Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 187 Canaan Valley & Environs 2015 Southeastern Naturalist 14(Special Issue 7):187–202 Vegetation of Managed Grasslands in the Canaan Valley National Wildlife Refuge Kelly A. Chadbourne1,2 and James T. Anderson1,* Abstract - Much of Canaan Valley was converted to agricultural uses following logging in the early 1900s. More recently, some land has undergone succession from grassland to scrub-shrub habitat. We evaluated vegetation and habitat structure in mowed and unmowed hayfields and idle pastures during 1999 and 2000. We observed 71 plant species on 3 hayfields and 3 pastures. Solidago ulginosa (Bog Goldenrod), Solidago rugosa (Wrinkle-leaved Goldenrod), Achillea millefolium (Yarrow), Dactylis glomerata (Orchard Grass), Phalaris arundinacea (Reed Canary Grass), and Hypericum densiflorum (Glade St. John’s Wort) were the most common taxa. Species composition and abundance varied by field type and mowing treatment. Vegetation was taller in pastures than in hayfields; standing dead vegetation was greater in unmowed plots than in mowed plots. Mowing is useful for maintaining vegetative structure for wildlife and may influence plant species composition and abundance. Introduction The condition of West Virginia’s grasslands is driven by mixes of anthropogenic changes and natural disturbances. Prior to European settlement, most of West Virginia was forested, although open-land habitats occupied large patches (Core 1949). Much of the treeless area was probably bog and other wetland habitats, but grasslands and grass-bald communities comprised some of these sites (Core 1949). Grass-bald communities are a unique high-elevation open-land feature vegetated by grasses and forbs. Although the origin of these communities is unknown, researchers have suggested that natural and anthropogenic disturbances created them (Core 1949, Mark 1958, Rentch and Fortney 1997). Since European settlement, the amount of grassland habitat in WV has increased. Surface mining has resulted in the development of many grassland habitats over the last 70 years (Whitmore 1980, Whitmore and Hall 1978). However, grasslands created by farming—specifically pastures and hayfields—have also increased dramatically since European settlement (Jones and Vickery 1997). In the Appalachian region, idle farmlands provide important habitat for grassland- dependent wildlife species (Bollinger et al. 1990). Idle hayfields and pastures are increasing in the East due to changes in farming practices and the purchase of farmland for other uses. These new and changing land practices provide habitats that benefit a variety of wildlife (Farris and Cole 1981, McCoy et al. 2001). Shrubs and trees become established on abandoned hayfields and pastures, and if open 1Wildlife and Fisheries Resources Program, Division of Forestry and Natural Resources, West Virginia University, PO Box 6125, Morgantown, WV 26506. 2Current address - US Fish and Wildlife Service, Great Bay National Wildlife Refuge, 100 Merrimac Drive, Newington, NH 03801. *Corresponding author – jim.anderson@mail.wvu.edu. Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 188 sites are not maintained through active management, habitat for grassland wildlife species will be depleted and may eventually disappear in West Virginia. Grazing, burning, and mowing or haying are 3 widely accepted techniques for grassland management. Implementing these treatments is necessary to prevent woody encroachment into prairie fragments (Burger et al. 1994). When haying is used as a management technique, it often has the same effect as burning because cut material is removed. Mowing fields can provide high-quality grassland habitat by controlling woody vegetation, lowering vegetative height, and reducing litter build-up, especially if the cuttings are collected and taken away (Sample and Hoffman 1989). This activity halts succession and provides suitable wildlife habitat, such as nesting grounds for grassland birds. When mowing is used as a tool for grassland bird conservation, it is important to delay mowing until the breeding season has concluded. Mowing has been implicated as a reason for the declines of grassland nesting species (Bollinger et al. 1990, Frawley and Best 1991) because mowing early in the season disrupts nesting and will ultimately lead to nest failure. Unfortunately, late season mowing often decreases the value of hay to landowners and farmers who use the hay. Managers of natural areas need quantitative estimates of vegetative structure and composition so that management recommendations to enhance wildlife benefits can be achi eved. Idle grasslands and hayfields have been documented on the Canaan Valley National Wildlife Refuge (hereafter called “the Refuge”), but their structure and importance to wildlife are relatively unknown. Fortney (1975) conducted a study of the vegetative communities in Canaan Valley (hereafter, the Valley); however, data on structural components important to wildlife and management effects were neither sampled nor evaluated. The objective of this study was to estimate the composition and structure of vegetation on the Refuge’s mowed and unmowed hayfields and pastures. Field Sites This study was conducted on the Refuge in the Valley, Tucker County, WV. The Valley is about 15 mi (24 km) long and 2–4 mi (3–6 km) wide and is oriented on a northeast–southwest axis. The Valley’s floor lies at an elevation of 3150– 3248 ft (960–990 m) and is surrounded by mountains up to 1000 ft (305 m) above the valley floor (USFWS 1979). A large wetland complex occurs on the Valley’s floor (~6025 acres [2438 ha]), consisting of palustrine emergent, scrub-shrub, forested, unconsolidated bottom, and unconsolidated shore wetlands (USFWS 1979). The Valley’s climate is similar to that of northern New York, Vermont, New Hampshire, and the northern half of Maine, with cold winters and cool summers (Thornthwaite 1948). Many plants reach the southern limit of their ranges in the Valley (Fortney 1993). The growing season is relatively short, with an average of 92 frost-free days from 31 May–1 September (Fortney 1975). Summer temperatures are moderate, with an average of 75–79 °F (24–26 °C) during the day and 50–55 °F (10–13 °C) at night. Annual total rainfall is 45 in (114 cm); Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 189 rainfall was 7 in (18 cm) below average in 1999 and at normal level in 2000 (NOAA 1999, 2000). The Valley was first visited by European explorers in 1746, and people settled as early as 1800 (Fortney 1993). Prior to logging, a forest of large Picea rubens (Sarg.) (Red Spruce) with a dense understory of Rhododendron maximum L. (Rhododendron) covered the Valley (Fortney 1993). Other species present here and associated with a boreal climate included Abies balsamea L. (Balsam Fir), Tsuga canadensis L. (Eastern Hemlock), Betula alleghaniensis (F. Michx.) (Yellow Birch), Acer saccharum (Marsh.) (Sugar Maple), Fagus grandifolia (Ehrh.) (American Beech), Vaccinium oxycoccos L. (Small Cranberry), Gaultheria hispidula L. (Creeping Snowberry), Geum strictum (Aiton) (Yellow Avens), Carex leptonervia (Fernald) (Nerveless Woodland Sedge), Scirpus microcarpus (J. Presl. & C. Presl) (Panicled Bulrush), Scirpus atrocinctus (Fernald) (Blackgirdle Bulrush), Caltha palustris L. (Marsh Marigold), and Arisaema triphyllum (L.) Schott (Jack-in-the-Pulpit) (Strausbaugh and Core 1977). Within the forest, natural openings or glades occurred, which consisted of grass balds or bogs too wet for forest development. Deep layers of needles and other plant and animal matter accumulated and, together with Sphagnum spp. (sphagnum mosses), built humus-rich acidic soil. Beginning in the 1880s, the railroad rendered the Valley more accessible and thereby more susceptible to logging (Vogelmann 1978). The Red Spruce was cut, leaving the Valley bare. By 1920, few trees remained and lumbering activity throughout West Virginia declined dramatically (Vogelmann 1978). The microclimate changed and the soil dried, creating a substrate where tree seedlings were unable to become established and where fire proliferated (Fortney 1993, Vogelmann 1978). Once-forested bottomlands became sphagnum bogs on wetter sites and Polytrichum spp. (haircap moss) hummocks on slightly drier sites (Fortney 1993). Drier uplands converted to forb and grass meadows and some heavily burned areas were planted with mixed grasses (Vogelmann 1978). Agriculture in the Valley has generally been unsuccessful, although some farms exist today. Most crops produced low yields in the short growing season, but cattle production was successful and has been the predominant form of livestock since the area was logged (Vogelmann 1978). Pastures and hayfields are found in the slightly drier southern portion of the Valley. Currently, the Refuge grasslands consist of dry upland areas, Crataegus spp. (hawthorn) savannahs, saturated wet meadows, and saturated scrub-shrub wetlands, all of which are included in the broad definition of grassland systems (V ickery et al. 1999). Methods Vegetation monitoring We studied 6 grassland plots, labeled the Beall, Cortland, Freeland, Harper, Herz, and Thompson tracts, which totaled 730 ac (295 ha) (mean = 121.6 ac/ field, SE = 31.4; [49.2 ha/field, SE = 12.7]). We classified each plot according to its previous land use as idle hayfield (Beall [230 ac (93 ha)], Harper [180 ac Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 190 (73 ha)], and Thompson [59 ac (24 ha)]) or as idle pasture (Cortland [40 ac (16 ha)], Freeland [69 ac (28 ha)], and Herz [151 ac (61 ha)]). At the conclusion of the 1999 bird breeding-season, in late August, Refuge staff mowed 50% of each tract except Herz to determine the effects of habitat manipulation on vegetative structure and composition. The Herz tract was not mowed because it was too wet for mowing equipment. We placed transects for vegetation surveys on mowed and unmowed portions of each field. Transects ran from edge to edge and parallel to the long axis of each field. Placement depended on plot size, and we placed the transects 164 ft (50 m) away from the short axis of each field ( Warren 2001). We sampled the vegetation once per month during June–August of 1999 and 2000 following the methods of Best et al. (1997). We measured the grassland vegetation’s capacity to obstruct vision (i.e., vertical density [cm]) and maximum height (cm), the litter depth (cm) using a 3.28-ft (1-m) tall Robel pole (Robel et al. 1970), and the canopy coverage using a Daubenmire frame (Daubenmire 1959). We recorded vertical density at 13 ft (4 m) from the Robel pole. We measured ground cover, classified as canopy, litter, or bare ground, to the nearest 5% based on Best et al. (1997). We classified canopy coverage as either living or standing-dead vegetation and we classified vegetative cover as forb, grass, or woody plant (Best et al. 1997). Thus, the sum of forbs, grasses, and woody vegetation percentages equaled the percent canopy cover, as did the sum of the percent living plus percent standing-dead vegetation. Litter included all dead and decomposing plant material on the soil surface. We included all dead plant material found above the litter layer in standing-dead vegetation and considered it in maximum height and vertical density readings. We recorded measurements every 16.4–65.6 ft (5–20 m) depending on field size and shape (Warren 2001). In general, we recorded the above measurements every 16.4 ft (5 m) on small fields (less than 69 ac [28 ha]) and every 65.6 ft (20 m) on large fields (≥151 ac [61 ha]) (Warren 2001). Regardless of field size or shape, we recorded species of vegetation touched by a 0.2-in- (0.5-cm-) thick rod placed every 3.28 ft (1 m) along each transect. Statistical analyses We checked the data for normality by using the Shapiro-Wilk statistic and for homogeneity of variances by plotting residuals (Cody and Smith 1991). To improve normality and the distribution of variances, we used an arcsine transformation on canopy, live vegetation, dead vegetation, wood, and vertical density, and used a log (x+1) transformation on percent bare ground, vertical density, maximum height, and frequency data (Zar 1999). We analyzed habitat characteristics, the dependent variables, of plots using three-way multivariate analysis of variance (MANOVA) or three-way analysis of variance (ANOVA) between treatments (mowed and unmowed), habitat types (pastures and hayfields), and months (June, July, and August) (independent variables). Then, we analyzed the data using three-way ANOVA to compare habitat characteristics between habitat types (pastures and hayfields), years (1999 and Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 191 2000), and months (June, July, and August). For combined 1999 and 2000 data, we used all data from 1999, but only data from the unmowed portion of each field in 2000. We analyzed vertical density, maximum height, and litter depth (dependent variables) using 3 separate ANOVAs. We used MANOVA to analyze cover type (canopy, litter, or bare ground), growth state (standing live or standing dead), and vegetative type (forb, grass, or woody plant); these categories were transformed because the variables within the groups were correlated (Warren 2001). We compared the frequency of occurrence (number of points occupied/ number of points sampled) between hayfields and pastures for selected species during July 1999 and 2000. We compared frequency between years and field types using two-way ANOVA. We considered all tests to be significant at P less than 0.05. Following a significant ANOVA, we used Tukey’s multiple comparison test to separate means. Results We identified 71 plant species, including 67 on hayfields and 58 on pastures during 1999 and 2000 (Appendix 1). Species with the most coverage in hayfields and pastures were similar for both years (Appendix 1). The dominant grassland plants present on the Refuge consisted of Dactylis glomerata (Orchard Grass), Danthonia compressa (Flattened Oatgrass), Anthoxanthum odoratum. (Sweet Vernal Grass), Phleum pratense (Timothy), Solidago ulginosa (Bog Goldenrod), Solidago rugosa (Wrinkle-leaved Goldenrod), Hypericum densiflorum (Bushy St. Johnswort), and Spiraea alba (Narrow-leaved Meadowsweet). Frequency of occurrence of plant species was similar between pastures and hayfields based on July 1999 and 2000 data. The biggest difference was for Reed Canary Grass, which occurred on 9.7% of the points on hayfields but only on less than 0.0001% of points in pastures (Table 1). There were no differences between mowed and unmowed treatments for percent cover-type (Wilks’ mean = 0.786, P = 0.196), percent vegetative type (Wilks’ mean = 0.796, P = 0.218), vertical density (F1, 21 = 2.37, P = 0.138), and maximum height (F1, 21 = 0.20, P = 0.657) (Table 2). There were differences between treatments for percent growth state (Wilks’ mean = 0.710, P = 0.033), with standing-dead vegetation taller in unmowed than in mowed plots (F1, 21 = 6.05, P = 0.023) (Table 2). There were no differences between habitat types for ground cover (Wilks’ mean = 0.865, P = 0.352), growth state (Wilks’ mean = 0.926, P = 0.411), vegetative cover (Wilks’ mean = 0.734, P = 0.074), or vertical density (F1, 24 = 0.62, P = 0.440) for combined 1999–2000 data (Table 3). Vegetation was taller in pastures than in hayfields (F 1, 24 = 6.10, P = 0.021; Table 3). However, litter was deeper in hayfields than in pastures (F1, 24 = 15.95, P < 0.001; Table 3). There were no differences between years for growth state (Wilks’ mean = 0.927, P = 0.416) and litter depth (F1, 24 = 0.27, P = 0.606). Differences were detected between years for ground cover (Wilks’ mean = 0.609, P = 0.011). Of the percent groundcover Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 192 Table 1. Frequency (%; number of points occupied/number of points observed) of selected abundant species of vegetation on pastures (3) and hayfields (3) on the Canaan Valley National Wildlife Refuge, July 1999 and 2000. tr = trace Pastures Hayfields Species Mean SE Mean SE F P Achillea millefolium 2.52 1.65 3.42 0.80 1.92 0.203 Agrostis gigantea 1.11 0.31 0.12 0.12 9.65 0.015 Anthoxanthum odoratum 2.97 1.59 2.84 1.94 0.37 0.561 Asclepias syriaca 0.51 0.51 0.19 0.14 0.12 0.743 Bromus spp. 0.15 0.15 0.19 0.19 0.02 0.897 Carex spp. 2.11 1.03 2.62 0.96 0.18 0.679 Cirsium spp. 0.18 0.18 0.02 0.02 0.68 0.433 Crataegus spp. 0.04 0.04 0.03 0.03 0.04 0.849 Dactylis glomerata 6.76 4.27 4.20 2.69 0.34 0.574 Danthonia compressa 3.67 2.09 2.12 1.38 0.38 0.678 Daucus carota 0.26 0.26 0.17 0.11 0.02 0.903 Dichanthelium clandestinum 1.27 1.27 0.33 0.17 0.10 0.759 Euthamia graminifolia 2.74 2.00 1.88 1.11 0.01 0.924 Schedonorus arundinaceus 1.86 1.26 0.44 0.26 0.94 0.362 Fragaria spp. 2.12 1.78 0.48 0.13 0.58 0.469 Hieracium spp. 0.87 0.62 0.33 0.22 0.59 0.460 Holcus lanatus 2.74 1.55 2.95 0.76 0.43 0.530 Houstonia caerulea tr - 0.08 0.07 1.00 0.347 Hypericum densiflorum 6.34 4.28 3.28 2.08 0.04 0.838 Hypericum ellipticum 0.08 0.08 tr - 1.00 0.347 Juncus effusus 1.01 0.65 1.49 0.70 0.30 0.597 Leucanthemum vulgare 0.45 0.33 1.45 0.53 2.67 0.118 Lolium perenne 0.91 0.54 1.94 0.97 0.67 0.436 Lotus corniculatus 2.01 1.14 6.44 3.95 2.49 0.153 Oxalis stricta 0.32 0.25 0.51 0.26 0.46 0.517 Packera aurea 0.17 0.17 0.06 0.04 0.29 0.604 Phalaris arundinacea tr - 9.72 4.39 12.91 0.007 Phleum pratense 9.79 3.92 6.95 3.50 0.05 0.821 Plantago virginica 0.09 0.09 tr - 1.00 0.347 Poa palustris 0.03 0.03 0.21 0.21 0.61 0.457 Potentilla spp. 3.69 1.62 13.34 3.36 6.23 0.037 Prunella vulgaris 0.17 0.13 0.39 0.29 0.45 0.523 Pteridium aquilinum 0.28 0.22 0.08 0.08 0.57 0.473 Ranunculus spp. 0.34 0.19 0.57 0.25 0.58 0.456 Rumex acetosella 0.23 0.12 0.09 0.07 0.80 0.399 Salix sericea tr - 0.42 0.34 1.60 0.242 Scirpus atrocinctus 0.14 0.13 tr - 1.34 0.280 Sisyrinchium spp. 0.67 0.35 0.18 0.17 1.22 0.302 Solidago rugosa 9.78 3.19 6.32 1.45 0.35 0.573 Solidago ulginosa 6.94 3.68 4.45 2.83 0.19 0.673 Spiraea alba 3.76 3.07 1.47 0.84 0.04 0.840 Stellaria graminea 0.09 0.09 0.32 0.28 0.52 0.493 Taraxacum officinale 1.62 0.70 0.19 0.10 3.71 0.090 Trifolium aureum 0.22 0.14 0.46 0.21 0.85 0.383 Trifolium pratense 0.01 0.01 0.92 0.64 2.40 0.160 Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 193 variables, bare ground did not differ between years (F1, 24 = 1.26, P = 0.273), but the percentages of canopy cover and litter were different. Percent canopy-cover was higher in 2000 than in 1999 (F1, 24 = 8.39, P = 0.008; Fig. 1), but percent litter was higher in 1999 than in 2000 (F1, 24 = 12.33, P = 0.002; Fig. 1). Vertical density (F1, 24 = 17.71, P < 0.001) and maximum height (F1, 24 = 24.95, P < 0.001) were higher in 2000 than in 1999 (Fig. 2). Table 2. Vegetative characteristics for mowed and unmowed treatments on the Canaan Valley National Wildlife Refuge, Tucker County, WV, June–August 2000. Mowed Unmowed Variable Mean SE Mean SE Groundcover (%) Canopy 79.20 3.31 73.21 3.27 Litter 17.81 3.14 20.19 3.30 Bare ground 3.10 1.39 6.68 2.37 Growth state (%) Live 77.28 3.16 66.01 4.83 Dead 1.20 0.38 3.09 0.61 Vegetative cover (%) Forbs 37.92 2.02 35.94 2.60 Grasses 39.84 2.20 33.07 3.48 Wood 0.73 0.40 4.20 1.52 Vertical density (cm) 18.43 2.65 22.65 2.76 Maximum height (cm) 50.57 4.98 51.80 4.56 Litter depth (cm) 1.73 0.16 2.67 0.20 Table 3. Vegetative characteristics for hayfields and pastures on the Canaan Valley National Wildlife Refuge, Tucker County, WV, June–August, 1999 and 2000. Hayfields Pastures Variable Mean SE Mean SE Groundcover (%) Canopy 75.35 2.84 76.63 4.02 Litter 20.38 2.83 17.58 3.73 Bare ground 4.30 1.33 5.96 2.81 Growth state (%) Live 71.97 2.49 70.13 6.31 Dead 2.83 0.56 1.52 0.56 Vegetative cover (%) Forbs 34.21 1.94 39.40 4.30 Grasses 38.55 1.80 33.26 4.30 Woody plants 1.99 0.82 3.37 1.71 Vertical density (cm) 18.14 2.20 23.83 3.24 Maximum height (cm) 47.45 3.83 55.79 5.57 Litter depth (cm) 2.53 0.23 1.90 0.16 Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 194 Discussion When the Refuge’s grasslands were subjected to mowing, we found that vegetative structure was similar for most variables in mowed and unmowed fields at one year after treatment. McCoy et al. (2001) found that most mowed fields in northeastern Missouri showed no difference in vegetative structure between the years before and after mowing. The higher percentage of standingdead vegetation in the unmowed plots, specifically in pastures, provided cover for grassland birds at the start of their breeding season when maximum height and vertical density were lower. However, a decrease in the percentage of standing-dead vegetation on mowed plots appeared to depress bird territory establishment on the mowed sites during the summer of 2000 (Warren 2001). On the Refuge, standing-dead vegetation found on unmowed treatments contributed to increased vertical density early in the bird-breeding season. This taller vegetation may have served as a cue that attracted some birds as they chose their nest sites and territories. Prior to acquisition by the Refuge in 1994, the pastures and hayfields were actively used for agriculture, thus maintaining much of both field types in similar successional stages during our study. In general, at our sites on the Refuge, plant species frequency was similar between hayfields and pastures, which reduced the distinction between the two habitat types. However, Reed Figure 1. Comparison of percent canopy and percent litter between years on the Canaan Valley National Wildlife Refuge, Tucker County WV, 1999–2000. Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 195 Canary Grass, an exotic invasive species (Kecher and Zadler 2004), was found at a greater rate on hayfields than on pastures. Both hayfields and pastures had exotic and native grasses that may have simulated the now-rare North American natural tallgrass prairie (Steinauer and Collins 1996). Madden et al. (2000) found that Dolichonyx oryzivorus L. (Bobolink) mainly used hayfields or pastures with exotic, tall grasses. These tall, rhizomatous, graminoids are structurally similar to the native grass species they replaced. Associated bird species have adopted this introduced vegetation as breeding habitat (Madden et al. 2000). McCoy et al. (2001) stated that succession on idle grass fields often results in monotypic stands of dense vegetation with limited habitat value for wildlife species. Similarly, on the Refuge, grassland songbirds used Reed Canary Grass when other plant species were present, but they did not use dense, monotypic stands of Reed Canary Grass for territory establishment or nest location (Warren 2001). Grassland birds treated idle hayfields and idle pastures equally with regard to territory establishment and bird densities, although they showed a preference for unmowed sites when selecting territories in 2000 (Warren 2001). The quality of vegetation in actively managed grasslands may be more desirable than in unmanaged sites, but wildlife used idle grass habitats for foraging, territory establishment, breeding, protection from weather and predators, and bedding (Warren 2001). Figure 2. Comparison of vertical density (cm) and maximum height (cm) between years on the Canaan Valley National Wildlife Refuge, Tucker County, WV, 1999–2000. Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 196 While the Cortland and Herz pastures offered abundant shrub cover not typical of grasslands, in general, the hayfields and pastures that we studied had similar species composition and structure. Furthermore, grassland birds appeared to respond similarly to both habitat types. According to Knopf et al. (1988), birds within a vegetative association may respond to differences in plant species and structural variables when selecting habitats. Vegetation was taller in pastures than hayfields due to the predominance of shrubby St. Johnswort and Narrowleaved Meadowsweet in the pastures. Also, litter depth was greater in hayfields than pastures, possibly because there was more grass and forb biomass in hayfields that contributed to the litter layer (Warren 2001). However, grassland birds on the Refuge selected hayfields and pastures regardless of any potential finescale differences in plant species and structure. Some bird species, like Sturnella magna L. (Eastern Meadowlark), avoided the shrub-dominated parts of pastures and were found on grass-dominated areas of hayfields. Pastures had more shrubs, which increased the amount of edge area and vertical diversity, and provided additional habitat for wildlife species. An increase in canopy cover, vertical density, and maximum height from 1999 to 2000 can largely be attributed to differences in weather conditions between the two years. West Virginia experienced a drought in 1999 and normal rainfall in 2000 (NOAA 2000). In the Valley, May–August total precipitation was 8.6 in (21.9 cm) greater in 2000 than in 1999. This difference in precipitation may have influenced grassland plant species growth and wildlife distributions in the Valley and surrounding areas in both years (O’Connor et al. 1999). In 1999, grassland plants produced seed and died back early in response to the drought. Additionally, these plants did not achieve their maximum height or form a complete canopy because of their physiological responses to the drought (Holtzer et al. 1988). In 2000, grassland plants showed a positive growth response to the increased rainfall and thereby provided better cover for a longer time. However, even with a drought effect, plant species composition appeared to be similar between 1999 and 2000 on unmowed fields. Management implications It is important to actively manage the grasslands to provide high-quality grassland habitat for the Refuge’s wildlife. To benefit grassland species, a combination of mowing, grazing, and prescribed burning should be implemented to reset succession and maintain an open setting. These management techniques will prevent woody encroachment into grassland habitat fragments (Burger et al. 1994). Herkert et al. (1996) found that providing a mosaic of mowed and unmowed, grazed and ungrazed, and burned and unburned areas provides a full range of habitats for wildlife species. Such habitat diversity complements the different responses of grassland plant and wildlife species to management techniques. Mowing may be the most feasible option because it eliminates the need for coordinating with livestock farmers. It also avoids having to deal with unpredictable weather for prescribed burning. Sample and Hoffman (1989) found that mowing can be used Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 197 to control woody vegetation, reduce vegetative height, and reduce litter build-up if the cuttings are harvested. Mowing should be conducted on a rotational basis, leaving fields or portions of fields idle for one to two growing seasons. On the Refuge, mowing should be conducted in mid-to-late August to avoid most of the grassland birds’ nesting seasons. Further research should be conducted on the Valley’s grasslands to determine the effect of surrounding active farmland management on overall vegetative structure and composition, as well as to investigate wildlife species’ responses to different management regimes. Additionally, the vegetative structure and composition of these grasslands should be monitored to determine the long-term effects of mowing. Acknowledgments We thank the US Fish and Wildlife Service (Canaan Valley National Wildlife Refuge); West Virginia Division of Natural Resources; West Virginia University’s Division of Forestry and Natural Resources; the Davis College of Agriculture, Natural Resources, and Design at West Virginia University (Brown Faculty Development Fund); and the West Virginia Agricultural and Forestry Experiment Station (McIntire-Stennis) for funding. We are indebted to the late W.N. Grafton for field help in identifying plants. We also thank C.A. Rhoads and S.K. Reilly for assisting with data collection. R.C. Whitmore and L. Butler reviewed an earlier version of this paper. This is manuscript number 3201 of the West Virginia University Agricultural and Forestry Experiment Station. Literature Cited Best, L.B., H. Campa III, K.E. Kemp, R.J. Robel, M.R. Ryan, J.A. Savidge, H.P. Weeks, Jr., and S.R. Winterstein. 1997. Bird abundance and nesting in CRP fields and cropland in the Midwest: A regional approach. Wildlife Society Bulletin 25:864–877. Bollinger, E.K., P.B. Bollinger, and T.A. Gavin. 1990. Effects of hay-cropping on eastern populations of the Bobolink. Wildlife Society Bulletin 18:142–150. Burger, L.D., L.W. Burger, and J. Faaborg. 1994. Effects of prairie fragmentation on predation on artificial nests. Journal of Wildlife Management 58:249–254. Cody, R.P., and J.K. Smith. 1991. Applied Statistics and the SAS Programming Language, 3rd Edition. Elsevier Scientific Publications, New York, NY. 403 pp. Core, E.L. 1949. Original treeless areas in West Virginia. Journal of the Elisha Mitchell Science Society 65:306–310. Daubenmire, R.F. 1959. A canopy-coverage method of vegetation analysis. Northwest Science 35:43–64. Farris, A.L., and J.H. Cole. 1981. Strategies and goals for wildlife habitat restoration on agricultural lands. Transactions of the North American Wildlife and Natural Resource Conference 46:130–136. Fortney, R.H. 1975. The vegetation of Canaan Valley, West Virginia: A taxonomic and ecological study. Ph.D. Dissertation, West Virginia University, Morgantown, WV. 210 pp. Fortney, R.H. 1993. Canaan Valley: An area of special interest within the upland forest region. Pp. 47–65, In S.L. Stephenson (Ed.). Upland Forests of West Virginia. Mc- Clain Printing Co., Parsons, WV. 295 pp. Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 198 Frawley, B.J., and L.B. Best. 1991. Effects of mowing on breeding bird abundance and species composition in alfalfa fields. Wildlife Society Bulletin 19:135–145. Herkert, J.R., D.W. Sample, and R.E. Warner. 1996. Management of Midwestern grassland landscapes for the conservation of migratory birds. Pp. 89–116, In F.R. Thompson III (Ed.). Management of Midwestern Landscapes for the Conservation of Migratory Birds. USDA Forest Service, North Central Forest Experiment Station. 207 pp. Holtzer,T.O., T.L. Archer, and J.M. Norman. 1988. Host plant suitability in relation to water stress. Pp. 111–137, In E.A. Heinrichs (Ed.). Plant Stress-Insect Interactions. John Wiley, New York, NY. 492 pp. Jones, A., and P. Vickery. 1997. Conserving grassland birds: Managing agricultural lands including hayfields, crop fields, and pastures for grassland birds. Grassland Conservation Program, Center for Biological Conservation, Massachusetts Audubon Society, Lincoln, MA. 17 pp. Kercher, S.M., and J.B. Zedler. 2004. Multiple disturbances accelerate invasion of Reed Canary Grass (Phalaris arundinacea L.) in a mesocosm study. Oecologia 138:455–464. Knopf, F.L., J.A. Sedgewick, and R.W. Cannon. 1988. Guild structure of a riparian avifauna relative to seasonal cattle grazing. Journal of Wildlife Management 52:280–290. Madden, E.M., R.K. Murphy, and L. Murray. 2000. Models for guiding management of prairie bird habitat in northwestern North Dakota. American Midland Naturalist 144:377–392. Mark, A.F. 1958. The ecology of the southern Appalachian grass balds. Ecological Monographs 28:293–336. McCoy, T.D, E.W. Kurzejeski, L.W. Burger, Jr., and M.R. Ryan. 2001. Effects of conservation practice, mowing, and temporal changes on vegetation structure on CRP fields in northern Missouri. Wildlife Society Bulletin 29:979–987. National Oceanic and Atmospheric Administration (NOAA). 1999. National Climatological Data: Canaan Valley, West Virginia. US Department of Commerce, Washington, DC. Available online at http://www.ncdc.noaa.gov/land-based-station-data. Accessed 14 October 2014. NOAA. 2000. National Climatological Data: Canaan Valley, West Virginia. US Department of Commerce, Washington, DC. Available online at http://www.ncdc.noaa.gov/ land-based-station-data. Accessed 14 October 2014. O’Connor, R.J., M.T. Jones, R.B. Boone, and T.B. Lauber. 1999. Linking continental climate, land use, and land patterns with grassland bird distribution across the conterminous United States. Studies in Avian Biology 19:45–59. Rentch, J.S., and R.H. Fortney. 1997. The vegetation of West Virginia grass bald communities. Castanea 62:147–160. Robel, R.J., J.N. Briggs, A.D. Dayton, and L.C. Hulbert. 1970. Relationships between visual obstruction measurements and weight of grassland vegetation. Journal of Range Management 23:295–298. Sample, D.W., and R.M. Hoffman. 1989. Birds of dry-mesic and dry prairies in Wisconsin. Passenger Pigeon 51:195–208. Steinauer, E.M., and S.L. Collins. 1996. Prairie ecology: The tallgrass prairie. Pp. 39–52, In F.B. Samson and F.L. Knopf (Eds.). Prairie Conservation: Preserving North America’s Most Endangered Ecosystem, Island Press, Washington, D.C. 352 pp. Strausbaugh, P.D., and E.L. Core. 1977. Flora of West Virginia, 2nd Edition. Seneca Books, Morgantown, WV. 1079 pp. Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 199 Thornthwaite, C.W. 1948. An approach toward a rational classification of climate. Geographical Review 38:55–94. United States Fish and Wildlife Service (USFWS). 1979. Final environmental impact statement: Acquisition of lands for the Canaan Valley National Wildlife Refuge, West Virginia. US Department of the Interior, Hadley, MA. 157 pp. Vickery, P.D., P.L. Tubaro, J.M. Cardoso da Silva, B.G. Peterjohn, J.R. Herkert, and R.B. Cavalcanti. 1999. Conservation of grassland birds in the western hemisphere. Studies in Avian Biology 19:2–26. Vogelmann, H.W. 1978. Evaluation of the Canaan Valley-Cabin Mountain wetland, Tucker County, West Virginia. US Department of the Army, Corps of Engineers, Huntington, WV. 81 pp. Warren, K.A. 2001. Habitat use, nest success, and management recommendations for grassland birds of the Canaan Valley National Wildlife Refuge, West Virginia. M.Sc. Thesis. West Virginia University, Morgantown, WV. 146 pp. Whitmore, R.C. 1980. Reclaimed surface mines as avian habitat islands in the eastern forest. American Birds 34:13–14. Whitmore, R.C., and G.A. Hall. 1978. The response of passerine species to a new resource: Reclaimed surface mines in West Virginia. American Birds 32:6–9. Zar, J.H. 1999. Biostatistical Analysis, 4th Edition. Prentice Hall, Upper Saddle River, NJ. 929 pp. Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 200 Appendix 1. Percent cover of vegetative species found on Canaan Valley National Wildlife Refuge’s idle hayfields and pastures, Tucker County, WV (1999–2000). H = hayfields and P = pastures. 1999 2000 Scientific name Species H P H P Family Apiaceae Daucus carota L. Queen Anne’s Lace 0.34 0.00 0.54 0.13 Zizia aptera (A. Gray) Golden Alexander 0.04 0.05 0.00 0.00 Family Asclepidaceae Asclepias syriaca L. Common Milkweed 0.05 0.00 0.03 0.18 Family Asteraceae Achillea millefolium L. Yarrow 6.46 3.18 4.27 0.48 Anaphalis margaritacea L. Pearly Everlasting 0.00 0.00 0.04 0.00 Aster spp. Aster 0.12 0.30 0.04 0.02 Cirsium spp. Thistle 0.19 0.29 0.21 0.22 Erigeron pulchellus (Michx.) Daisy Fleabane Trace 0.00 0.12 0.00 Euthamia graminifolia (L.) Nutt. Goldenrod, Grass-leaved 1.12 2.03 0.90 0.57 Hieracium spp. Hawkweed 0.15 0.00 1.00 1.08 Leucanthemum vulgare Lam. Ox-eye Daisy 1.46 0.22 1.69 0.24 Packera aurea L. Golden Ragwort 0.00 0.05 0.14 0.00 Rudbeckia hirta L. Black-eyed Susan 0.01 0.02 0.07 0.08 Solidago bicolor L. Silverrod 0.00 0.01 0.00 0.05 Solidago rugosa (Mill.) Wrinkle-leaved Goldenrod 5.32 6.25 7.04 3.65 Solidago uliginosa (Nutt.) Bog Goldenrod 6.42 13.60 8.37 13.49 Taraxacum officinale (F.H. Wigg.) Dandelion 0.62 1.14 0.68 0.68 Tragopogon pratensis L. Yellow Goat’s Beard 0.04 0.00 0.19 0.00 Family Brassicaceae Brassica rapa L. Bird’s Rape 0.00 0.00 0.01 0.00 Family Caryophyllaceae Dianthus armeria L. Deptford Pink Trace 0.00 0.00 0.00 Stellaria graminea L. Lesser Stitchwort 0.02 0.00 0.22 0.03 Stellaria longifolia (Muhl.) Longleaf Stitchwort 0.00 0.00 0.00 0.00 Family Cyperaceae Carex spp. Sedge 1.07 2.39 0.71 2.26 Scirpus atrocinctus (Fernald) Blackgirdle Bulrush 0.00 0.00 0.01 0.10 Family Dennstaedtiaceae Pteridium aquilinum L. Bracken Fern 0.02 0.12 0.10 0.10 Family Ericaceae Vaccinium spp. Blueberry 1.09 1.14 1.68 8.87 Family Fabaceae Lotus corniculatus L. Bird’s-foot Trefoil 0.00 0.00 0.00 1.24 Medicago sativa L. Alfalfa Trace 0.00 0.00 0.54 Trifolium aureum Pollich Yellow Hop Clover 0.38 0.00 0.36 0.08 Trifolium pratense L. Red Clover 0.68 0.07 1.00 0.19 Trifolium repens L. White Clover 0.00 0.00 0.01 0.00 Family Gentianaceae Gentiana clausa Closed Gentian 0.00 0.00 0.03 0.00 Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 201 1999 2000 Scientific name Species H P H P Family Hypericaceae Hypericum densiflorum (Pursh) Bushy St. Johnswort 4.28 10.30 5.25 11.69 Hypericum ellipticum (Hook.) Elliptic-leaved St. Johnswort 0.00 0.07 0.00 0.05 Family Iridaceae Sisyrinchium spp. L. Blue-eyed Grass 0.01 0.32 0.00 0.11 Family Juncaceae Juncus effusus L. Common Rush 0.11 0.62 1.55 2.61 Family Labiatae Mentha spp. L. Mint 0.13 1.06 0.11 0.02 Prunella vulgaris L. Heal-all 0.10 0.07 0.29 5.19 Satureja vulgaris L. Field-basil 0.28 0.04 0.22 5.34 Stachys palustris L. Marsh Woundwort 0.00 0.00 0.00 2.79 Family Lycopodiaceae Lycopodium digitatum (Fernald) Fan Clubmoss 0.00 0.16 0.01 2.98 Lycopodium spp. Clubmoss 5.48 4.56 1.37 0.00 Family Onagraceae Oenothera perennis L. Small Sundrops 0.07 0.00 0.04 0.00 Family Oxalidaceae Oxalis stricta L. European Yellow Wood Sorrel 0.16 0.37 0.24 11.69 Family Plantaginaceae Plantago virginica L. Buck Plantain, Buck 0.15 0.48 0.12 0.05 Family Poaceae Agrostis giantea Roth Redtop 0.18 1.24 1.44 1.18 Anthoxanthum odoratum L. Sweet Vernal Grass 0.08 4.31 3.01 6.18 Dactylis glomerata L. Orchard Grass 6.35 1.62 6.81 5.19 Danthonia compressa Austin Flattened Oatgrass 0.44 1.69 6.94 5.34 Dichanthelium clandestinum Deertongue 0.00 0.02 0.25 0.19 (L.) Gould Holcus lanatus L. Velvetgrass 0.28 1.88 1.50 1.24 Lolium perenne L. Eastern Ryegrass 0.32 0.41 2.09 0.54 Phalaris arundinacea L. Reed Canary Grass 11.93 0.00 13.12 0.00 Phleum pratense L. Timothy 1.99 5.84 3.21 2.98 Poa palustris L. Fowl Bluegrass 2.40 0.02 0.00 0.00 Schedonorus arundinaceus Tall Fescue 0.00 0.00 1.36 2.79 (Schreb.) Bromus spp. Brome 0.20 0.20 0.00 0.00 Family Polygalaceae Polygala sanguinea L. Rose Polygala 0.00 0.00 0.14 0.00 Family Polygonaceae Rumex acetosella L. Field Sorrel, Sheep Sorrel 0.00 0.00 0.17 0.19 Family Ranunculaceae Ranunculus spp. Buttercup 0.17 0.38 1.41 1.41 Family Rosaceae Crataegus spp. Hawthorne 0.00 0.00 0.08 0.08 Fragaria spp. Wild strawberry 0.33 2.34 0.61 0.61 Potentilla spp. Cinquefoil 11.12 4.20 9.86 9.86 Southeastern Naturalist K.A. Chadbourne and J.T. Anderson 2015 Vol. 14, Special Issue 7 202 1999 2000 Scientific name Species H P H P Rubus spp. Dewberry 0.55 2.41 1.22 1.22 Rubus spp. Blackberry 0.00 0.00 0.00 0.00 Spiraea alba (Du Roi) Narrow-leaved Meadowsweet 1.90 7.56 2.59 2.59 Family Rubiaceae Galium mollugo L. White Bedstrawe 3.61 2.02 0.96 0.96 Houstonia caerulea L. Bluet 0.08 0.26 0.06 0.06 Family Salicaceae Salix sericea (Marshall) Silky Willow 0.43 0.04 1.01 1.01 Family Scrophulariaceae Linaria vulgaris (Mill.) Yellow Toadflax 0.00 0.00 0.00 0.00 Family Violaceae Viola sororia var. sororia Willd. Common Blue Violet 0.02 0.00 0.03 0.03