Influence of inundation potential and forest overstory on the ground-layer vegetation of Allegheny Plateau riparian forests.
Riparian zones, occurring at the interface of terrestrial and aquatic ecosystems, are among the most dynamic, diverse and productive of ecological systems (Gregory et al., 1991; Naiman et al., 1993). High diversity in the riparian flora is a function of the great horizontal heterogeneity usually inherent to these systems (Gregory et al., 1991). In riparian forests, for example, complex gradients that can include variation in elevation, flooding frequency and duration, substrate conditions, nutrients, light and other factors, help to create a mosaic of plant communities along the stream to forest gradient (Peterson and Rolfe, 1982; Menges and Waller, 1983; Menges, 1986; Nilsson et al., 1989). Productivity varies along this gradient, and is typically greater in riparian plant communities than in adjacent upland communities (Johnson and Bell, 1976; Peterson and Rolfe, 1982; Gregory et al., 1991). On a regional basis, riparian systems often support a distinct flora that may be uncommon or rare in the landscape, particularly in those heavily altered by humans (Gregory et al., 1991; Naiman et al., 1993; Bratton et al., 1994), and thus may have highly significant conservation value.
Headwater riparian habitat is abundant in the nonglaciated Allegheny Plateau of northwestern Pennsylvania. In the 200,000 ha Allegheny National Forest alone, riparian habitat flanks over 1600 km of headwater streams (Williams and Moriarity, 1997). Forests of the region have been well-studied in regard to Allegheny hardwood silviculture, but riparian forests, particularly ground-layer plant communities, have received little attention. Emerging information indicates that riparian forests contain some of the richest ground-layer plant communities of the region, and thus, may be important foci of vascular plant richness in Allegheny Plateau landscapes (Rubino, 1997; Williams and Moriarity, 1997; Hanlon et al., 1998; Walters and Williams, 1999; Williams et al., 1999). Given the documented importance of ground-layer vegetation in the structure and function of forest ecosystems (Siccama et al., 1970; Peterson and Rohlf, 1982), and its potential value for assessing site conditions and monitoring environmental change (Pregitzer and Barnes, 1982; Strong et al., 1991; Gilliam and Turrill, 1993; Gilliam et al., 1994; McClenahan and Long, 1995; Meier et al., 1995; Adkison and Jackson, 1996), information on ground-layer characteristics may provide key insight into the environment, dynamics and integrity of riparian forests in managed landscapes of the Allegheny Plateau.
The focus of this study was to examine variation in plant species richness, biomass and cover in the ground-layer (all vascular plants [less than or equal to]1 m tall) of headwater riparian forests of the Allegheny Plateau in northwestern Pennsylvania. Primary objectives were: (1) to compare riparian ground-layer species richness, biomass and cover across a gradient of potential inundation frequency; and (2) to examine the relationship of these ground-layer attributes with forest overstory characteristics. We chose inundation potential and forest overstory as study variables because of their documented effects on ground-layer communities across a range of forest types (Bratton, 1976; Hicks, 1980; Whitney and Foster, 1988; Gilliam and Turrill, 1993; Harmon and Franklin, 1995) and because virtually nothing was known of their influence on the ground-layer vegetation of Allegheny Plateau riparian forests. An additional objective was to identify ground-layer guilds that could serve as potential indicators of site inundation potential for riparian area delineation and management (e.g., Forested Wetlands Task Force, 1993).
Study area. - This study was conducted in the Allegheny National Forest (ANF) in northwestern Pennsylvania (41 [degrees] 45[minutes]N, 79 [degrees] 00[minutes]W). ANF lies in the nonglaciated Allegheny Plateau Physiographic Province, a landscape typified by relatively flat to gently rolling plateaus dissected by deep, u- or v-shaped, stream valleys (Hough and Forbes, 1943). Plateau elevations range from 500 to 700 m above sea level; stream valley floors generally range from 300 to 400 m in elevation (Cerutti, 1985; Kopas, 1985; Whitney, 1990). Mean annual temperatures for the region range from 6 C to 9.5 C, depending on aspect and exposure (Cerutti, 1985; Kopas, 1985). Precipitation is distributed fairly evenly throughout the year and averages between 100 and 110 cm (Cerutti, 1985; Kopas, 1985).
Sandstones and shales of Pennsylvanian age are the dominant parent material from which soils of the region are derived (Aguilar and Arnold, 1985; Whitney, 1990). Soils of well-drained to moderately well-drained sites are acidic, relatively infertile gray-brown to red-yellow podzolics of the Cookport (fragic aquidult), Gilpin (typic hapludult) and Hazelton (typic dystrochrept) series (Hough and Forbes, 1943; Aguilar and Arnold, 1985; Cerutti, 1985; Kopas, 1985; Whitney, 1990). In stream valley floodplains, soils may be colluvial or alluvial and of the Philo (fluvaquentic dystrochrepts) and Pope (fluvaventic dystrochrepts) series, or of mixed colluvial/alluvial origin (udifluvent complexes) (Lattman, 1960; Aguilar and Arnold, 1985; Cerruti, 1985; Kopas, 1985).
ANF is within the hemlock-northern hardwood forest type of Kuchler (1964) and the hemlock-white-pine-northern hardwood forest region of Braun (1950). Presettlement forests of the nonglaciated Allegheny Plateau were dominated by eastern hemlock (Tsuga canadensis) and American beech (Fagus grandifolia) on moister plateaus and stream valleys, and oak-chestnut forests (Quercus rubra, Q. prinus, Castanea dentata) on drier ridges and outcrops (Marquis, 1975; Aguilar and Arnold, 1985; Whitney, 1990; Abrams and Ruffner, 1995) (scientific nomenclature follows Gleason and Cronquist, 1991; common names follow Rhoads and Klein, 1993). The extensive logging that occurred from 1880 to 1930 produced the Allegheny hardwood forest type that presently covers much of the region (Marquis, 1975; Whitney, 1990). Dominant tree species include black cherry (Prunus serotina), red maple (Acer rubrum), sugar maple (A. sacccharum) and American beech (Fagus grandifolia) (Marquis, 1975; Whitney, 1990; Abrams and Ruffner, 1995).
Study sites and geomorphic surface classification. - Six riparian sites in ANF were selected for study (Table 1). Sites were chosen to represent a range of geomorphic surfaces, inundation potential and forest overstory characteristics (Table 1). Stream order among sites ranged from second to fourth and active channel width ranged from 7 m to 15 m. Chief criteria for selecting study sites were lack of recent timber management or other major human disturbance, absence of recent canopy defoliation by insects and a minimum stand age of at least 40 y.
To facilitate comparisons across riparian study sites, we categorized habitats within systems with a site classification that considered inundation potential, geomorphic surface type and forest cover type (Table 2). Three inundation classes were recognized: (1) high inundation potential - streamside geomorphic surfaces with relatively saturated soils subject to fluctuating water levels and inundation at recurring intervals due to seasonal (e.g., March to May spring thaw) or storm-related stream pulses; (2) moderate inundation potential - floodplain geomorphic surfaces that may experience periodic inundation due to seasonal high flows but which generally lack saturated soils for extended periods; and (3) low inundation potential-hillslope geomorphic surfaces that are beyond the influence of normal high stream flows. Three forest cover types were distinguished: (1) deciduous forest - sites at which deciduous trees comprise over 70% of the overstory basal area; (2) coniferous forest - sites at which coniferous trees, primarily Tsuga canadensis, comprise over 70% of the overstory basal area; and (3) mixed forest - sites at which the overstory is comprised of both deciduous and coniferous species with similar basal area. Our forest cover types are similar to basal area criteria currently used by ANF (Allegheny National Forest, 1986).
Classification of geomorphic surfaces generally follows Osterkamp and Hupp (1984), Harris (1987) and Gregory et al. (1991) with the addition of savanna wet depressions. Savannas, dominated by a sparse canopy of scattered Prunus serotina and/or Acer rubrum, develop sporadically in stream valleys, at the heads of drainage basins and on upland plateaus of the region. Primarily anthropic in origin, savannas arose as a result of tree regeneration failures that followed the extensive logging and hot slash fires of the turn of the century (Hough, 1945, 1949; Horsley, 1985). The floodplain savanna we studied at Salmon [TABULAR DATA FOR TABLE 1 OMITTED] [TABULAR DATA FOR TABLE 2 OMITTED] Creek was typified by an undulating microtopography in which xeric ridges alternate with depressions that are saturated for much of the growing season (Williams et al., 1999), similar to the hummocky floodplain category of Harris (1987).
Field sampling. - An array of 10 to 20 permanent plots was established at each study site during 1993 to 1994. Plot size varied according to area of each geomorphic surface sampled (10 x 10 m plots for floodplain and mesic hillslope plots; 1 x 1 m or 2 x 2 m plots for smaller, constrained surfaces such as depositional bars and savanna wet depressions). Within each geomorphic surface 7 plots were selected at random for sampling. Woody plant stems [greater than or equal to]2.5 cm dbh were measured in each 100 [m.sup.2] plot on forested floodplain and hillslope geomorphic surfaces. Woody vegetation in the Salmon Creek floodplain savanna was sampled using the point-quarter method at 30 random points along three 200 m transects.
Field sampling of the ground-layer was conducted during a two wk period from 2 to 16 August 1994, when the summer riparian vegetation in the study area is generally at its peak development. We recognize that vernal herbs can produce an early season peak in biomass in some Allegheny Plateau forests, and our study time frame would have excluded them from sampling. However, vernal herbs are sporadically distributed in Allegheny Plateau riparian forests, particularly among geomorphic surfaces of high inundation potential (C. E. Williams and W. J. Moriarity, pers. obs.), and thus provide less opportunity to document the influence of inundation potential than summer herbs. Ground-layer vegetation was sampled in 0.25 [m.sup.2] quadrats, one quadrat placed randomly within each of the seven randomly selected permanent plots per geomorphic surface. Quadrats of this size are widely used in studies of ground-layer biomass and species richness (e.g., Walker and Peet, 1983; Moore and Keddy, 1989). Within each quadrat, total ground cover was estimated visually to the nearest 5%, and the presence of each vascular plant species was recorded. A spherical densiometer was used to estimate percent forest overstory cover at the center of each permanent plot (Lemmon, 1956). All vascular plants [less than or equal to]1 m in height were clipped to ground level, sorted by growth form (ferns, forbs, graminoids and woody plants) and bagged. Ground-layer harvests were returned to the laboratory, oven-dried at 80C for 48 to 72 h and weighed to the nearest 0.1 gram. Voucher specimens are deposited in the Clarion University herbarium.
Data analysis. - To provide a compositional summary of the forest overstory at each study site, importance values (the sum of relative basal area + relative density + relative frequency/3) were calculated for all overstory species (Curtis and McIntosh, 1951). Depositional bar sites were influenced by the canopy of adjacent forest but were nonforested, thus overstory composition data are not presented for these surfaces. Ground-layer species composition was summarized by relative frequency of occurrence of plant species in sample plots across study sites and geomorphic surfaces. Percent similarity in ground-layer species composition among inundation classes was determined with Sorensen's index using presence-absence data (Krebs, 1989). Wetland indicator species were determined using Reed (1988; see Table 2 for a list of wetland indicator species classes and their abbreviations). The G-test (P [less than or equal to] 0.05) was used to test the null hypothesis that the proportion of wetland and upland indicator ground-layer species [total wetland species = sum of obligate wetland species (OBL) + facultative wetland species (FACW, FACW+, FACW-); total upland species = sum of facultative species (FAC, FAC+, FAC-) + facultative upland species (FACU, FACU+, FACU-) + upland species (UPL)] did not differ across inundation classes (Zar, 1984).
It is not possible to replicate our study sites for statistical analysis (e.g., Hurlbert, 1984). However, since we were interested in inundation as an integrating variable in this study, we chose to consider inundation potential as the main variable for analysis and used sites within inundation classes as replicates. Thus, each inundation class was represented by four replicates (sites) with seven observations (plots) per replicate. One-way analysis of variance (ANOVA) was used to determine significant differences (P [less than or equal to] 0.05) in ground-layer characteristics across inundation classes using sample means pooled from respective geomorphic surfaces in each class. When ANOVAs were significant, the Tukey multiple range test was used to separate means. The Tukey test was chosen because it is generally robust in regard to departures from normality and homogeneity of variance and is conservative in declaring differences in means (Chew, 1976; Zar, 1984). Within each inundation class, the relationship of ground-layer species richness, cover and total biomass to forest overstory characteristics (percent overstory cover above permanent plots; overstory basal area and stem density, and basal area of Tsuga canadensis per each 100 [m.sup.2] plot from which ground-layer samples were taken) was assessed with Pearson product-moment correlations (P [less than or equal to] 0.05) using individual plot values (Zar, 1984). The Bonferroni method was used to adjust probabilities associated with matrix correlations (Wilkinson, 1997). Data were transformed [square root (x + 0.5) for count data; arcsin transformation for percentage data; log (x + 0.5) for other continuous data] before analysis to stabilize variances (Zar, 1984). Statistical analyses were conducted using SYSTAT version 7.0 (Wilkinson, 1997).
Vegetation composition. - Fourteen tree species were recorded from the overstory of mesic floodplain and hillslope forests (Table 2). based on importance values, six species were dominant across sites: Tsuga canadensis, Betula alleghaniensis, Acer rubrum, A. saccharum, Fagus grandifolia and Prunus serotina. None of the overstory tree species recorded is a wetland species (Reed, 1988). Overstory basal area ranged from 7.1 [m.sup.2]/ha for the floodplain savanna at Salmon Creek to 50.4 [m.sup.2]/ha for the hillslope coniferous forest at Coon Run (Table 2).
Eighty-one species were recorded from the ground-layers of the six riparian study sites. Forbs (37 species; 45.7% of total species) were dominant followed by graminoids (22 species; 27.2%), woody plants (16 species; 19.8%) and ferns (6 species; 7.4%). Similarity in ground-layer species composition ranged from 29.9% between high and moderate inundation classes, 36.0% between high and low inundation classes and 57.1% between moderate and low inundation classes. Wetland and upland ground-layer species were not distributed independently across inundation classes (G = 28.075, d.f. = 2, P [less than or equal to] 0.0001; Table 3). Wetland species occurred most often on geomorphic surfaces of high inundation potential (66.7% of total species for which indicator status has been determined; Table 3). In contrast, upland plant species, primarily tree seedlings, occurred principally on geomorphic surfaces of moderate and low inundation potential (81.5% and 74.3%, respectively, of total species for which indicator status has been determined; Table 3). Wetland plant species of common occurrence on geomorphic surfaces of high inundation potential included Glyceria melicaria, Galium asprellum, Impatiens capensis, Polygonum sagittatum, Aster umbellatus and Chelone glabra (Table 3). Upland species common on geomorphic surfaces of moderate inundation potential included: Thelypteris noveboracensis, Brachyelytrum erectum, Mitchella repens, Oxalis acetosella, Acer rubrum and Fagus grandifolia. Thelypteris noveboracensis, Dennstaedtia punctilobula, O. acetosella, A. rubrum and Maianthemum canadense were among the most widespread upland species of the ground-layer of low inundation potential geomorphic surfaces (Table 3).
Variation in ground cover and overstory cover. - Percent ground cover was significantly greater for geomorphic surfaces of high inundation potential but did not differ between those of moderate and low inundation potential (Table 4). Ground cover ranged from a high of 92.1% for savanna wet depressions at Salmon Creek to a low of 3.4% for the floodplain coniferous forest at Irwin Run (Table 3). Percent overstory cover was significantly lower for geomorphic surfaces of high inundation potential but did not differ between those of moderate and low inundation potential (Table 4). Overstory cover ranged from a low of 31.9% in the floodplain savanna at Salmon Creek to a high of 93.1% for the mesic hillslope coniferous forest at Coon Run (Table 3).
Variation in species richness. - Total species richness (number of species/0.25 [m.sup.2]) was significantly greater for geomorphic surfaces of high inundation potential but did not differ between those of moderate and low inundation potential (Table 4, [ILLUSTRATION FOR FIGURE 1 OMITTED]). Species richness of forb and graminoid species was significantly greater in the high inundation class. Fern and woody plant species richness did not differ significantly among inundation classes (Table 4, [ILLUSTRATION FOR FIGURE 1 OMITTED]). Ground-layer species richness varied across sites and geomorphic surfaces [ILLUSTRATION FOR FIGURE 1 OMITTED]. Total species richness was greatest in the depositional bar geomorphic surface at Salmon Creek (10.3 +- 0.6 SE species/0.25 [m.sup.2]) and lowest in the floodplain coniferous forest at Irwin Run (0.7 + 0.6 SE species/0.25 [m.sup.2]). Species richness of graminoids and forbs was strongly skewed toward geomorphic surfaces of high inundation potential [ILLUSTRATION FOR FIGURE 1 OMITTED].
Variation in biomass. - As with total species richness, total biomass (gm. dry wt./0.25[m.sup.2]) [TABULAR DATA FOR TABLE 3 OMITTED] [TABULAR DATA FOR TABLE 4 OMITTED] was significantly greater for geomorphic surfaces of high inundation potential but did not differ between those of moderate and low inundation potential (Table 4). Forb biomass and graminoid biomass were significantly greater in the high inundation class, whereas fern biomass and woody plant biomass did not differ significantly across inundation classes (Table 4).
Ground-layer biomass varied considerably across sites and geomorphic surfaces (Table 4, [ILLUSTRATION FOR FIGURE 1 OMITTED]). Total biomass was significantly greater for geomorphic surfaces of high inundation potential than for those of moderate and low inundation potential (Table 4, [ILLUSTRATION FOR FIGURE 1 OMITTED]). Total [TABULAR DATA FOR TABLE 5 OMITTED] biomass was greatest for savanna wet depression and depositional bar surfaces. Forb biomass was greatest in the savanna wet depression surface and was 2 to 10 times higher than the forb biomass of other geomorphic surfaces [ILLUSTRATION FOR FIGURE 1 OMITTED]. Graminoid biomass was greatest for savanna wet depression and depositional bar geomorphic surfaces but was negligible for all others [ILLUSTRATION FOR FIGURE 1 OMITTED].
Ground-layer relationships to overstory characteristics. - Forest overstory had no significant association with ground-layer characteristics within inundation classes (Table 5). Although the majority of correlations between ground-layer and overstory characteristics were strongly non-significant, the negative correlations between overstory cover and total biomass for high inundation potential sites (r = -0.51; P = 0.06) and overstory stem density and total biomass for moderate inundation sites (r = -0.53; P = 0.07) were only marginally nonsignificant based on p-values close to 0.05 (e.g., Motulsky, 1995).
Inundation potential has a significant influence on the structure and composition of ground-layer vegetation in headwater riparian forests of the Allegheny Plateau. Geomorphic surfaces of high inundation potential support greater ground-layer species richness, biomass and cover and a relatively distinct wetland flora compared to mesic floodplains and hillslopes of moderate inundation potential. High species richness and productivity in riparian plant communities have been correlated with pulsed hydroperiod (Johnson and Bell, 1976; Mitsch et al., 1991), substrate heterogeneity (Nilsson et al., 1989) and available light (Menges, 1986). It is likely that these factors determine in large part the mosaic nature of the ground-layer in Allegheny Plateau riparian forests, in particular, the distinctiveness of the flora of high inundation potential geomorphic surfaces (e.g., Huston 1979).
Many of the ground-layer characteristics of forested geomorphic surfaces of moderate to low inundation potential quantified during this study, particularly cover and biomass, are comparable to similar data collected for upland deciduous and coniferous forests in the eastern United States (Gilliam and Turrill, 1993 and references therein). However, biomass, cover and to a lesser extent species richness, are strikingly higher for high inundation potential depositional bar and savanna wet depression surfaces than for most regional forests, and are more comparable to values recorded for open graminoid wetland and wet savanna vegetation types (e.g., Wheeler and Giller, 1982; Walker and Peet, 1983). This dichotomy in biomass, cover and species richness suggests the presence of two broad ecological groupings in the ground-layer of Allegheny Plateau headwater riparian forests: a fairly rich and productive, but shade-intolerant, graminoid and forb-dominated wetland flora; and a relatively species-poor, less productive, upland forest flora dominated by tree seedlings, ferns and relatively shade-tolerant forbs.
Bratton (1976) and Hicks (1980) classified ground-layer species in southern Appalachian forests into guilds according to their patterns of resource partitioning and occurrence along a range of soil moisture and forest overstory conditions. A guild concept can also be applied to our data with respect to the ground-layer flora of different inundation classes. For example, the ground-layer guild of geomorphic surfaces of high inundation potential, typified by relatively high species richness, biomass, cover and the general shade-intolerance of the flora (especially the graminoids), would include dominant wetland species such as Glyceria melicaria, Galium asprellum, Impatiens capensis, Oxalis stricta, Aster umbellatus, Chelone glabra and Polygonum sagittatum, among others (Table 3). The ground-layer guild of geomorphic surfaces of moderate inundation potential, typified by relatively low to modest species richness, biomass and cover, and dominated by moderately shade-tolerant mesic, upland species, would include Thelypteris noveboracensis, Mitchella repens, Oxalis acetosella, Acer rubrum and occasional wetland species such as Arisaema triphyllum, Chelone glabra and Glyceria melicaria. The ground-layer guild of geomorphic surfaces of low inundation potential, having low species richness, biomass and cover, would be represented by shade-tolerant upland species including Oxalis acetosella, Dryopteris intermedia and Maianthemum canadense. Thus, the ground-layer vegetation of Allegheny Plateau riparian forests can be viewed as a series of wetland and upland species guilds that replace each other along the stream to mesic forest gradient in response to soil moisture and forest overstory conditions.
In contrast to studies of riparian forests associated with larger river systems (e.g., Pysek and Prach, 1993; Pyle, 1995), nonnative plant species were of relatively low importance in the ground-layer of the headwater riparian forests we studied. We recorded only five non-native species (6.2% of the total flora) from our study sites including the graminoids Agrostis gigantea and Microstegium vimineum, and the forbs Myosotis scorpioides, Rumex obtusifolius and Stellaria graminea. Of these five species, only the shade-tolerant M. vimineum is an invasive pest of riparian zones and mesic forests (Barden, 1987). The establishment and maintenance of nonnative species in riparian forests, particularly those species that are shade-intolerant, are often facilitated by a combination of human disturbance, such as recreational impacts and forest fragmentation (Pyle, 1995), and natural flooding and scouring events that create open colonization sites (Pysek and Prach, 1993). The low frequency of nonnative species that we observed is not unusual for the ground-layer of headwater riparian forests of the Allegheny Plateau (e.g., Rubino, 1997; Hanlon et al., 1998; Williams et al., 1999), and may be due to lack of suitable disturbance for nonnative plant establishment and recruitment, or to regionally low sources of nonnative species adapted to riparian zones and/or closed canopy mesic forest.
Heavy browsing by white-tailed deer (Odocoileus virginianus) has been a stress for forest communities in northwestern Pennsylvania for over 60 y (Marquis, 1975; Tilghman, 1989; Rooney and Dress, 1997a). Intense deer browsing can decrease ground-layer species richness, biomass and cover in forests (e.g. Alverson et al., 1988). At Heart's Content in northwestern Pennsylvania, Rooney and Dress (1907b) found a dramatic decline in the species richness and abundance of the ground-layer vegetation after 60 y of deer browsing, including the near loss of species such as Viburnum alnifolium (hobble-bush), a once widespread shrub of riparian zones and mesic uplands. It is possible that some of the differences in ground-layer characteristics that we recorded during our study may have been due in part to past and/or present deer browsing. For example, the lack of significant differences in ground-layer characteristics between sites of moderate and low inundation potential may have been influenced by a similar history of browsing that caused a convergence in species richness, biomass and cover relative to high inundation potential sites. However, myriad other biotic and abiotic factors are involved in the development of plant communities, particularly in riparian habitats (e.g., Gregory et al., 1991). Thus to attribute differences in riparian ground-layer characteristics primarily to deer browsing probably oversimplifies a more complicated set of factors that determines pattern and process in these dynamic communities.
Forest management activities in and around headwater riparian forests of the Allegheny Plateau could potentially alter the structure and composition of ground-layer plant communities. Stand thinning, a standard silvicultural prescription in mid-aged Allegheny hardwood stands (Horsley, 1977), may cause divergent responses in ground-layer species richness, biomass and cover according to site inundation potential. For example, by increasing light levels, thinning of mesic floodplain and hillslope forests could increase the density, biomass and cover of the ferns Thelypteris noveboracensis and Dennstaedtia punctilobula and the grass Brachyelytrum erectum at the expense of a more diverse ground-layer flora (Horsley 1977), a generally undesirable condition when managing for system diversity. In contrast, stand thinning adjacent to geomorphic surfaces of high inundation potential, sites where invasion of facultative ferns and grasses is generally less problematic, may increase the abundance of the shade-intolerant, graminoid-dominated flora. Depending on ecological and/or aesthetic management goals (e.g, Reader and Bricker, 1992), it is possible that small-scale low-impact thinning which mimics natural gap processes could be used as a management tool for enhancing the diversity and mosaic nature of the ground-layer in Allegheny Plateau riparian forests.
Acknowledgments. - Anna Eberle and April Moore assisted with vegetation sampling. John Knox, Eric Menges, Kim Williams and an anonymous referee provided helpful comments on drafts of the manuscript. Thomas Wieboldt, Associate Curator, Virginia Tech Herbarium, verified or identified difficult plant taxa. The interlibrary loan staff at Clarion University provided access to key references. This study was supported in part by a participating agreement between the USDA Forest Service, Allegheny National Forest, and Clarion University of Pennsylvania.
ABRAMS, M. D. AND C. M. RUFFNER. 1995. Physiographic analysis of witness tree distribution (1765-1798) and present forest cover through north central Pennsylvania. Can. J. For. Res., 25:659-668.
ADKISON, G. P. AND M. T. JACKSON. 1996. Changes in ground-layer vegetation near trails in midwestern U.S. forests. Nat. Areas J., 16:14-23.
AGUILAR, R. AND R. W. ARNOLD. 1985. Soil-landscape relationships of a climax forest in the Allegheny High Plateau, Pennsylvania. Soil. Sci. Am. J., 49:695-701.
ALLEGHENY NATIONAL FOREST. 1986. Land and resource management plan. USDA Forest Service. Warren, PA. 224 p.
ALVERSON, W., D. S. WALLER AND S. L. SOLHEIM. 1988. Forests too deer: edge effects in northern Wisconsin. Conserv. Biol., 2:348-358.
BARDEN, L. S. 1987. Invasion of Microstegium vimineum (Poaceae), an exotic, annual, shade-tolerant, [C.sub.4] grass, into a North Carolina floodplain. Am. Midl. Nat., 118:40-45.
BRATTON, S. P. 1976. Resource division in an understory herb community:. responses to temporal and microtopographic gradients. Am. Nat., 110:679-693.
-----, J. R. HAPEMAN AND A. R. MAST. 1994. The lower Susquehanna River gorge and floodplain (U.S.A.) as a riparian refugium for vernal, forest floor herbs Conserv. Biol., 8:1069-1077.
BRAUN, E. L. 1950. Deciduous forests of eastern North America. The Blakiston Co. Philadelphia, PA. 596 p.
CERUTTI, J. R. 1985. Soil survey of Warren and Forest Counties, Pennsylvania. USDA Soil Conservation Service. 148 p.
CHEW, V. 1976. Comparing treatment means: a compendium. HortScience, 11:348-357.
CURTIS, J. T. AND R. P. MCINTOSH. 1951. An upland forest continuum in the prairie-forest border region of Wisconsin Ecology, 32:476-496.
FORESTED WETLANDS TASK FORCE. 1993. Best management practices for silvicultural activities in Pennsylvania's forested wetlands. Pennsylvania State University, School of Forest Resources and Pennsylvania Hardwoods Development Council. State College, PA. 56 p.
GILLIAM, F. S. AND N. L. TURRILL. 1993. Herbaceous layer cover and biomass in a young versus a mature stand of a central Appalachian hardwood forest. Bull. Torrey Bot. Club, 120:445-450.
-----, -----, S. D. AULICK, D. K. EVANS AND M. B. ADAMS. 1994. Herbaceous layer and soil response to experimental acidification in a central Appalachian hardwood forest J. Environ. Qual., 23:835-844.
GLEASON, H. A. AND A. CRONQUIST. 1991. Manual of vascular plants of northeastern United States and adjacent Canada. 2nd ed. The New York Botanical Garden. Bronx, NY. 910 p.
GREGORY, S. V., F. J. SWANSON, W. A. McKEE AND K. W. CUMMINS. 1991. An ecosystem perspective of riparian zones. BioScience, 41:540-551.
HANLON, T. J., C. E. WILLIAMS AND W. J. MORIARITY. 1998. Species composition of soil seed banks of Allegheny Plateau riparian forests. J. Torrey Bot. Soc., 125:199-215.
HARMON, J. M. AND J. F. FRANKLIN. 1995. Seed rain and seed bank of third- and fifth-order streams on the western slope of the Cascade Range. Res. Pap. PNW-RP-480. Portland, OR. USDA Forest Service, Pacific Northwest Research Station. 27 p.
HARRIS, R. R. 1987. Occurrence of vegetation on geomorphic surfaces in the active floodplain of a California alluvial stream. Am. Midl. Nat., 118:393-405.
HICKS, D. J. 1980. Intrastand distribution patterns of southern Appalachian cove forest herbaceous species. Am. Midl. Nat., 104:209-223.
HOUGH, A. F. 1945. Frost pocket and other microclimates in forests of the northern Allegheny Plateau. Ecology, 26:235-250.
-----. 1949. Deer and rabbit browsing and available winter forage in Allegheny hardwood forests. J. Wildl. Manage., 13:135-141.
----- AND R. D. FORBES. 1943. The ecology and silvics of forests in the high plateaus of Pennsylvania. Ecol. Monogr., 13:301-320.
HORSLEY, S. B. 1977. Allelopathic inhibition of black cherry. II. Inhibition by woodland grass, ferns and club moss. Can. J. For. Res., 7:515-519.
-----. 1985. Reforestation of orchard stands and savannas on Pennsylvania's Allegheny Plateau. Northern J. Appl. For., 2:22-26.
HUSTON, M. 1979. A general hypothesis of species diversity. Am. Nat., 113:81-101.
HURLBERT, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecol. Monogr., 54:187-211.
JOHNSON, F. L. AND D. T. BELL. 1976. Plant biomass and net primary production along a flood-frequency gradient in the streamside forest. Castanea, 43:150-165.
KOPAS, F. A. 1985. Soil survey of Cameron and Elk Counties, Pennsylvania. USDA Soil Conservation Service. 132 p.
KREBS, C. J. 1989. Ecological methodology. Harper and Row Publ. New York. 654 p.
KUCHLER, A. W. 1964. Potential natural vegetation of the conterminous United States. Am. Geogr. Soc. Spec. Publ. 36. 156 p.
LATTMAN, L. H. 1960. Gross section of a floodplain in a moist region of moderate relief. J. Sediment. Petrol., 30:275-282.
LEMMON, P. E. 1956. A spherical densiometer for estimating forest overstory density. For. Sci., 2:314-320.
MARQUIS, D. A. 1975. The Allegheny hardwood forests of Pennsylvania. USDA Forest Service Gen. Tech. Rep. NE-15. Northeastern Forest Exp. Stn., Upper Darby, PA. 32 p.
MCCLENAHAN, J. R. AND R. P. LONG. 1995. Variability in oak forest herb layer communities, p. 60-78. Proceedings of the 10th Central Hardwood Forest Conference, 5-8 March 1995, Morgantown, WV. USDA Forest Service Gen. Tech. Rep. NE-197.
MEIER, A. J., S. P. BRATTON AND D. C. DUFFY. 1995. Possible ecological mechanisms for loss of vernal-herb diversity in logged eastern deciduous forest. Ecol. Appl., 5:935-946.
MENGES, E. S. 1986. Environmental correlates of herb species composition in five southern Wisconsin floodplain forests. Am. Midl. Nat., 115:106-117.
----- AND D. M. WALLER. 1983. Plant strategies in relation to elevation and light in floodplain herbs. Am. Nat., 122:454-473.
MITSCH, W. J., J. R. TAYLOR AND K. B. BENSON. 1991. Estimating primary productivity of forested wetland communities in different hydrologic landscapes. Landscape Ecology, 5:75-92.
MOORE, D. R. J. AND P. A. KEDDY. 1989. The relationship between species richness and standing crop in wetlands: the importance of scale. Vegetatio, 79:99-106.
MOTULSKY, H. 1995. Intuitive biostatistics. Oxford University Press. New York. 386 p.
NAIMAN, R. J., H. DECAMPS AND M. POLLOCK. 1993. The role of riparian corridors in maintaining regional biodiversity. Ecol. Appl., 3:209-212.
NILSSON, C., G. GRELSSON, M. JOHANSSON AND U. SPERENS. 1989. Patterns of plant species richness along riverbanks. Ecology, 70:77-84.
OSTERKAMP, W. R. AND C. R. HUPP. 1984. Geomorphic and vegetative characteristics along three northern Virginia streams. Geol. Soc. Amer. Bull., 95:1093-1101.
PETERSON, D. L. AND G. L. ROLFE. 1982. Nutrient dynamics of herbaceous vegetation in upland and floodplain forest communities. Am. Midl. Nat., 107:325-339.
PREGITZER, K. S. AND B. V. BARNES. 1982. The use of ground flora to indicate edaphic factors in upland ecosystems of the McCormick Experimental Forest, Upper Michigan. Can. J. For. Res., 12:661-672.
PYLE, L. L. 1995. Effects of disturbance on herbaceous exotic plant species on the floodplain of the Potomac River. Am. Midl. Nat., 134:244-253.
PYSEK, P. AND K. PRACH. 1993. Plant invasions and the role of riparian habitats: a comparison of four species alien to central Europe. J. Biogeogr., 20:413-420.
READER, R. J. AND B. D. BRICKER. 1992. Value of selectively cut deciduous forest for understory herb conservation: an experimental assessment. For. Ecol. Manage., 51:317-327.
REED, P. B. 1988. National list of plant species that occur in wetlands: National summary. United States Fish and Wildlife Service Biological Report 88. 244 p.
RHOADS, A. F. AND W. M. KLEIN, JR. 1993. The vascular flora of Pennsylvania: annotated checklist and atlas. Mem. Amer. Philos. Soc. Vol. 207. Philadelphia, PA. 636 p.
ROONEY, T. P. AND W. J. DRESS. 1997a. Patterns of plant diversity in overbrowsed primary and mature secondary hemlock-northern hardwood forest stands. J. Torrey Bot. Soc., 124:43-51.
----- AND -----. 1997b. Species loss over sixty-six years in the ground-layer vegetation of Heart's Content, an old-growth forest in Pennsylvania, USA. Nat. Areas J., 17:297-305.
RUBINO, D. L. 1997. Impacts of power and petroleum transmission corridors on riparian vegetation in northwestern Pennsylvania. M.S. thesis. Clarion University of Pennsylvania. Clarion. 63 p.
SICCAMA, T. G., F. H. BORMANN AND G. E. LIKENS. 1970. The Hubbard Brook ecosystem study: productivity, nutrients, and phytosociology of the herbaceous layer. Ecol. Monogr., 40:389-402.
STRONG, W. L., D. J. BLUTH, G. H. LAROI AND I. G. W. CORNS. 1991. Forest understory plants as predictors of lodgepole pine and white spruce site quality in west-central Alberta. Can. J. For. Res., 21:1675-1683.
TILGHMAN, N. G. 1989. Impacts of white-tailed deer on forest regeneration in northwestern Pennsylvania. J. Wildl. Manage., 53:524-532.
WALKER, J. AND R. K. PEET. 1983. Composition and species diversity of pine-wiregrass savannas of the Green Swamp, North Carolina. Vegetatio, 55:163-179.
WALTERS, G. L. AND C. E. WILLIAMS. 1999. Riparian forest overstory and herbaceous layer of two upper Allegheny River Islands in northwestern Pennsylvania. Castanea, in press.
WHEELER, B. D. AND K. E. GILLER. 1982. Species richness of herbaceous fen vegetation in Broadland Norfolk, in relation to the quantity of above-ground plant material. J. Ecol., 70:179-200.
WHITNEY, G. G. 1990. The history and status of the hemlock-hardwood forests of the Allegheny Plateau. J. Ecol., 78:443-458.
----- AND D. R. FOSTER. 1988. Overstorey composition and age as determinants of the understorey flora of woods of central New England. J. Ecol., 76:867-876.
WILKINSON, L. 1997. SYSTAT 7.0. New statistics. SPSS, Inc. Chicago, IL.
WILLIAMS, C. E. AND W. J. MORIARITY. 1997. Riparian research and management on the Allegheny National Forest. Pennsylvania For., 88:6-8.
-----, ----- AND G. L. WALTERS. 1999. Overstory and herbaceous layer of a riparian savanna in northwestern Pennsylvania. Castanea, in press.
ZAR, J. H. 1984. Biostatistical analysis. 2nd ed. Prentice-Hall, Inc. Englewood Cliffs, NJ. 718 p.
|Printer friendly Cite/link Email Feedback|
|Author:||Williams, Charles E.; Moriarity, William J.; Walters, Gary L.; Hill, Lori|
|Publication:||The American Midland Naturalist|
|Date:||Apr 1, 1999|
|Previous Article:||Effects of larval food-limitation on Vanessa cardui linnaeus (Lepidoptera: nymphalidae).|
|Next Article:||Effects of habitat area on the occurrence of grassland birds in Illinois.|