2009 NORTHEASTERN NATURALIST 16(4):491–500 Canada Warbler Habitat Use of Northern Hardwoods in Vermont Jameson F. Chace1,2,4,*, Steven D. Faccio3,5, and Abraham Chacko1 Abstract - We examined habitat use by Wilsonia canadensis (Canada Warbler), a migratory songbird population in a 40-year decline. We used a long-term forest bird monitoring program in Vermont to compare the structural components of sites of warbler presence and absence. Habitats occupied included lowland Picea-Abies (spruce-fir), northern hardwood, and Quercus-Carya (oak-hickory) forests, and Acer rubrum-Thuja (Red Maple-cedar) and cedar-fir swamps. Northern hardwood forest detections were explored in greater detail due to the greater extent of coverage in Vermont, the higher number of survey points (n = 80), and high percentage of Canada Warbler detections at those points (29%). Within the northern hardwood forests, warblers occurred in patches with a lower canopy height and higher percent ground cover of shrubs and ferns than patches where warblers were not detected. These three parameters were also the strongest set of competing Akaike’s information criterion model scores based on the patch attributes. In the northern hardwoods of the northeast, the conditions of reduced average canopy height and increased ground cover are created naturally by wind throw, ice storms, and insect damage, as well as under some forms of timber management. Canada Warblers appear to prefer these forest structural conditions because they provide abundant foraging strata, conceal nesting sites, and expose song perches. Introduction Wilsonia canadensis L. (Canada Warbler) is a Nearctic-Neotropical migratory bird that winters in South America and breeds in the northeastern United States, southeastern Canada, and through the transitional hardwoodboreal forests extending west into Alberta. This species inhabits a variety of upland and lowland habitats in the Atlantic Northern Forests of New England (Conway 1999, Dettmers 2003). Patchily distributed across the breeding range, the Canada Warbler has highest densities in swamps where excessive wet conditions and poor soils limit canopy closure and favor shrub growth and in areas having forest floors with pronounced relief. In upland forests, Canada Warblers appear to be disturbance specialists, moving into regenerating forest patches following wind-throw (Hagan and Grove 1999), ice damage (Faccio 2003), or timber removal (King and DeGraaf 2000). The Canada Warbler is one among several migratory birds breeding in North America that have exhibited long-term population declines, raising 1Department of Biology, Villanova University, Villanova, PA 19085. 2Center for Northern Forest Research, PO Box 192, Island Pond, VT 05846. 3Conservation Biology Department, Vermont Institute of Natural Science, Woodstock Road, Quechee, VT, 05059. 4Current address - Department of Biology and Biomedical Science, Salve Regina University, 100 Ochre Point Avenue, Newport, RI 02840. 5Current address - Vermont Center for Ecostudies, PO Box 420, Norwich, VT 05055. *Corresponding author - jameson.chace@salve.edu. 492 Northeastern Naturalist Vol. 16, No. 4 concerns about habitat loss and degradation in breeding and nonbreeding areas (Martin and Finch 1995, Robbins et al. 1989). Results from the North American Breeding Bird Survey indicate nearly four decades of Canada Warbler decline throughout the northeastern portion of its range. Estimates of annual population change since 1980 range between -3.8% and -7.3% (Sauer et al. 2005). Reasons for the declines are unknown; however, loss of breeding habitat may be a contributing factor (Conway 1999). The Northeast Endangered Species and Wildlife Diversity Technical Committee, Partners in Flight, and the North American Bird Conservation Initiative (NABCI) have identified the Canada Warbler as one of the region’s highest priorities for conservation and research (Dettmers 2003, Rich et al. 2004, Therres 1999). Our objective was to measure patch occurrence by Canada Warblers across a diversity of forested habitat types within large contiguous blocks of mature forest in Vermont and determine specific structural requirements within the upland northern hardwood forests. Methods Study area We surveyed for all bird species at 27 study sites distributed throughout large contiguous patches of forest in Vermont. Nineteen sites (70%) were in two of the state’s biophysical regions (Northern Green Mountains and Champlain Valley), and from 1 to 3 sites were located in 5 of the 6 other biophysical regions of the state (Fig. 1). Each site was in mature, contiguous, interior forest, and consisted of at least 4 point-count stations (124 total point counts) spaced at least 200 m apart. Elevations ranged from 310 to 810 m. Sites were located in forest stands representing eight cover types: northern hardwood (n = 18), lowland Picea-Abies (spruce-fir; n = 2), Quercus- Carya (oak-hickory; n = 2), Thuja-Abies (cedar-fir) swamp (n = 1), dry oak (n = 1), lakeside floodplain (n = 1), Tsuga-Pinus (hemlock-pine; n = 1), and Acer rubrum L.-Thuja (Red Maple-cedar) swamp (n = 1). Northern hardwood forests are the state’s dominant forest type, consisting primarily of Fagus grandifolia Ehrh. (American Beech), Acer saccharum Marsh (Sugar Maple), and Betula lutea Michaux (Yellow Birch), with lesser amounts of B. papyrifera Marsh (Paper Birch) and Picea rubens Sargent (Red Spruce) at higher elevations. Lowland spruce-fir forests are restricted to the colder regions of the state, primarily the Northeast Highlands (Fig. 1), and were dominated by Abies balsamea Miller (Balsam Fir ) and Red Spruce, with lesser amounts of Pinus strobus L. (White Pine), Red Maple, Larix laricina K. Koch (Tamarack), and Betula spp. (birches). Both lowland spruce-fir sites in our study were forested wetlands. Oak-hickory forests, which occur in the warmer valleys, are composed primarily of Quercus rubra L. (Red Oak), Carya ovata Miller (Shagbark Hickory), and Ostrya virginiana Miller (Hophornbeam). Cedar-fir swamps are dominated by Thuja occidentalis L. (Northern White Cedar) and Balsam Fir, with lesser amounts of Red Spruce, Tamarack, and Red Maple. Dry oak forests, found on rocky ridge tops, are composed of Red Oak and Q. alba L. (White Oak), occasionally with lesser 2009 J.F. Chace, S.D. Faccio, and A. Chacko 493 amounts of Q. prinus L. (Chestnut Oak). Lakeside floodplain forests occur primarily along the shores of Lake Champlain, and are dominated by A. saccharinum L. (Silver Maple) and Fraxinus pennsylvanica Marsh (Green Ash), with lesser amounts of Ulmus americana L. (American Elm) and Salix spp. (willows). Hemlock-pine forests are dominated by Tsuga canadensis L. (Eastern Hemlock) and White Pine, often with a mix of Red or Sugar Maple and American Beech. Red Maple-cedar swamps occur primarily in the Champlain Valley, and are dominated by Red Maple, Northern White Cedar, and F. nigra Marsh (Black Ash). Figure 1. Distribution of study sites (black dots) in Vermont by biophysical region. 494 Northeastern Naturalist Vol. 16, No. 4 Bird surveys Point counts (10 min) were conducted twice annually during June as part of the Vermont Forest Bird Monitoring Program (FBMP). The FBMP uses point-count surveys at mature, interior forest sites throughout the state to collect habitat-specific baseline data on species composition and relative abundance (Faccio et al. 1998). Surveys began by 0600 hrs and concluded by 1000 hrs (EDT), and were not conducted on days with rain or moderate to high winds. Experts in visual and aural bird identification recorded all individual birds within two distinct distance classes from the center point: within and beyond the 50-m radius. Only Canada Warblers detected within 50 m were included in this analysis. Each point was scored for Canada Warbler “presence” (≥1 warbler detected) or “absence” (no detections) on either of the two survey days per year in either 2001 or 2002. Vegetation measurement Vegetation was measured at each point-count location in June and July 2001 or 2002 and generally followed the BBIRD protocol (Martin et al. 1997). Four 11.3-m radius plots were established at each point-count location, one centered on the census point, and three located 35 m distant at angles of 0, 120, and 240 degrees from the center point. Litter depth, ground cover estimates, and shrub/sapling density were measured within 5 m of the center of each vegetation plot; canopy metrics were obtained within the 11.3-m radius. Average organic-litter depth (cm) was calculated from 12 evenly spaced locations along two 10-m perpendicular lines oriented to the cardinal directions and crossing at the plot center. Ground-cover metrics consisted of visual estimates of the percent cover between 0 and 50 cm in height, and within 5 m of the plot center, for the following 10 variables: total green vegetation (Total); woody perennial plants (Shrub); ferns (Fern); grasses and sedges (Grass); broad-leaf, non-woody plants (Forb); mosses (Moss); ground covered with leaf litter (Leaf Litter); downed logs (>12 cm dbh, >2.5 m long; Logs); open ground and rocks (Bare); and standing water (Water). Each ground-cover type was measured independent of all other categories, such that the sum would typically be greater than the “Total.” All shrubs and saplings (>50 cm in height, <8 cm dbh) were identified to species and counted within 5 m of the plot center. Shrubs and saplings were classified by species (or “dead”) as either small (<2.5 cm diameter at 10- cm height) or large (≥2.5 cm). Trees were classified by size class (Class 1: 8–23 cm dbh; Class 2: 23–38 cm dbh; Class 3: >38 cm dbh) for each species (including dead standing snags) and counted within the 11.3-m radius plot. Canopy height of an average tree in the plot was measured with a clinometer. Percent canopy cover was estimated from the average of four readings of a convex, spherical crown forest densiometer held at waist height, measured at plot center oriented to the cardinal directions; canopy cover was partitioned as the total canopy cover and upper canopy cover (>5 m from the ground). Statistical analysis All tests were nonparametric because data were not normally distributed, even after log-transformation. All statistics were conducted using 2009 J.F. Chace, S.D. Faccio, and A. Chacko 495 SAS JMP 5.0 (SAS Institute Inc., Cary, NC). Descriptive statistics include means and standard error unless otherwise noted; univariate statistics used a Bonferroni-adjusted alpha. The Wilcoxon-Mann-Whitney two-sample test was used to compare occupied and unoccupied point-count locations in univariate analyses (Bonferroni-adjusted alpha). Using an information-theoretic approach, we evaluated sets of competing models investigating Canada Warbler detections in northern hardwood forests of Vermont by comparing Akaike’s Information Criterion scores (AIC; Burnham and Anderson 2002). The models were based on the most likely factors derived from univariate tests (P < 0.05) after highly correlated variables (Pearson product-moment correlation) were removed. Through an iterative process using ≤5 variables, the most robust model was selected. Results Regional scale Canada Warblers were detected within 50 m of 31 of the 124 census points. The 31 points occurred at 16 of the 27 study sites (Table 1). Canada Warblers were detected at study sites in all biophysical regions sampled except Vermont Valley (one site); the Southern Vermont Piedmont was not sampled. The majority of detections (71%) occurred at sites within the Northern Green Mountains (Fig. 1), the biophysical region with the greatest coverage of study sites and predominately covered by northern hardwood forest. Overall, habitats occupied by Canada Warblers included northern hardwood forest, lowland spruce-fir forest, oak-hickory forest, cedar-fir swamp, and Red Maple-cedar swamp (Table 1). Within habitat type: Northern hardwoods We limited quantitative analysis to northern hardwood forests because the majority of study sites occupied by Canada Warblers were northern hardwood forests (Table 1), while other habitat types were relatively undersampled. Northern hardwoods comprised 80 of the 124 of the survey points (65%), of which Canada Warblers occupied 23 points (29%). In northern hardwood forests, Canada Warblers were present at points that had signifi- cantly greater total ground cover and lower canopy height (Table 2). Table 1. Canada Warbler presence (n = 31) and absence (n = 93) in major forest habitat types in Vermont. Habitat type Survey points % occupied Northern hardwood 80 29 Lowland spruce-fir 10 30 Oak-hickory 10 20 Cedar-fir swamp 5 40 Dry oak 5 0 Lakeside floodplain 5 0 Hemlock-pine 5 0 Red Maple-cedar swamp 4 25 496 Northeastern Naturalist Vol. 16, No. 4 We evaluated sets of competing AIC models investigating Canada Warbler detections in northern hardwood forests of Vermont based on the most significant variables (P < 0.2), as determined in univariate tests (Table 2), to explain Canada Warbler presence or absence. No more than five variables were used to model Canada Warbler presence in the northern hardwood points (n = 80). Total ground cover was removed from models because of a high correlation with low (<50 cm) shrub cover (r = 0.69), fern cover (0.51), and forb cover (0.56). The best model for predicting Canada Warbler detections at northern hardwood points was based on vegetative structure, including shrub and fern cover, and canopy height (Table 3). Canopy height was highly correlated with total ground cover (r = -0.38; F1, 82 = 13.447, P = 0.0004), shrub cover (r = -0.39; F1, 82 = 15.1733, P = 0.0002), fern cover (r = -0.21; F1, 82 = 3.9348, P = 0.051) and number of small alive stems (r = -0.34 ; F1, 82 = 10.452, P = 0.0018). Additional variables included in the model made only marginal changes in log-likelihood with less predictive power than the model presented (Table 3). Table 2. Habitat characteristics associated with Canada Warbler presence and absence in northern hardwood forest (n = 80); means (SE) of four 11.3-m radius (0.04-ha) plots at each point. Ground-cover estimates and number of saplings were measured within a 5 m radius of census point (0.0078 ha). * indicates statistical significance (Bonferroni-adjusted alpha, P = 0.0021). Absent Present Character (n = 57) (n = 23) Wilcoxon Z P Average leaf litter depth (cm) 3.7 (0.2) 4.1 (0.4) 0.6084 0.5500 Ground cover: % total ground cover 53.6 (2.5) 72.6 (3.9) 3.7838 0.0002* % shrub cover 15.2 (1.5) 28.4 (4.4) 2.6911 0.0071 % fern cover 12.1 (1.4) 20.5 (3.1) 2.9779 0.0029 % grass cover 1.9 (0.5) 4.4 (1.8) 1.7986 0.0700 % forb cover 15.9 (1.6) 23.2 (3.1) 2.1023 0.0360 % moss cover 6.7 (0.5) 10.1 (1.3) 2.4388 0.0150 % leaf cover 85.1 (1.1) 85.6 (1.5) 0.0643 0.9000 % downed logs 6.3 (0.7) 6.3 (0.7) 0.8212 0.4100 % bare ground 3.0 (0.4) 3.6 (1.0) 0.2426 0.8100 % water cover 0.6 (0.3) 0.05 (0.05) 2.0380 0.0420 Canopy: Average canopy height (m) 20.34 (0.4) 16.77 (0.97) 3.4919 0.0005* % canopy cover 91.7 (0.4) 92.1 (1.2) 1.0932 0.2700 % canopy cover >5 m 88.7 (0.6) 85.1 (1.6) 1.7856 0.0700 Sapling/shrubs: No. dead small stems 1.4 (0.2) 1.6 (0.4) 0.1291 0.9000 No. dead large stems 1.1 (0.1) 2.0 (0.5) 1.0258 0.3100 No. living small stems 38.3 (2.9) 47.3 (5.9) 1.1871 0.2300 No. living large stems 12.5 (0.8) 16.3 (1.8) 1.6969 0.0890 Trees: No. trees 8–22.9 cm dbh 14.6 (0.7) 18.7 (1.8) 2.0241 0.0430 No. trees 23.0–37.9 cm dbh 3.7 (0.2) 3.7 (2.1) 0.2476 0.8000 No. trees >38.0 cm dbh 1.3 (0.1) 1.2 (0.1) 0.2493 0.8000 No. dead trees 8–22.9 cm dbh 1.8 (0.2) 2.8 (0.4) 1.9140 0.0600 No. dead trees 23.0–37.9 cm dbh 0.6 (0.1) 0.7 (0.1) 0.5727 0.0600 No. dead trees >38.0 cm dbh 0.2 (0.03) 0.2 (0.05) 0.7127 0.4700 2009 J.F. Chace, S.D. Faccio, and A. Chacko 497 Discussion In Vermont, Canada Warblers occupy breeding sites in a variety of forest types, including northern hardwoods, Red Maple-cedar swamps, lowland spruce-fir forests, cedar-fir swamps, and oak-hickory forests (Table 1). Within northern hardwood forests, warbler detections are related to the structure of the canopy and understory. Canada Warblers use sites that have lower average canopy height and higher coverage of low (<50 cm) ground cover (Table 2), primarily ferns and shrubs (Table 3). The structural attributes described here may be important to Canada Warblers for several reasons. First, Canada Warblers sing from exposed, elevated perches (J.F. Chace and S.D. Faccio, unpubl. data; Kendeigh 1945), and lower average canopy height may be related to canopy openings, snapped tops, and other structural features associated with canopy disturbance events (e.g., wind-throw, ice storms, insect damage). Second, high densities of shrubs and saplings, such as those found in disturbed forests or forested wetlands, provide cover and favorable foraging structure. Two studies of warbler foraging behavior, conducted in upland forests of New Hampshire and Wisconsin, found that Canada Warblers concentrate their feeding effort in shrubs and low tree branches at heights of 3–5 m (Sabo and Holmes 1983, Sodhi and Paszkowski 1995). This range corresponds with the leafy stratum often formed by wetland shrubs (Miller 1999) and by regenerating forests 6–20 years after harvest (Hagan et al. 1997). Lastly, Canada Warblers nest on the ground, usually on raised hummocks under dense ground cover (Conway 1999; J.F. Chace, pers. observ.). Structurally complex forest floors, with hummocks, rootballs, downed woody debris, and dense patches of vegetation may provide concealment for nests and young. The use of sites with higher ground cover of ferns and shrubs may provide cover for Canada Warbler nests. Several previous studies have shown that Canada Warblers occupy a variety of forest types with structural conditions similar to those found in this study. In this study, we found warblers at points with lower than average canopy height, perhaps indicative of past wind-throw and ice storm damage. Canopy disturbance resulting from wind-throw, insect outbreaks, ice storms, Table 3. Aikiake’s information criterion (AICc) for structural habitat components associated with Canada Warbler presence in northern hardwood forests sites (n = 85) in Vermont; best ten models of 3 to 5 variables are presented. K = number of model parameters. Model Log-likelihood K AICc Δ AICc SHRUB, CANHT, FERNS 38.158615 5 13.9227 0 SHRUB, CANHT, FERNS + SHRUBS ALIVE large 36.314240 6 16.1971 -2.2744 SHRUB, CANHT, FERNS + WATER 36.469787 6 16.2008 -2.2781 SHRUB, CANHT, FERNS + MOSS 36.480940 6 16.2011 -2.2784 SHRUB, CANHT, FERNS + FORBS 36.956112 6 16.2123 -2.2896 SHRUB, CANHT, FERNS + TREES (Class 1) 37.249450 6 16.2192 -2.2965 SHRUB, CANHT, FERNS + HIGH CANOPY 37.728050 6 16.2302 -2.3076 SHRUB, CANHT, FERNS + GRASS 38.059350 6 16.2378 -2.3152 SHRUB, CANHT, FERNS + DEAD TREES (Class 1) 38.158603 6 16.2401 -2.3175 SHRUB, CANHT, FERNS, MOSS + WATER 34.128043 7 18.5208 -4.5981 498 Northeastern Naturalist Vol. 16, No. 4 and timber harvest often create these conditions through increased sunlight penetration that results in vigorous shrub and sapling growth, and through an increase to forest floor complexity due to downed woody material. Hagan and Grove (1999) frequently encountered Canada Warblers in small treefall gaps within large blocks of mature forest in Maine, and Hagan and Meehan (2002) reported a positive correlation with dead-tree basal area and understory stem density. In another Vermont study, Faccio (2003) found Canada Warblers in small (0.1–0.2 ha) canopy gaps within an extensively forested landscape during three years following a damaging ice storm in 1998. However, these gaps did not increase Canada Warbler abundance within the forest as a whole (Faccio 2003). While all points surveyed in this study were not managed for timber harvesting, timber removal may create suitable structural conditions for the Canada Warbler. In the industrial forests of Maine, Hagan et al. (1997) found Canada Warblers most abundant in young, scrubby re-growth 6–20 years following both partial-cuts and clearcuts, particularly when some unharvested trees remained. This finding mirrored results from a previous study of birdhabitat relations in Maine timberlands, in which Canada Warblers were found only in regenerating stands dominated by stems <10 cm dbh and >2 m in height; highest counts occurred in areas where saplings exceeded 4.5 m and where loggers had left some trees in the overstory (Titterington et al. 1979). In a New Hampshire study, King and DeGraaf (2000) found Canada Warblers to be significantly more abundant in 3- to 5-year-old clearcuts and shelterwood cuts, than in mature, closed canopy hardwood stands. These other studies are consistent with our finding of Canada Warblers detected in northern hardwood patches with high ground cover and high density of large stems. Canada Warbler populations appear to be decreasing across their range (Sauer et al. 2005), precipitating a need for understanding the habitat conditions that promote occupancy. Canada Warblers in the mature northern hardwood forests of Vermont may benefit from factors that create small canopy gaps and promote understory growth. Perhaps small-group-selection harvesting is a candidate management tool for this songbird with a high regional conservation emphasis. We caution that Canada Warbler detections at a single point does not necessarily equate with successful forest-patch occupancy nor is a proxy for successful pairing or reproductive success. The reproductive response of these warblers to different harvesting regimes remains untested. Bock and Jones (2004), in a review of the relationship between avian density and reproduction, found that 30% of the studies had a negative correlation. Clearly, studies that measure the reproductive success of Canada Warblers across different forest types and harvesting regimes is needed. The decline of the Canada Warbler population is perplexing. Long-term declines on the breeding grounds of the United States and Canada occur in regions where mature forests are abundant and habitat conditions seem to be suitable. Certainly, impacts from wind throw, ice storms, and insect damage in these forests today create small canopy gaps that these warblers appear to respond to (Faccio 2003, Hagan and Grove 1999). Mature forests are extensive in Canada, where 80% of the global Canada Warbler population breeds. 2009 J.F. Chace, S.D. Faccio, and A. Chacko 499 Yet this portion of the population declined by 4.5% between 1968 and 2007 (Savignac 2008). Habitat loss and degradation on the breeding grounds have impacted the population; however, population limitation most likely stems from habitat loss on the species’ winter range in South America (Lambert and Faccio 2005). The primary over-wintering grounds of the Canada Warbler includes the mid-elevation subtropical forests of the northern Andes, especially Columbia (Conway 1999, Robinson et al. 1995). These are among the most threatened habitats in the world, where 90% of forest has been cleared by activities relating to agriculture, fuel wood harvesting, cultivation of illegal drugs, and herbicide treatments to eradicate drug crops (Davis et al. 1997). Clearly, there is a need for research on habitat use and survivorship of nonbreeding populations of the Canada Warbler. Acknowledgments We thank James Gillis, Rob Gregory, Cristin Chiha, Jenn McLaughlin, and Jennifer Schomp for measuring vegetation at the Forest Bird Monitoring Project point-count stations. We are especially grateful to the following volunteers and field staff who helped conduct bird surveys: Carl Anderson, Jayson Benoit, Ernie Buford, Chip Darmstadt, Brett Engstrom, Ted Gaine, Mary Gaudette, Jim Graves, Warren King, Mark LaBarr, Everett Marshall, Sean Macfaden, Bryan Pfeiffer, Alan Quackenbush, Charlie Rabatin, Chris Rimmer, Sue Staats, Ruth Stewart, and Ned Swanberg. Clait Braun and four anonymous reviewers provided very valuable comments on earlier drafts. Financial support for this project was generously provided by the William P. Wharton Foundation, Sweet Water Trust, and Villanova University. Literature Cited Bock, C.E., and Z.F. Jones. 2004. Avian habitat evaluation: Should counting birds count? Frontiers in Ecology: 2:403–410. Burnham, K.P., and D.R. Anderson. 2002. Model Selection and Multimodel Inference: A Practical Information-theoretic Approach. 2nd Edition. Springer-Verlag, New York, NY. Conway, C.J. 1999. Canada Warbler. In A. Poole and F. Gill (Eds.). The Birds of North America, No. 421. The Birds of North America, Inc. Philadelphia, PA. Davis, S.D., V.H. Haywood, O. Herrera-MacBryde, J. Villa-Lobos, and A. Hamilton. 1997. Centres of Plant Diversity: A Guide and Strategy for Their Conservation. Volume 3: The Americas. IUCN Publications Unit, Cambridge, UK. Available online at http://www.nmnh.si.edu/botany/projects.cpd/. Dettmers, R. 2003. Priority bird species in Bird Conservation Region 14. North American Bird Conservation Initiative, Arlington, VA. Faccio, S.D. 2003. Effects of ice storm-created gaps on forest breeding-bird communities in Central Vermont. Forest Ecology and Management 186:133–145. Faccio, S.D., C.C. Rimmer, and K.P. McFarland. 1998. Results of the Vermont Forest Bird Monitoring Program, 1989–1996. Northeastern Naturalist 5:293–312. Hagan, J.M., and S.L. Grove. 1999. Bird abundance and distribution in managed and old-growth forest in Maine. Report No. MM-9901. Manomet Center for Conservation Sciences, Brunswick, ME. Hagan, J.M., and A.L. Meehan. 2002. The effectiveness of stand-level and landscape- level variables for explaining bird occurrence in an industrial forest. Forest Science 48:2002. 500 Northeastern Naturalist Vol. 16, No. 4 Hagan, J.M., P.S. McKinley, A.L. Meehan, and S.L. Grove. 1997. Diversity and abundance of landbirds in a northeastern industrial forest. Journal of Wildlife Management 61:718–735. Kendeigh, S.C. 1945. Community selection by birds on the Helderberg Plateau of New York. Auk 62:418–436 King, D.I., and R.M. DeGraaf. 2000. Bird species diversity and nesting success in mature, clearcut, and shelterwood forest in northern New Hampshire, USA. Forest Ecology and Management 129:227–235. Lambert, J.D., and S.D. Faccio. 2005. Canada Warbler population status, habitat use, and stewardship guidelines for the northeastern forests. VINS Technical Report 05-4. Vermont Institute of Natural Science. Woodstock, VT. Martin, T.E., and D.M. Finch (Eds.). 1995. Ecology and Management of Neotropical Migratory Birds: A Synthesis and Review of Critical Issues. Oxford University Press, New York, NY. Martin, T.E., C.R. Paine, C.J. Conway, W.M. Hochachka, P. Allen, and W. Jenkins. 1997. BBIRD field protocol. Montana Cooperative Wildlife Research Unit, University of Montana, Missoula, MT. Miller, N.A. 1999. Landscape and habitat predictors of Canada Warbler (Wilsonia canadensis) and Northern Waterthrush (Seiurus noveboracensis) occurrence in Rhode Island swamps. M.Sc. Thesis. University of Rhode Island, Kingston, RI. Robbins, C.S., J.R. Sauer, R.S. Greenberg, and S. Droege. 1989. Population declines in North American birds that migrate to the Neotropics. Proceedings of the National Academy of Sciences USA 86:7658–7662. Robinson. S.K., J.W. Fitzpatrick, and J. Terborgh. 1995. Distribution and habitat use of Neotropical migrant landbirds in the Amazon basin and Andes. Bird Conservation International 5:305–323. Rich, T.D., C.J. Beardmore, H. Berlanga, P.J. Blancher, M.S. W. Bradstreet, G.S. Butcher, D.W. Demarest, E.H. Dunn, W. C. Hunter, E.E. Inigo-Elias, J.A. Kennedy, A.M. Martell, A.O. Panjabi, D.N. Pashley, K.V. Rosenberg, C.M. Rustay, J.S. Wendt, and T.C. Will. 2004. Partners in Flight North American landbird conservation plan. Cornell Lab of Ornithology, Ithaca, NY. Sabo, S.R., and R.T. Holmes. 1983. Foraging niches and the structure of forest bird communities in contrasting montane habitats. Condor 85:121–138. Sauer, J.R., J.E. Hines, and J. Fallon. 2005. The North American Breeding Bird Survey, Results and Analysis 1966 – 2004. Version 2005.2, USGS Patuxent Wildlife Research Center, Laurel, MD. Savignac, C. 2008. COSEWIC assessment and status report on the Canada Warbler Wilsonia canadensis in Canada. Canadian Wildlife Service Environment Canada, Ottawa, ON, Canada. Sodhi, N.S., and C.A. Paszkowski. 1995. Habitat use and foraging behavior of four Parulid warblers in a second-growth forest. Journal of Field Ornithology 66:277–288. Therres, G.D. 1999. Wildlife species of conservation concern in the Northeastern United States. Northeast Wildlife 54:93–100. Titterington, R.W., H.S. Crawford, and B.N. Burgason. 1979. Songbird response to commercial clear-cutting in Maine spruce-fir forests. Journal of Wildlife Management 43:602–609.