logging). We expected stands more severely damaged by beech bark disease to possess the greatest abundance of invertebrates and insectivores; however, the Green Mountain site, with the highest invertebrate yield, resulted in moderate insectivore captures. Additionally, the Adirondack site, which possessed 1.8 times fewer invertebrates, produced no insectivore captures. Pitfall traps were successful in capturing Carabids and Pulmonata at these sites. 406 Northeastern Naturalist Vol. 19, No. 3 In conclusion, results from this study demonstrate that linking small-mammal and invertebrate biodiversity patterns to regional forests experiencing beech bark disease can be diffi cult. Patterns of forest structure and composition, mast, forest-floor invertebrates, and higher-order predators may be linked in both direct and indirect pathways, making it diffi cult to predict wildlife patterns from one year to the next (Lucas et al. 2005). Mast was not found to be the most signifi cant predictor of wildlife richness patterns at beech bark disease-infested sites, rather date of infestation, as well as abiotic (e.g., previous-year precipitation and temperature, land-use history, recent cutting regimes) and edaphic (e.g., soils) factors were important. This study provides a baseline inventory of wildlife species inhabiting forests in the aftermath phase of beech bark disease. Dynamics of forest structure, mast, seed predators, and prey are influenced by year and season, as well as abiotic and other biotic conditions, which call for multi-year surveys to reveal correlations between forest pathogens and wildlife patterns. Acknowledgments Many thanks to researchers who granted permissions and support with fi eld site history and logistics, specifi cally Benjamin Snyder, Director of Natural Resources and Environmental Science at the Frost Valley YMCA, Stacy McNulty and Charlotte Demers of the Adirondack Ecology Center SUNY Environmental Science and Forestry (ESF), Herbert Boyce of Northwoods Forest Consultants, both Kirk Beattie and Stephen Kramer of the W.H. Miner Agricultural Research Institute, as well as Lisa Thornton, the stewardship forester at Vermont fish and Wildlife. Many thanks to Rachel Schultz and Timothy Mihuc (SUNY Plattsburgh), as well as Kristine Hopfensperger (Northern Kentucky University) for their assistance with statistical analysis and Erin Bradshaw Settevendemio (SUNY Plattsburgh/University of Florida) for assistance in seed-trap collection. Additional gratitude to Robert Fuller (SUNY Plattsburgh) for his tutelage in soil terminology, Janet Mihuc (Paul Smith’s College) for invertebrate consultation, and Drs. Kenneth Adams and Christopher Martine (SUNY Plattsburgh/Bucknell University) for manuscript review. We wish to thank the Colgate University research council for discretionary funds that supported this project. finally, we wish to thank the editor and 2 anonymous reviewers for valuable constructive criticism, which resulted in a thorough revaluation of our data and fi ndings. Literature Cited Arnett, R.H., Jr., and R.L. Jacques, Jr. 1981. Simon and Schuster's Guide to Insects. fireside New York, NY. 512 pp. Balvanera, P., A.B. Pfi sterer, N. Buchmann, J.S. He, T. Nakashizuka, D. Raffaelli, and B. Schmid. 2006. Quantifying the evidence for biodiversity effects on ecosystem functioning and services. Ecology Letters 9:1146–1156. Bocock, K.L. 1964. Changes in the amounts of dry matter, nitrogen, carbon, and energy in decomposing woodland leaf litter in relation to the activities of the soil fauna. Journal of Ecology 52(2):273–284. Brannon, M.P. 2005. Distribution and mictrohabitat of Woodland Jumping Mouse, Napaeozapus insignius, and the White-footed Mouse, Peromyscus leucopus, in the southern Appalachians. Southeastern Naturalist 4(3):479–486. 2012 D.E. Garneau et al. 407 Brower, J.E., and T.J. Cade. 1966. Ecology and physiology of Napaeozapus insignis (Miller) and other woodland mice. Ecology 47:46–63. Clotfelter, E.D., A.B. Pedersen, J.A. Cranford, N. Ram, E.A. Snajdr, V. Nolan, Jr., and E.D. Ketterson. 2007. Acorn mast drives long-term dynamics of rodent and songbird populations. Oecologia 154(3):493–503. Conard, J.M., J.A. Baumgardt, P.S. Gipson, and D.P. Althoff. 2008. The influence of trap density and sampling duration on the detection of small-mammal species richness. Acta Theriologica 53(2):143–156. Diamond, S.J., and R.H. Giles. 2000. Hard-mast production before and after the chestnut blight. Southern Journal of Applied Forestry 24:196–201. Eaton, E.R., and K. Kaufman. 2007. Kaufman field Guide to Insects of North America. Houghton Mifflin Harcourt. New York, NY. 392 pp. Ellison, A.M., M.S. Bank, B.D. Clinton, E.A. Colburn, K.A. Elliot, C.R. Ford, D.R. Foster, B.D. Kloeppel, J.D. Knoepp, G.M. Lovett, J. Mohan, D.A. Orwig, N.L. Rodenhouse, W.V. Sobczak, K.A. Stinson, J.K. Stone, C.M. Swan, J. Thompson, B. Von Holle, and J.R. Webster. 2005. Loss of foundation species: Consequences for the structure and dynamics of forested ecosystems. Frontiers in Ecology and the Environment 3(9):479–486. Evans, A.V. 2007. National Wildlife Federation field Guide to Insects and Spiders of North America. Sterling Publishing Co., New York, NY. 496 pp. Feldhammer, G.A. 1979. Vegetative and edaphic factors affecting abundance and distribution of small mammals in southeast Oregon. Western North American Naturalist 39(3):207–218. Goodrum, P.D., V.G. Reid, and C.E. Boyd. 1971. Acorn yield characteristics, and management criteria of oaks for wildlife. Journal of Wildlife Management 35:520–532. Griffi n, J.M., G.M. Lovett, M.A. Arthur, and K.C. Weathers. 2003. The distribution and severity of beech bark disease in the Catskill Mountains, NY. Canadian Journal of Forest Research 33:1753–1760. Hagen, S.B., I. Folstad, and S.W. Jakobsen. 2003. Autumn colouration and herbivore resistance in Mountain Birch (Betula pubescens). Ecology Letters 6:807–811. Hane, E. 2005. The effects of land-use history on beech bark disease severity. Pp. 138–141, In C.A. Evans, J.A. Lucas, and M.J. Twery (Eds.). Beech Bark Disease: Proceedings of the Beech Bark Disease Symposium. General Technical Report NE- 331. USDA Forest Service, Northern Research Station, Newtown Square, PA. 149 pp. Holmsgaard. E., and H.C. Olsen. 1966. Experimental induction of flowering in beech. Der. Forstlige Forsogsvaesen i Danmark 30(1):1–18. Iiada, S. 2006. Dispersal patterns of Quercus serrata acorns by wood mice in and around canopy gaps in a temperate forest. Forest Ecology and Management 227:71–78. Jenkins, J.C., J.D. Aber, and C.D. Canham. 1999. Hemlock Woolly Adelgid impacts on community structure and N cycling rates in Eastern Hemlock forests. Canadian Journal of Forest Research 29:630–645. Jensen, P.G., C. Demers, S. McNulty, W. Jakubas, and M.M. Humphries. 2012. Marten and fisher responses to fluctuations in prey populations and mast crops in northern hardwood forest. Journal of Wildlife Management 76(3):489–502. Jensen, T.S. 1982. Seed production and outbreaks of non-cyclic rodent populations in deciduous forests. Oecologia 54(2):184–192. Kaminski, J.R., M.L. Davis, M. Kelly, and P.D. Keyser. 2007. Disturbance effects on small-mammal species in a managed Appalachian forest. American Midland Naturalist 157:385–397. 408 Northeastern Naturalist Vol. 19, No. 3 Kasson, M.T., and W.H. Livingston. 2011. Relationships among beech bark disease, climate, radial growth response, and mortality of American Beech in northern Maine, USA. Forest Pathology (early view). 14 pp. Available online at DOI: 10.1111/j.1439- 0329.2011.00742.x. Kelly, D., W.D. Koenig, and A.M. Liebold. 2008. An intercontinental comparison of the dynamic behavior of mast-seeding communities. Population Ecology 50:329–342. Kirkland, G.L., Jr., and R.J. Griffi n. 1974. Microdistribution of small mammals at the coniferous-decidious forest ecotone in northern New York. Journal of Mammalogy 55(2):417–427. Koivula, M., P. Puntilla, Y. Haila, and J. Niemela. 1999. Leaf litter and the small-scale distribution of carabid beetles (Coleoptera, Carabidae) in the boreal forest. Ecography 22:424–435. Kovalenko, K.E., E.D. Dibble, and J.G. Slade. 2010. Community effects of invasive macrophyte control: Role of invasive plant abundance. Journal of Applied Ecology 47:318–328. Le Blanc, M.-L., D. Fortin, M. Darveau, and J.-C. Ruel. 2010. Short-term response of small mammals and forest birds to silviculture practices differing in tree retention in irregular boreal forests. Ecoscience 17(3):334–342. Lewis, J., J. Licitra, A.R. Tuininga, A. Sirulnik, D. Turner, and J. Johnson. 2008. Oak seedling growth and ectomycorrhizal colonization are less in Eastern Hemlock stands infested with Hemlock Woolly Adelgid than in adjacent oak stands. Tree Physiology 28:629–636. Lovett, G.M., M.A. Arthur, K.C. Weathers, and R.D. fitzhugh. 2005. Effects of beech bark disease on carbon and nitrogen cycling in the Catskill forests. P. 68, In C.A. Evans, J.A. Lucas, and M.J. Twery (Eds.). Beech Bark Disease: Proceedings of the Beech Bark Disease Symposium. General Technical Report NE-331. USDA Forest Service, Northern Research Station, Newtown Square, PA. 149 pp. Lucas, J.A., C.A. Evans, and M.J. Twery. 2005. The status of beech bark disease in northern hardwood forests. Adirondack Journal of Environmental Studies Spring/ Summer:9–12. Margurran, A.E. 2004. Measuring Biological Diversity. Blackwell Publishing Malden, MA. 260 pp. McCay, T., and G.L. Storm. 1997. Masked Shrew (Sorex cinereus) abundance, diet, and prey selection in an irrigated forest. American Midland Naturalist 138(2):268–275. McCune, B., and M.J. Mefford. 2011. PC-ORD. Multivariate Analysis of Ecological Data. Version 6. MjM Software, Gleneden Beach, OR. McNulty, S.A., and R.D. Masters. 2005. Changes to the Adirondack forest: Implications of beech bark disease on forest structure and seed production. Pp. 52–57, In C.A. Evans, J.A. Lucas, and M.J. Twery (Eds.). Beech Bark Disease: Proceedings of the Beech Bark Disease Symposium. General Technical Report NE-331. USDA Forest Service, Northern Research Station, Newtown Square, PA. 149 pp. McShea, W.J. 2000. The influence of acorn crops on annual variation in rodent and bird populations. Ecology 81(1):228–238. Milne, L., and M. Milne. 2000. National Audubon Society field Guide to Insects and Spiders of North America. Knopf, New York, NY. 992 pp. Morin, R.S., A.M. Liebhold, P.C. Tobin, K.W. Gottschalk, and E. Luzader. 2007. Spread of beech bark disease in the eastern United States and its relationship to regional forest composition. Canadian Journal of Forest Research 37:726–736. 2012 D.E. Garneau et al. 409 Nowacki, G.J., M.D. Abrams, and C.F. Lorimer. 1990. Composition, structure, and historical development of Northern Red Oak stands along an edaphic gradient in Northcentral Wisconsin. Forest Science 36(2):276–292. Orrock, J.L., D. Farley, and J.F. Pagels. 2003. Does fungus consumption by the Woodland Jumping Mouse vary with habitat type or the abundance of other small mammals? Canadian Journal of Zoology 81:753–756. Orwig, D.A., and M.D. Abrams. 1994. Land-use history (1790–1992), composition, and dynamics of oak-pine forests within the Piedmont and Coastal Plain of northern Virginia. Canadian Journal of Forest Research 24(6):1216–1225. Ostfeld, R.S., C.G. Jones, and J.O. Wolff. 1996. Of mice and mast. Bioscience 46:323–330. Petrillo, H.A., and J.D. Witter. 2004. Invertebrate biodiversity in northern hardwood ecosystems under varying disturbance regimes. Pp. 88–91, In C.A. Evans, J.A. Lucas, and M.J. Twery (Eds.). Beech Bark Disease: Proceedings of the Beech Bark Disease Symposium. General Technical Report NE-331. USDA Forest Service, Northern Research Station, Newtown Square, PA. 149 pp. Rodrigo, A., F. Sarda-Palomera, J. Bosch, and J. Retana. 2008. Changes in dominant ground beetles in Black Pine forests with fi re severity. Ecoscience 15(4):442–452. Schmidt, K.A., and R.S. Ostfeld. 2008. Numerical and behavioral effects within a pulsedriven system: Consequences for direct and indirect interactions among shared prey. Ecology 89:635–646. Seldon, D.S., and J.R. Beggs. 2010. The effi cacy of baited and live capture pitfall traps in collecting large-bodied forest carabids. New Zealand Entomologist 33:30–37. Smock, L.A., and C.M. MacGregor. 1988. Impact of the American chestnut blight on aquatic shredding macroinvertebrates. Journal of North American Benthological Society 7:212–221. Sork, V.L., J. Bramble, and O. Sexton. 1993. Ecology of mast-fruiting in three species of North American deciduous oaks. Ecology 74(2):528–541. Stadler, B., T. Muller, D. Orwig, and R. Cobb. 2005. Hemlock Woolly Adelgid in New England forests: Canopy impacts transforming ecosystem processes and landscapes. Ecosystems 8:233–247. Storer, A.J., J.N. Rosemier, B.L Beachy, and D.J. Flaspohler. 2004. Potential effects of beech bark disease and decline of beech on abundance of birds and small mammals. Pp. 72–78, In C.A. Evans, J.A. Lucas, and M.J. Twery (Eds.). Beech Bark Disease: Proceedings of the Beech Bark Disease Symposium. General Technical Report NE- 331. USDA Forest Service, Northern Research Station, Newtown Square, PA. 149 pp. Suchomel, J., and M. Heroldova. 2008. Effect of seed crop of trees on the abundance and body parameters of granivorous mammals in isolated forest stands of southern Moravia (Czech Republic). Polish Journal of Ecology 56(1):181–186. Takahashi, K., and M.J. Lechowicz. 2007. Do interspecifi c differences in sapling growth traits contribute to the co-dominance of Acer saccharum and Fagus grandifolia. Annals of Botany 101:103–109. USDA Forest Service. 2004a. Betula lenta. L. Available online at http://www.na.fs. fed.us/pubs/silvics_manual/volume_2/fagus/grandifolia.html. Accessed 30 December 2011. USDA Forest Service. 2004b. Fagus grandifolia. Available online at http://www.na.fs. fed.us/pubs/silvics_manual/volume_2/fagus/grandifolia.html. Accessed 30 December 2011. 410 Northeastern Naturalist Vol. 19, No. 3 Wang, G., J.O. Wolff, S.H. Vessey, N.A. Slade, J.W. Witham, J.F. Merritt, M.L. Hunter, Jr., and S.P. Elias. 2009. Comparative population dynamics of Peromyscus leucopus in North America: Influences of climate, food, and density dependence. Population Ecology 51:133–142. Werner, S.M., and K.F. Raffa. 2000. Effects of forest management practice on the diversity of ground-occuring beetles in mixed northern forests of the Great Lakes Region. Forest Ecology and Management 139(1–3):135–155. White, R.E. 1983. Peterson field Guide to Beetles. Houghton Mifflin Harcourt, Boston, MA. 384 pp. Whitaker, J.O, Jr. 1963. A study of the Meadow Jumping Mouse, Zapus hudsonius (Zimmerman), in central New York. Ecological Monographs 33(3):215–254. Whitaker J.O., Jr., and W.J. Hamilton. 1998. Mammals of the Eastern United States. Comstock Publishing Associates. Cornell University Press, Ithaca, NY. 583 pp. Wilson, J.A., and K.E. Mabry. 2010. Trapping mammals in a cautious world: The effects of disenfectants on trap success. Western North American Naturalist 70(4):467–473. Wolff, J.O. 1996. Population fluctuations of mast-eating rodents are correlated with production of acorns. Journal of Mammalogy 77(3):850–856.