Xeric limestone prairies of eastern United States: review and synthesis.
Abstract Introduction Physiography Bedrock Geology Soils Climate Vegetation Ozark Plateaus Central Lowland Interior Low Plateaus Ridge and Valley Appalachian Plateaus Coastal Plain Endemic Taxa and Other Noteworthy Floristic Elements Origin Maintenance Physical Environment Factors and Anthropogenic Influence Conceptual Model Affinities of Xeric Limestone Prairies to Other Vegetation Types Eastern North America Western North America Direction for Future Research Literature Cited
Xeric limestone prairies (XLPs) (sensu Baskin et al., 1994; Baskin & Baskin, 2000) are open, nonforested areas in which herbaceous plant communities are developed on shallow, rocky soils derived from calcareous substrates including limestone, dolomite, and shale (Lawless et al., 2004) (Fig. 1). This vegetation type is characterized by dominance of [C.sub.4] perennial grasses (Baskin & Baskin, 2000; Laughlin & Uhl, 2003; Baskin & Baskin, 2004; Lawless et al., in press) and relatively high taxonomic richness of heliophytic [C.sub.3] perennial forbs, many of which are rare at the state, regional, and/or global levels (Skinner et al., 1983; White et al., 1994; Laughlin & Uhl, 2003; Lawless et al., 2004). XLPs often contain islands of woody vegetation within the grassland matrix and typically are surrounded by dry, rocky upland forests. Expansion of islands of woody vegetation and encroachment of adjacent forests into XLPs has been documented in some regions and can result in drastic reductions in the size of openings (e.g., Kimmel & Probasco, 1980; Annala & Kaputska, 1983; Laughlin, 2004).
[FIGURE 1 OMITTED]
This vegetation type is broadly distributed throughout a large portion of the eastern deciduous forest formation (sensu Braun, 1950) (Fig. 2), where it has been referred to as "glades" (Missouri, Arkansas, Illinois, Indiana, Kentucky, West Virginia, Tennessee, Georgia, and Alabama), "barrens" (Illinois, Indiana, Ohio, West Virginia, Virginia, Kentucky, Tennessee, Georgia, and Alabama), "prairies" (Alabama and Ohio), "prairie barrens" (Alabama), and "xeric limestone prairies" (Pennsylvania and Kentucky). Inconsistency in terminology used to describe XLPs throughout their broad geographic range has resulted in failure to recognize similarities among sites (Ludwig, 1999) and in confusion of these herbaceous communities with limestone cedar glades (Baskin et al., 1994; Baskin & Baskin, 1999, 2004), an herbaceous edaphic climax community in north-central southeastern United States characteristically dominated by [C.sub.4] summer annual grasses (sensu Baskin & Baskin, 1999, 2003).
[FIGURE 2 OMITTED]
Although many individual studies have been conducted on the flora, vegetation, and physical environment of XLPs in various states and physiographic regions, heretofore no attempt has been made to synthesize information on this vegetation type throughout eastern United States. Thus, the purpose of this paper is to synthesize data on the geographic/physiographic distribution, geology, soils, climate, vegetation, and endemic and other noteworthy plant taxa of XLPs in eastern United States and to provide a conceptual model of the origin, maintenance, and successional trajectories of this vegetation type. In addition, affinities of XLPs to other nonforested herbaceous vegetation types in North America are discussed. Plant nomenclature follows United States Department of Agriculture, Natural Resources Conservation Service (USDA, NRCS, 2004).
Physiographic provinces are in accordance with Fenneman (1938, 1946 ) except for the Interior Low Plateaus, which follows Quarterman and Powell (1978). Lower tiers in the hierarchy of land-form classification are in accordance with various sources (cited in Table I), and physiographic regions (sections and subsections) of the Interior Low Plateaus follow Quarterman and Powell (1978).
XLPs occur in six physiographic provinces in eastern United States: Ozark Plateaus, Central Lowland, Interior Low Plateaus, Appalachian Plateaus, Ridge and Valley, and Coastal Plain (Fig. 2; Table I). In the Ozark Plateaus, this vegetation type is distributed broadly on the Springfield and Salem plateaus and extends south into the Boston Mountains of northern Arkansas (Keeland, 1978; Kurz, 1981; Ladd & Nelson, 1982; Nelson & Ladd, 1982a, 1983; Heikens, 1991). Sites in the Central Lowland occur on the Osage Plains of west-central Missouri, on the Eastern Glaciated Plains and Lincoln Hills of east-central Missouri, and the Muscatatuck Flats and Canyons and Switzerland Hills of southeastern Indiana (Ladd & Nelson, 1982; Homoya, 1987; Maxwell, 1987). In the Interior Low Plateaus, XLPs range from the Mammoth Cave Plateau in southern Indiana to the Northeastern Blue Grass in southeastern Ohio and south to the Moulton Valley in northern Alabama (Braun, 1928; Bacone et al., 1983; Webb et al., 1997). This vegetation type is distributed over its maximum latitudinal range in the Ridge and Valley, where it extends from north-central Pennsylvania to central Alabama (DeSelm, 1993; Allison & Stevens, 2001; Laughlin, 2002). The distribution of XLPs in the Coastal Plain and Appalachian Plateaus is extremely limited; they apparently are restricted to the Southern Red Hills (Coastal Plain) in southern Alabama (Harper, 1920) and to the Marietta Plateau (Appalachian Plateaus) in eastern Ohio (Rick Gardner, personal communication).
XLPs are developed primarily on Paleozoic calcareous substrates (Table II). In the Ozark Plateaus, this vegetation type occurs on Cambrian, Ordovician, and Mississippian limestones and dolomites; in the Central Lowland (principally) on Mississippian and Ordovician limestones; in the Interior Low Plateaus on Mississippian, Silurian, and Ordovician limestones and dolomites; and in the Ridge and Valley on limestones and dolomites of Cambrian through Devonian age. In the East Gulf Coastal Plain in Alabama, Harper (1920) described XLPs restricted to the Tertiary (Eocene) Midway Limestone. Apparently, limestone of the Pennsylvanian Conemaugh Formation is the only substrate upon which this vegetation type is developed in the (unglaciated) Allegheny Plateaus (in Belmont County, Ohio) (Rick Gardner, pers. commun.).
Soils of XLPs typically are shallow ([less than or equal to] 1.0 m, mostly [much less than] 1.0 m) and rocky, and often they are mapped as rock outcrop complexes with one or more associated soil series (Baskin et al., 1994; Lawless et al., 2004). Despite circumneutral pH values (~6.0-8.0), soil fertility is often quite low, and many researchers have documented low soil phosphorus levels (Keeland, 1978; Heikens, 1991; Ludwig, 1999; Allison & Stevens, 2001; Rhoades et al., 2004; Trammell et al., 2004; Rhoades et al., 2005). Soil moisture varies considerably throughout the growing season, characteristically decreasing from saturated conditions in early spring to xeric conditions in summer and autumn (Aldrich et al., 1982). A literature review and personal communication revealed this vegetation type to be associated with five soil orders and 37 series: 22 Alfisols, six Ultisols, five Mollisols, three Inceptisols, and one Vertisol (Table III).
The vast majority of sites occur on moderate to steep slopes, where soil erosion generally exceeds soil genesis (Ware, 2002). Therefore, topographic position and soil texture may profoundly affect the distribution of XLPs. Many of the soil series upon which sites occur are relatively fine-textured clay loams and silty clay loams, which are extremely susceptible to erosion when they occur on moderate to steep slopes. In Kentucky, many of the soil mapping units associated with XLPs are moderately to severely eroded (Lawless et al., 2004). However, other studies include only the soil series, rather than mapping units per se, upon which sites occur and thus do not include detailed descriptions of these series under localized conditions. Heikens (1991) reported only soil associations for sites in southern Illinois, but each of the five series (Alford, Hormer, Muren, Stoy, and Weir) that make up these associations are Alfisols.
Climates in which XLPs occur in eastern United States are categorized as humid subtropical and humid temperate in Koppen's classification system (Ackerman, 1941). In areas of the southeastern United States (Arkansas, Alabama, Georgia, Tennessee, Kentucky, and Virginia) and in southern portions of certain midwestern states (Missouri, Illinois, Indiana, and Ohio), this vegetation type occurs in humid subtropical climates (Cfa) with rain in all seasons and hot summers (mean temperature of warmest months [less than or equal to] 22.2[degrees]C) (Trewartha, 1968, from Ackerman, 1941). XLPs in central to northern portions of midwestern United States (Missouri, Illinois, Indiana, and Ohio) and in the Ridge and Valley of West Virginia and Pennsylvania have a humid temperate climate (Dfa) with rain in all seasons and hot summers (Trewartha, 1968, from Ackerman, 1941). The primary difference between Cfa and Dfa climatic regions is the number of inclusive months with mean monthly temperatures >10[degrees]C. Subtropical climates (Cfa) have 8 to 12 and temperate climates (Dfa) 4 to 7 inclusive months with mean monthly temperatures >10[degrees]C (Trewartha, 1968, from Ackerman, 1941).
Precipitation patterns and amounts differ considerably between XLP sites in the eastern United States and grasslands of the Central Plains and Prairies Formation (sensu Barbour & Billings, 2000). Seasonality of precipitation throughout the range of XLPs is considerably less than that in the Central Plains and Prairies Formation (Markham, 1970), where amounts of precipitation are relatively high during the growing season and significantly less during the remainder of the year (Borchert, 1950). Mean annual precipitation generally decreases from south to north in the range of XLPs, varying from about 100 cm in midwestern United States to 150 cm in southeastern United States. Conversely, mean annual precipitation in the Central Plains and Prairies Formation decreases from east to west and typically is [less than or equal to] 90 cm (National Climatic Data Center, 2001).
[C.sub.4] perennial grasses (Schizachyrium scoparium, Bouteloua curtipendula, Andropogon gerardii, Sorghastrum nutans, and Sporobolus clandestinus) are typically dominant in XLPs. Schizachyrium scoparium is the characteristic dominant species in XLPs in eastern United States, having the highest cover, frequency, and/or importance value in 18 of 20 studies for which quantitative vegetation data are available (Table IV). However, in some sites [C.sub.3] perennial herbs (e.g., Silphium terebinthinaceum and Monarda fistulosa var. brevis) and/or taxa in the Cyperaceae, particularly Carex spp. and Fimbristylis spp., have cover values that equal or exceed those of dominant grasses. When examined at fine spatial scales (<1.0 [m.sup.2]), the [C.sub.4] summer annual grasses Sporobolus vaginiflorus and Sporobolus neglectus are locally dominant in shallow-soil zones of some sites (Skinner, 1979; Ver Hoef et al., 1993; Baskin & Baskin, 2000; Lawless, 2005).
In any particular site, species richness of dominant and subdominant perennial graminoids is low and that of perennial [C.sub.3] forbs is relatively high. However, the forb component of XLP floras varies considerably among sites and especially across ecoregions (Lawless et al., 2004; Lawless, 2005). Therefore, vegetation of XLPs, like that of the tall grass prairie (Glenn & Collins, 1990), is characterized by a stable matrix of dominant graminoids and a forb component that varies considerably both within and among sites (Lawless, 2005).
In the following section, vegetation of XLP communities in the eastern United States will be reviewed and discussed by physiographic province. This review focuses on quantitative vegetation data, but it also includes qualitative descriptions for regions where quantitative data are not available. For quantitative studies conducted by DeSelm (1988, 1991, 1993) and DeSelm and Webb (1997), mean cover values have been recalculated to incorporate quadrats that did not include the taxon of interest (i.e., cover = 0%), because DeSelm included only quadrats in which the taxon of interest was present.
Hall (1955) compared ground-layer species composition and Juniperus virginiana seedling distribution and population structure in an old field and "glade" (= XLP) in the Missouri Botanical Garden Arboretum in the northeastern section of the Salem Plateau. In the glade, Sporobolus neglectus had the highest frequency (1.00) in 1-[m.sup.2] quadrats, followed by Carex crawei (0.90), Schizachyrium scoparium (0.80), Houstonia longifolia (0.75), Euphorbia corollata (0.60), and Rudbeckia missouriensis (0.55). Hall described glade community structure and vegetation dynamics as follows: "In a good year the clumps of Andropogon [= Schizachyrium scoparium] are larger, denser, and more frequent. The [annual] Sporobolus effectively fills in between the clumps. In a dry or unseasonable year the Andropogon may lose ground, the Sporobolus to a lesser extent, but the Bouteloua [curtipendula] may increase considerably." Of 6694 J. virginiana seedlings recorded in the glade, 99.7% were in the smallest size class ([less than or equal to] 0.46 m in height), thus indicating extremely high seedling mortality in the glade. Hall concluded, "In general, juniper population density is proportional to the degree of land abuse, so that glades may support junipers distributed as dense 'brakes,' open stands with evenly but widely spaced individuals, or scattered, clumped colonies."
Schizachyrium scoparium had both the highest mean cover and frequency in sites sampled by Kucera and Martin (1957) in the White River region of the Salem Plateau in southwestern Missouri. Locally dominant graminoids included Andropogon gerardii, Bouteloua curtipendula, Panicum virgatum, Sorghastrum nutans, Sporobolus heterolepis, and Tridens flavus. Hedyotis nigricans, a [C.sub.3] herbaceous perennial, was the only forb with a relatively high frequency (57%). Kucera and Martin (1957) also reported increased frequencies of [C.sub.3] summer annuals (Ambrosia artemisiifolia, Croton capitatus, and Croton monanthogynus) "during the drouth years." They also noted increased densities of Juniperus virginiana in some sites, which in some instances resulted in "enclosure."
Skinner (1979) sampled vegetation associated with three threatened glade species (Penstemon cobaea var. purpureus, Centaurium texense, and Stenosiphon linifolius) using 0.01-[m.sup.2] and 0.1-[m.sup.2] quadrats centered around individuals of focal species. As noted by Baskin and Baskin (2000), "Sporobolus neglectus had the highest percent occurrence in quadrats of both sizes placed around Penstemon and Centaurium, and Schizachyrium scoparium had the highest percent occurrence in those of both sizes placed around S. linifolius, which grows in deeper soil (15.7 [+ or -] 6.1 cm) than Penstemon (10.9 [+ or -] 5.6 cm) or Centaurium (5.3 [+ or -] 2.8 cm)."
Schizachyrium scoparium had the highest mean importance value (29.2%) in 20 sites sampled by Hicks (1981) in the Hercules Glade Wilderness in the Salem Plateau. Importance values were obtained from frequency and cover data collected in fifty 20 cm x 50 cm quadrats per site. Other taxa with mean importance values [greater than or equal to] 5% included Hedyotis nigricans (10.5%), Sporobolus neglectus (9.5%), Rudbeckia missouriensis (7.3%), Panicum virgatum (6.4%), and Croton spp. (C. capitatus and C. monanthogynus) (5.1%). Woody species data were collected in one or two 15 m x 25 m macroplots per site. Juniperus virginiana had the highest mean importance value (41.2%) of all tree species, followed by Cotinus obovatus (19.2%) and Diospyros virginiana (12.5%). The mean importance value of Rhus aromatica (58.8%) exceeded those of all other shrubs and of all vines.
Ver Hoef et al. (1993) sampled 24 sites on Gasconade Dolomite and seven sites on Eminence Dolomite in the vicinities of the Current and Jack's Fork rivers in the Salem Plateau of southeastern Missouri. Mean percent cover values reported in this study actually represent average geometric means (range = 0 to 31.6), since cover class midpoints were converted to geometric midpoints prior to calculations of means. On both substrates, Schizachyrium scoparium had the highest mean cover value (31.6 on each) in 7 m x 7 m quadrats. Mean cover of Rudbeckia missouriensis (Gasconade, 31.6; Eminence, 14.1) and of the perennial [C.sub.4] grass Sporobolus clandestinus (Gasconade = 9.6; Eminence = 15.4) also was high on both substrates. Fimbristylis caroliniana had the second highest mean cover value (23.0) on Eminence Dolomite glades, followed by Hedyotis nigricans (15.4) and Liatris cylindracea (15.4). TWINSPAN analysis of cover class data collected in 7 m x 7 m quadrats identified Bouteloua curtipendula and Sorghastrum nutans as significant indicator species of Gasconade Dolomite sites. Aristida purpurascens was the only significant indicator species of Eminence Dolomite sites. Mean cover of Sporobolus vaginiflorus exceeded that of all perennial grasses in 70 cm x 70 cm quadrats sampled in the "shallow soil glade" (mean soil depth = 7.5 cm) microhabitat in the Brandewiede Hollow East A site and in the "rocky glade" (5.9 cm) and "shallow soil glade" (10.7 cm) microhabitats in the Cave Spring E site, both of which are on Gasconade Dolomite. Juniperus virginiana had the highest mean basal area on both substrates. However, woody species richness and total basal area in Gasconade Dolomite sites was significantly higher than those in Eminence Dolomite sites.
The only vegetation data for the Springfield Plateau region in Missouri were collected by George (1996). Sampling was conducted in two sites in both spring (June) and autumn (October) using 1-[m.sup.2] quadrats. Panicum virgatum had the highest mean cover in both seasons in the Cross Timbers site and in autumn in the Silver Mine Ridge Road site. Schizachyrium scoparium had the second highest cover in both seasons at the Cross Timbers site and the highest cover in spring in the Silver Mine Ridge Road site. Cover of [C.sub.4] perennial grasses increased from 10% to 70% between spring and autumn sampling. Other species with relatively high cover values in one of these sites included Rudbeckia missouriensis, Echinacea paradoxa var. paradoxa, Bouteloua curtipendula, Liatris aspera, Liatris squarrosa, Fimbristylis puberula, Hedyotis nigricans, Andropogon gerardii, and Silphium terebinthinaceum. Juniperus virginiana was the dominant woody taxon in both sites.
Of the four "glade" community types (grassland-cedar, cedar, cedar-hardwood, and hardwood) described by Keeland (1978) in the Springfield Plateau, Salem Plateau, and Boston Mountains regions of northern Arkansas, only the grassland-cedar type fits the description of XLPs. The other community types presumably represent seral stages from a grassland-cedar community type to a hardwood forest (Baskin & Baskin, 2000). In the grassland-cedar community type, Schizachyrium scoparium had the highest mean cover value (22%), followed by Desmodium rotundifolium, Sorghastrum nutans, and Desmodium paniculatum. Juniperus virginiana and Quercus stellata were the dominant overstory species, and Rhus glabra had the highest importance value of all understory species.
In the Buffalo National River in north-central Arkansas, Logan (1992) estimated abundance of species in limestone and sandstone "glades" "using a subjective five point scale," ranging from "5 [=] abundant (widespread with high cover, a dominant or near-dominant species" to "1 [=] rare (only one or two individuals)." Schizachyrium scoparium was in the highest abundance class in 18 of 20 limestone "glades." Andropogon gerardii, Sporobolus vaginiflorus, and Bouteloua curtipendula were dominant or codominant in these sites. Coreopsis tinctoria, Acalypha gracilens, Eupatorium altissimum, and Rudbeckia missouriensis were the only forbs in abundance classes 4 or 5 in three or more sites. Abundant woody species included Juniperus virginiana, Quercus marilandica, and Quercus stellata.
Only one of five limestone "barrens" sampled by Heikens (1991) in southern Illinois is in the Central Lowland (Fults Limestone Glade, Monroe County, Illinois). The only taxon with a mean cover value [greater than or equal to] 1.0% in fifteen 50-[m.sup.2] circular plots was Schizachyrium scoparium (12.0%). Bouteloua curtipendula, Dalea purpurea, Lespedeza capitata, Rudbeckia missouriensis, Solidago nemoralis, and the site dominant, S. scoparium, were present in 100% of these quadrats. Woody taxa present in five or more of the 15 quadrats were Comus drummondii (7 quadrats) and Ceanothus americanus (6).
Community structure differed markedly among three "limestone glades" sampled by McClain and Ebinger (2002) in Calhoun County, Illinois. In Schleeper Glade, Hedyotis nigricans [percent importance value (%IV) = 35.9] and Bouteloua curtipendula (35.6) were codominant, and Croton capitatus (9.4) was the only other taxon with a %IV[greater than or equal to]5. Bouteloua curtipendula had the highest %IV (= 20.2) in Kamp's Glade, followed by Ruellia humilis (11.2), Croton capitatus (10.6), Schizachyrium scoparium (8.0), and Hedyotis nigricans (5.5). Lead Hollow Glade was the only site in which S. scoparium was dominant (%IV = 30.5). Hedyotis nigricans (%IV = 16.0), and B. curtipendula (13.1) were subdominant, and R. humilis (7.5), Eupatorium altissimum (6.0), and Croton capitatus (5.7) were the only other taxa with %IV [greater than or equal to] 5.
Vegetation data are not available for sites in the Central Lowland in Indiana. Homoya (1987) qualitatively described the vegetation of a "limestone glade" in Versailles State Park (Ripley County) as follows: "... Andropogon gerardii, Aster sagittifolius [Symphyotrichum urophyllum], Silphium trifoliatum, Physostegia virginiana, Lithospermum canescens, Kuhnia eupatorioides, Helianthus hirsutus, and Euphorbia corollata are the herbaceous species with the highest estimated importance values." Important woody species included Quercus muhlenbergii, Viburnum rufidulum, Celastrus scandens, Cercis canadensis, Ceanothus americanus, and Juniperus virginiana." In two "glades" in the Indiana Army Ammunition Plant (Clark County), Maxwell (1987) reported, "Little bluestem [Schizachyrium scoparium] dominates both glades as a xeric, bunch-grass surrounded by patches of rocky pavement." The following herbaceous species were present in both the "little bluestem glades" and the limestone cedar glades surveyed by Baskin and Baskin (1975) in nearby Bullit County, Kentucky: Ruellia humilis, Manfreda virginica, Isanthus brachiatus, Ophioglossum engelmannii, Sporobolus vaginiflorus, Heliotropium tenellum, Scutellaria parvula, Houstonia canadensis, Sisyrinchium albidum, Asclepias verticillata, and Croton monanthogynus. Juniperus virginiana also was common in the "open glades" surveyed by Maxwell (1987).
The distribution map of "limestone glades" in Missouri compiled by Ladd and Nelson (1982), shows a concentration of sites in the vicinity of the Salt, Missouri, and Mississippi rivers in the glaciated portion of the state. However, to our knowledge, vegetation data are not available for this region. Density of glades (measured as areal coverage per 7.5-minute quadrangle) in this region is low ([less than or equal to]10 acres/7.5-minute quadrangle) in comparison with the Ozark Plateaus (Ladd & Nelson, 1982). This inequity in glade development between regions (Ozark Plateaus and Central Lowland) may have prompted vegetation scientists to focus their research on the Ozarks, where glades are both more numerous and typically larger in comparison with those in the Central Lowland (Nelson & Ladd, 1982a). However, collection of vegetation data in XLPs in the glaciated portion of the Central Lowland in Missouri would fill a gap in our knowledge of this vegetation type in eastern United States.
INTERIOR LOW PLATEAUS
Four of five "limestone glades" sampled by Heikens (1991) in southern Illinois are in the Interior Low Plateaus. According to Heikens, diagnostic features of these limestone glades include: cover of "prairie species" [greater than or equal to]10%, cover of woody species < 50%, soil depth < 10 cm, and exposed rock [greater than or equal to]5%. Schizachyrium scoparium was the dominant taxon in three sites, with mean cover values ranging from 15.7% to 25.5%. Silphium terebinthinaceum (mean cover = 9.1%) and S. scoparium (7.9%) were codominant in the fourth site (Whoopie Cat Bluff). Symphyotrichum oblongifolium (mean cover = 6.9%) and Echinacea pallida (4.9%) were subdominant taxa in the Cave Creek site. Quercus muhlenbergii and Juniperus virginiana were the only additional taxa with mean cover values [greater than or equal to]5% in one or more sites. TWINSPAN analysis of "natural forest openings" sampled by Heikens (1991) identified Physostegia virginiana, Manfreda virginica, and Asclepias verticillata as significant indicator species of the limestone glade community type.
Twenty-three of 32 "limestone glades" sampled by Kurz (1981) are in extreme southern Illinois in the Interior Low Plateaus. The remaining nine sites occur "along the Mississippi River valley" (Jersey, Monroe, St. Clair, and Union counties) and "on the uplands above the Illinois River" (Calhoun and Pike counties) in the Central Lowland. For select taxa, Kurz (1981) provided frequency data averaged across all sites. However, as mentioned above, the majority of sites (71.9%) are in the Interior Low Plateaus (Shawnee Hills), and thus the cursory data provided are discussed in the context of this physiographic province. In 30 of 32 sites, Schizachyrium scoparium had the highest frequency in 0.25-[m.sup.2] circular quadrats. Bouteloua curtipendula and Sorghastrum nutans each had the highest frequency in one of the other sites. According to Kurz (1981), "Dominant species of forbs occurring in several glades were Aster oblongifolius [= Symphyotrichum oblongifolium], Echinacea pallida, Croton monanthogynus, Hedyotis nigricans, Manfreda virginica, Euphorbia corollata, Physostegia virginiana, and Brickellia eupatorioides [= Kuhnia eupatorioides]."
Braun (1928) sampled "prairies" and "cedar barrens" on Crab Orchard Shale (= Estill Shale), Cedarville Dolomite (= Peebles Dolomite), and Monroe Dolomite (= Lilley and Bisher formations) in the Mineral Springs Region of Adams County, Ohio. She described sites developed on Crab Orchard Shale as "barren cleared slopes, everywhere showing the effects of slumping." According to Braun (1928), Juniperus virginiana, Panicum flexile, Allium cernuum, Cuphea viscosissima, Sabatia angularis, Ratibida pinnata, Monarda fistulosa, and Senna marilandica are characteristic, and Hedyotis nigricans is "particularly characteristic" of sites developed on this substrate. She reported frequency data for sites developed on Cedarville [Turkey Creek (two sites) and Cedar Fork (three sites)] and Monroe [Turkey Creek (two sites)] dolomites (Table IV, pages 427-431). Andropogon gerardii had the highest frequency (100%) in both sites on Monroe Dolomite (= Lilley and Bisher formations), followed by Silphium terebinthinaceum (90% and 82%, respectively). Schizachyrium scoparium had the highest frequency in three (Turkey Creek, site 1; Cedar Fork, sites 1 and 3) of five sites on Cedarville Dolomite. In Turkey Creek site 2, Fragaria virginiana had the highest frequency (55%), followed by Blephilia ciliata (52%), Helianthus occidentalis (50%), Lithospermum canescens (47%), Schizachyrium scoparium (43%), Andropogon gerardii (40%), and Delphinium exaltatum (40%). In Cedar Fork site 1, Manfreda virginica and Bouteloua curtipendula had the highest frequency values (90% and 78%, respectively). Remaining taxa with frequency values [greater than or equal to]50% in one or more sites include Agrimonia pubescens, Carex crawei, Leucanthemum vulgare (normative species), Comandra umbellata, Pycnanthemum tenuifolium, Ruellia humilis, Euphorbia corollata, Helianthus hirsutus, Isanthus brachiatus, Scutellaria parvula, Packera (Senecio) plattensis, Solidago nemoralis, and Sorghastrum nutans. With regard to woody species encroachment, Braun stated: "Among the tree invaders, Juniperus virginiana and Thuja occidentalis are most prominent; with these are Ostrya, Cercis, and Quercus muhlenbergii."
As is the case for the Central Lowland, vegetation data are lacking for XLPs in the Interior Low Plateaus in Indiana. Bacone et al. (1983) surveyed "glades" and "barrens" of Crawford and Perry counties and provided the following qualitative description: "On the barrens, prairie dock [Silphium terebinthinaceum] was very common, and Elliott's bluestem [Andropogon elliottii] was an important grass, in addition to big bluestem [Andropogon gerardii], little bluestem [Schizachyrium scoparium], and Indian grass [Sorghastrum nutans]." Aldrich et al. (1982) mention very little about vegetation in "limestone glades" in Harrison County, except for the following general statement: "They are dominated by prairie grasses and forbs."
Lawless et al. (in press) used a multiple-scale sampling design to identify XLP community types (0.01- to 100-[m.sup.2]) in 18 sites in the Interior Low Plateaus in Kentucky. Cluster analysis of cover class data obtained from 100-[m.sup.2] quadrats identified 12 community types. Schizachyrium scoparium had the highest mean cover value in 10 of these community types, and Andropogon gerardii and Silphium terebinthinaceum each had the highest mean cover value in one of the two remaining 100-[m.sup.2] community types. Dominant taxa (mean cover [greater than or equal to]10%) in three or more community types include S. scoparium (10 community types), Sporobolus vaginiflorus (5), Sorghastrum nutans (5), Echinacea simulata (5), Andropogon gerardii (4), and Carex crawei (3). Andropogon gerardii was subdominant (5% [less than or equal to]mean cover < 10%) in seven community types, and six taxa, including Echinacea simulata, Carex meadii, Hedyotis nigricans, Sorghastrum nutans, Juniperus virginiana, and Cercis canadensis, were subdominant in three community types. Sporobolus vaginiflorus was locally dominant in portions of many sites, as evidenced by its high percent occurrence values in numerous fine-scale community types (0.01- and 0.1-[m.sup.2]). In addition, high percent co-occurrence values of S. vaginiflorus and S. scopariumin certain fine-scale community types suggest that S. vaginiflorus is capable of coexisting with the dominant perennial grass over a range of environmental conditions. Characteristic trees and shrubs in this vegetation type in Kentucky include Juniperus virginiana, Cercis canadensis, Diospyros virginiana, Fraxinus americana, Rhamnus caroliniana, Rhus aromatica, Cornus florida, and Rosa carolina (Lawless, 2005).
Schizachyrium scoparium also is the dominant taxon in XLPs in the Interior Low Plateaus in Tennessee. DeSelm (1991) studied "small remnants of barrens" in the Central Basin, which "occur on shallow soil ... as an alternative to the cedar forest which borders glades." Schizachyrium scoparium had the highest mean cover value in each of the three sites for which data were reported: Mt. Juliet Road "glade" (mean cover = 40%); Mt. Juliet Road barren (40%); and Cedars of Lebanon glade border (76%). Sporobolus asper was codominant (mean cover = 35%) in Mt. Juliet Road barren. Although Bouteloua curtipendula was subdominant in Cedars of Lebanon glade border (mean cover = 15%), its distribution apparently was restricted within the site, as evidenced by its low frequency (15%). Hedyotis nigricans had the highest frequency (100%) and highest mean cover (19%) in the Mt. Juliet Road "glade." No other taxon had a mean cover value [greater than or equal to]5% in any site. Schizachyrium scoparium, Sporobolus vaginiflorus, and the [N.sub.2]-fixing cyanobacterium Nostoc commune had the highest frequency values in another site in Cedars of Lebanon State Forest sampled by Baskin and Baskin (1977). Schizachyrium scoparium also was the dominant taxon in three "barrens" in the Western Highland Rim (Perry and Decatur counties) sampled by DeSelm (1988). Other taxa with percent importance values [greater than or equal to]5% in one or more sites include Galactia volubilis, Opuntia humifusa, Euphorbia corollata, Sporobolus clandestinus, Solidago nemoralis, Croton monanthogynus, Fimbristylis puberula, Andropogon gerardii, and Sorghastrum nutans.
DeSelm and Webb (1997) sampled the "lower flat" (0-5% slope), "footslope" (11-20%), and "steep sideslope" (32-70%) in two grasslands in the Southern Highland Rim in northwestern Alabama. In both sites (Littleville and Cedar Creek), footslopes and steep sideslopes had the lowest mean soil depths (6.5-13.5 cm) and highest mean combined cover values of rock and gravel (24.4-86.0%), and thus portions of sites occupying these slope positions can be categorized as XLPs. In the footslope of Littleville Barren, mean total plant cover was only 46.3%, and Schizachyrium scoparium (mean cover = 5.0%) was the only taxon with a mean cover value [greater than or equal to]2.0%. Sporobolus clandestinus (mean cover = 30.2%) and Schizachyrium scoparium (16.2%) were the dominant and subdominant taxa, respectively, in the steep sideslope. Schizachyrium scoparium was the dominant taxon in both the footslope (mean cover = 18.7%) and steep sideslope (21.3%) of Cedar Creek Barren. Bouteloua curtipendula, Fimbristylis puberula, Sorghastrum nutans, and Celtis tenuifolia were the only other taxa with mean cover values >1.0% in either landform in this site.
RIDGE AND VALLEY
Bouteloua curtipendula, Danthonia spicata, and Schizachyrium scoparium "dominated in terms of abundance" in XLPs in Pennsylvania studied by Laughlin and Uhl (2003). However, quadrat sampling was not performed, and, therefore, abundance values reported in this study are semiquantitative. According to Laughlin (2002), common forbs in these sites are Anemone virginiana, Solidago nemoralis, Monarda fistulosa, Lithospermum canescens, Asclepias tuberosa, A. verticillata, A. viridiflora, Penstemon hirsutus, and Senecio obovatus, and common trees are Juniperus virginiana, Juglans nigra, Celtis occidentalis, Pinus strobus, and P. virginiana. Percentage of nonnative species in the flora of XLPs in Pennsylvania is particularly high (25% and 28% in two sites) (Laughlin, 2004). Consequently, Laughlin and Uhl (2003) concluded, "Exotic invasion is likely a significant reason for the loss of native species from limestone prairies."
Schizachyrium scoparium was the dominant taxon in 14 of 16 plots (11 sites total) sampled by Ludwig (1999) in southwestern Virginia. Andropogon gerardii was the dominant taxon in two plots. Other herbaceous taxa with cover values [greater than or equal to]5% in one or more sites include Symphyotrichum sericeum, Carex eburnea, C. hirsutella, Liatris aspera, Minuartia patula, Oxypolis rigidior, Scleria verticillata, Sorghastrum nutans, Sporobolus clandestinus, and S. vaginiflorus. In the "shrub layer" (height, 1-6 m), Cercis canadensis, Comus florida, Frangula caroliniana, Juniperus virginiana, and Quercus muhlenbergii were the only taxa with cover values [greater than or equal to]5%. In 12 of 16 plots, no taxa were present in the "tree layer" (6-10 m), and in the remaining four plots J. virginiana and/or Q. muhlenbergii were present.
Bartgis (1993) sampled "glades," "barrens," and "cedar woodlands" on Knobly and Cave mountains in West Virginia. Glades had shallower soils (mean depth, 3.9 cm) and higher cover of cobble and bedrock (mean cover, 74%) than did limestone barrens (mean soil depth, 8.9 cm; mean cover cobble/bedrock, 14%). However, perennial [C.sub.4] grasses and/or perennial [C.sub.3] herbs were the dominant taxa in both of these shallow-soil community types, and thus they can be classified as XLPs. In Knobly Mountain cedar glades, Bouteloua curtipendula (mean cover, 8%) and Solidago arguta var. harrisii (4%) were the dominant herbaceous taxa. Phlox subulata and Paronychia virginica were the only remaining herbaceous taxa with mean cover values > 1%. According to Bartgis (1993), "Bouteloua is much less important at the Cave Mountain cedar glades, where the dominants are typically Solidago arguta vat. harrisii, Monarda fistulosa var. brevis, and Paronychia virginica; Carex eburnea is locally dominant." Important woody taxa in these sites include Juniperus virginiana, Cercis canadensis, Quercus muhlenbergii, and Thuja occidentalis (Cave Mountain sites only). Total herbaceous cover was considerably higher in the Knobly Mountain limestone barrens, where B. curtipendula (mean cover, 35%), Elymus hystrix (12%), and Schizachyrium scoparium (5%) were the dominant taxa.
Eighty-six sites, including "glades" and "barrens," in the Ridge and Valley of Alabama, Georgia, Tennessee, and Virginia were studied by DeSelm (1993). He distinguished barrens from glades based upon percentage of perennial grass cover, with the former having perennial grass cover > 50% and the latter < 50%. Data were reported for three barrens, all of which are consistent with our definition of XLPs. In Crowder Cemetery Barren (Roane County, Tennessee), Andropogon gerardii (mean cover, 17.5%) and Schizachyrium scoparium (16.1%) were codominant, and Liatris cylindracea and Hypericum dolabriforme were the only other taxa with mean cover values > 1%. Schizachyrium scoparium (mean cover, 20.7%) and Bouteloua curtipendula (16.6%) were codominant in the Elementary School Barren (Catoosa County, Georgia). Subdominant taxa (5% < mean cover [less than or equal to]10%) include Panicum virgatum, Andropogon gerardii, and Hypericum dolabriforme. Schizachyrium scoparium (22.2%) and Aristida purpurascens (21.8%) were codominant in the third site (Centre, Alabama), and they were the only taxa with average cover values [greater than or equal to]1%.
Schizachyrium scoparium is also the dominant graminoid in the Ketona Dolomite "Glades" in the Ridge and Valley near its southwestern terminus, in Alabama. However, as noted by Allison and Stevens (2001), "... it usually does not achieve great density and is an aspect dominant only in late fall and winter, when the strong forb component is muted." Onosmodium decipiens and Erigeron strigosus var. dolomiticola, both Ketona Dolomite Glade endemics, were abundant at the majority of sites. Allison and Stevens (2001) also report, "Amsonia ciliata var. tenuifolia is often abundant and dense enough to be an aspect dominant in spring, and Rudbeckia triloba var. pinnatiloba is occasionally an aspect dominant in summer." Ketona Dolomite Glades also are noteworthy for supporting a number of woody taxa that typically do not occur in the remainder of the range of xeric limestone prairies, including Pinus palustris, Sabal minor, and Croton alabamensis var. alabamensis.
According to Allison and Stevens (2001), "Ketona Glades fail several to many criteria" for each of the three community types (limestone cedar glades, XLPs, and deep-soil barrens) described by Baskin et al. (1994) in the Big Barrens Region of Kentucky and Tennessee. The authors conclude that "Ketona Glades come closest to the 'xeric limestone prairie' class" but point out the following characteristics of Ketona Glades that differ from those of XLPs described by Baskin et al. (1994): (1) their occurrence on dolomite rather than limestone, (2) multiple endemic taxa, and (3) presence of two species of Leavenworthia. However, as discussed above, XLPs occur on dolomite in various regions throughout the geographic range of this community type in the eastern United States. Furthermore, XLP endemics (Table IV) occur in regions other than the Cahaba River valley in Alabama, particularly in the Ozark Plateaus in Missouri and Arkansas. In addition, Leavenworthia spp., some of which primarily are limestone cedar glade endemics (e.g., Leavenworthia alabamica, Leavenworthia exigua vat. exigua, and Leavenworthia exigua var. laciniata), occur in other XLPs in the eastern United States (Baskin & Baskin, 1977; Maxwell, 1987; DeSelm, 1988, 1991, 1993; DeSelm & Chester, 1993; Ver Hoef et al., 1993; Webb et al., 1997; Gardner & Minnie, 2004; Lawless et al., 2004).
Vegetation data are not available for XLPs in the unglaciated Allegheny Plateau, which apparently are restricted to Belmont County, Ohio (Rick Gardner, pers. commun.).
Harper (1920) is the only source of information on vegetation of "limestone prairies" in the Coastal Plain, in Alabama. He provided a cursory species list and estimated "the commonest species in the areas of natural prairie, in order of abundance," the first five of which are as follows: Schizachyrium scoparium, Polygala grandiflora, Fimbristylis puberula, Hedyotis nigricans, and Prunella vulgaris. He also states, "Juniperus virginiana is practically the only tree" throughout the geographically restricted area in which "natural prairie vegetation" is developed on limestone, a very uncommon substrate in the Coastal Plain of Alabama.
Endemic Taxa and Other Noteworthy Floristic Elements
Thirteen taxa are endemic/near endemic to XLPs of eastern United States (Table V). The Cahaba River valley in the Ridge and Valley in Alabama and the Ozark Plateaus in Missouri and Arkansas are centers of endemism in this vegetation type. Allison and Stevens (2001) described eight new taxa endemic to XLPs on Ketona Dolomite rock outcrops in the Cahaba River valley in Bibb County, Alabama, including Castilleja kraliana, Coreopsis grandiflora var. inclinata, Dalea cahaba, Erigeron strigosus vat. dolomiticola, Liatris oligocephala, Onosmodium decipiens, Silphium glutinosum, and Spigelia gentianoides vat. alabamensis. Delphinium treleasei, Echinacea paradoxa var. paradoxa, and Scutellaria bushii are restricted to XLPs in the Ozark Plateaus in Missouri and Arkansas. Valerianella ozarkana occurs in the Ozark Plateaus in Missouri, Arkansas, and Oklahoma and in the Ouachita Mountains in eastern Oklahoma (USDA, NRCS, 2004; Hoagland et al., 2004), and thus should be considered an Ozark-Ouachita endemic. According to Bartgis (1993), "Monarda fistulosa vat. brevis appears to be endemic to cedar glades, limestone barrens, glade woodlands, and dry limestone cliffs of West Virginia and Virginia" (also see Kimball et al., 2001). This is the only XLP endemic outside the Ozarks Plateaus not restricted to Ketona Dolomite.
The flora of XLPs also contains taxa considered to be endemic to other vegetation types, most notably shale barrens and limestone cedar glades. Bartgis (1993) noted the occurrence of eight mid-Appalachian shale barren endemics (sensu Keener, 1983) in XLPs in West Virginia, including Calystegia spithamaea subsp, purshianus, Oenothera argillicola, Solidago arguta var. harrisii, Taenidia montana, Trifolium virginicum, Antennaria virginica, Helianthus laevigata (not an endemic), and Paronychia montana. Twelve of the 19 taxa considered to be limestone cedar glade endemics by Baskin and Baskin (1999) also occur in XLPs of eastern United States (Table VI), including Dalea gattingeri, Echinacea tennesseensis, Leavenworthia alabamica, Leavenworth& exigua var. exigua, Leavenworthia exigua var. laciniata, Leavenworthia exigua var. lutea, Leavenworthia stylosa, Lobelia appendiculata var. gattingeri, Onosmodium molle, Pediomelum subacaule, Talinum calcaricum, and Trifolium calcaricum. Interestingly, Allison and Stevens (2001) suggested Pediomelum subacaule, a [C.sub.3] perennial cryptophyte lacking long-distance dispersal mechanisms, was introduced in XLPs on Ketona Dolomite. The limestone cedar glade near-endemics Astragalus tennesseensis and Dalea foliosa (sensu Baskin & Baskin, 1999) also are known from XLPs in the Cumberland River Basin in Tennessee (Rutherford and Davidson counties) (DeSelm, 1991).
As noted by DeSelm (1991, 1993), limestone cedar glade endemic/near-endemic taxa often occur in XLPs that are adjacent or in close proximity to limestone cedar glades. Exceptions include Pediomelum subacaule and Leavenworthia exigua var. lutea in sites on Ketona Dolomite in Bibb County, Alabama, and Talinum calcaricum in an XLP in Logan County, Kentucky. In virtually all instances, occurrence of limestone cedar glade endemic/near-endemic taxa are restricted to microsites that fit the description of limestone cedar glades yet exist within the perennial grass matrix of XLPs [see Fig. 1 (middle) in Lawless et al., 2004].
Other floristic elements of XLPs also are restricted to vegetation types developed on calcareous rock outcrops. Catling et al. (1993) described a new sedge, Carex juniperorum, based on collections from alvars in the Napanee limestone plain in Ontario and from XLPs and similar habitats in southeastern Ohio (Adams County) and northeastern Kentucky (Lewis and Bath counties). Recently, Wieboldt (Belden et al., 2004) reported a new station for this species in Montgomery County, Virginia, where it occurs "in dry or seasonally wet spots in a small barren or in the thin chinquapin oak-redbud-red cedar woodland." Echinacea simulata apparently is restricted (or nearly so) to XLPs and similar habitats in Missouri, Illinois, Indiana, Kentucky, and Tennessee. However, inconsistent taxonomic treatment of this taxon (Baskin et al., 1993; Kim et al., 2004), its remarkable similarity to Echinacea pallida (Binns et al., 2002), and sympatry of Echinacea pallida and Echinacea simulata throughout much of the latter's geographic range (Missouri, Illinois, Indiana, and Tennessee) make it difficult to discern the true distribution of this species and assess its fidelity to XLPs. The winter annual Lesquerella filiformis is narrowly distributed in southwestern Missouri (Jasper, Dade, Lawrence, Greene, and Christian counties) (United States Fish and Wildlife Service [USFWS], 1988) and northwestern Arkansas (Izard and Washington counties) (Arkansas Natural Heritage Commission, 2004), where it "... occurs in shallow soils on limestone glades, outcrops in pastures and rarely in rocky open woods (USFWS, 1988)." Viola egglestonii ranges from southern Indiana (Harrison County) south to northern Alabama (Franklin County) and northern Georgia (Catoosa and Walker counties) and was once considered to be a limestone cedar glade endemic (sensu Baskin & Baskin, 1978, 1986, 1989). However, this species also occurs in XLPs and other gladelike habitats (DeSelm, 1993; Baskin & Baskin, 2003; Lawless et al., 2004), and thus it is no longer considered to be a limestone cedar glade endemic or even a near endemic (sensu Baskin & Baskin, 2003).
Floras of XLPs in the Ozark Plateaus in Missouri and Arkansas have western and southwestern affinities. Many taxa reach the eastern edge of their geographic range in XLPs in this region (Nelson & Ladd, 1982b); examples are Acacia angustissima var. hirta, Castilleja purpurea, Centaurium texense, Juniperus ashei, Marshallia caespitosa var. signata, Nemastylis nuttallii, Palafoxia callosa var. callosa, Parthenium hispidum, Penstemon arkansana, and Yucca glauca var. glauca. Evolvulus nuttallianus, Onosmodium molle var. subsetosum, Oenothera macrocarpa subsp, macrocarpa, and Solidago gattingeri occur in XLPs in the Ozarks and are absent further east with the exception of limestone cedar glades in Middle Tennessee (Baskin & Baskin, 1986; Adams et al., 2003).
Origin and Maintenance
PHYSICAL ENVIRONMENTAL FACTORS AND ANTHROPOGENIC INFLUENCE
There is no general consensus regarding the origin of XLPs. Current distribution of this vegetation type in the eastern United States likely reflects interaction of multiple factors over variable time intervals ranging from decades to millennia. DeSelm and Murdock (1993) astutely acknowledged, "The scattered occurrence of grassland and related communities [in southeastern United States] does not argue strongly for climate nor great soil group control over community distribution?' XLPs in the Ozark and mid-western regions of United States also are scattered geographically and occur on numerous soil series and orders. Shallow soils (including 37 series in five orders), calcareous bedrocks resistant to weathering, moderate to steep slopes (typically with south to west aspects), anthropogenically increased fire frequencies (native Americans and/or European settlers), land clearing for cultivation or pasturage, grazing (by both native and nonnative mammals), and erosion are factors frequently mentioned, singly or in concert, as potential sources of origin and/or maintenance for this vegetation type (Fig. 3).
[FIGURE 3 OMITTED]
Braun (1928), Steyermark (1940), and Logan (1992) suggest XLPs are a primary vegetation type. Logan (1992) states, "Glades are primary natural communities" which are "... maintained indefinitely at an early stage of succession by the substrate or by natural forces such as erosion or microclimate?' Steyermark (1940) regarded "glades" as the first two seral stages [(1) Bouteloua curtipendula-Rudbeckia missouriensis and (2) Rhus aromatica-Diospyros virginiana-Juniperus virginiana] of a six-stage primary succession culminating in an Acer saccharum-Quercus alba climatic climax (see reviews in Baskin & Baskin, 2000, and Hicks, 1981). Braun (1928) offered a compelling argument against the notion of a recent, anthropogenic origin for "extensive prairie vegetation" developed on Cedarville [= Peebles] Dolomite in the Mineral Springs region of Ohio.
Clearing, the entrance and occupancy by a prairie community and its gradual elimination by forest invasion--which on the edges is extremely slow--would have had to take place within a period of 125 years. It seems far more plausible that the forest is a stage in primary succession, representing the slow migration of the sub-climax prairie of ridges.
Braun (1928) goes on to state, "There is no vegetational evidence that this community ["prairie"] is not primary; and the oldest inhabitants say that there never have been anything but tall grasses and cedars in these places." Conversely, Braun (1928) refers to the "cedar barrens" developed on Crab Orchard [= Estill] Shale in this same region as "barren cleared slopes, everywhere showing the effects of slumping." Thus, she suggests an anthropogenic origin for these community types. Furthermore, Braun makes the following statement regarding the "Andropogon gerardii-Silphium [terebinthinaceum] prairies" developed on Monroe Dolomite [= Lilley and Bisher formations] in the region: "Their position in relation to cultivated fields and to cleared slopes suggests they may be secondary." Therefore, she (Braun 1928) considered XLPs developed on Peebles Dolomite to be primary and those developed on other substrates in Adams County to be secondary.
Other researchers also implicate agriculture as a potential source of origin and/or expansion for this vegetation type. B askin et al. (1994, 1997) attributed formation of this vegetation type in the Kentucky Karst Plain [= Pennyroyal Plain + Elizabethtown (Plain) sensu Quarterman and Powell, 1978] to the following sequence of events: clearing of marginal agricultural lands by European settlers [right arrow] cultivation and/or grazing significant erosion of the topsoil [right arrow] abandonment [right arrow] colonization of these degraded areas by XLP flora [right arrow] succession to hardwood forest in the absence of disturbance or maintenance of XLP with periodic management (i.e., disclimax) (Lawless et al., 2004). According to Rhoades et al. (2005), "Soil erosion associated with land clearing may be an alternative explanation [in contrast to geologic substrate] for the genesis of these glade openings [Crooked Creek Barrens; Lewis County, Kentucky]." DeSelm (1993) suggested land-use patterns of both native American and colonial European cultures affected the maintenance and possible expansion, respectively, of "barrens" (including XLPs) in the Ridge and Valley of Alabama, Georgia, Tennessee, and Virginia.
Landscape disturbances caused by Native Americans, over and above such natural disturbances as lightning-caused fires, helped maintain necessary openings. European man pastured and burned these xeric sites and perhaps increased their area by plowing that resulted in soil loss.
For the Central Basin of Tennessee, DeSelm (1991) stated, the "... crescentic ring of barren-like vegetation between the glade and cedar thicket or forest" was the result of "clearing of the cedar or burning and grazing of this vegetation" and subsequent extension of perennial grasses. According to Bartgis (1993), "Most glades and barrens in West Virginia have been pastured, typically for sheep." However, he attributes the maintenance of XLPs to "periodic severe droughts," such as the one recorded in 1987 "which promote[s] prominence of Juniperus virginiana over deciduous species."
The vast majority of researchers attribute the origin and maintenance of XLPs to interaction of two or more of the following endogenous or exogenous factors: shallow soils, weathering-resistant bedrock, moderate to steep slopes (typically with south to west aspects), erosion, fire, and grazing. According to Heikens (1999), "The Ozark Plateaus supports a mosaic of prairies, forests, glades, barrens, and savannas depending on such factors as topography, bedrock, soils, fire, and native herbivores." Ver Hoef et al. (1993) stated, "Studies have indicated that bedrock type, topography, weathering, soil conditions, drought, and fire are some of the major factors influencing both the distribution of glade types and patterns within and around glades." Laughlin (2004) focused on maintenance rather than origin(s) of XLPs, yet he suggested anthropogenic fire and/or overgrazing (by nonnative ungulates) in combination with extremely xeric edaphic conditions (particularly in sites on Opequon soil series) were likely responsible for establishment of XLPs in the Ridge and Valley of Pennsylvania. Despite Logan's (1992) previously mentioned assertion that XLPs are primary communities, he also stated, "Xeric prairie and glade floras have evolved with drought and fire and need one or both to survive and compete with or displace encroaching species." According to Hicks (1981), "Prior to settlement by people of European origin, fires, either naturally occurring or set by Indians, and grazing by large herbivores such as elk and bison, now absent in the area, may have played an important role in maintaining the glade ecosystem." Kurz (1981) offers two possible explanations for the origin of sites in the glaciated portion of Illinois, both of which are plausible and not mutually exclusive.
The slope of the glades usually faces south and it is suggested that the wind-blown silt [loess] was not deposited in these areas because of their position. Another suggestion is that the soils eroded over a period of time because of some disturbance leaving little loess covering the bedrock.
According to Aldrich et al. (1982), the persistence of limestone glades in Harrison County, Indiana, "... involves a complex interplay of factors including thin soils, a southerly aspect, seven to twenty degree slopes, xeric conditions and fire." Homoya (1994) attributes the xeric, open nature of barrens in Indiana to "... excessive drainage ...; southern or western aspects ...; excessive steepness of slope ... ; presence of a hardpan ...; presence of bedrock at or near the surface ...; and a fire regime that increases radiant heating and consumes moisture-retaining litter."
Mann et al. (1999) and Lowell and Astroth (1989) developed geographic information system models to predict locations of XLPs in the Cedar Creek Preserve in Fort Knox Military Reservation (Kentucky) and the Hercules Glade Wilderness Area in the Mark Twain National Forest (Missouri), respectively. Soil series were not a significant indicator of this habitat in Cedar Creek Preserve. However, when soil subgroup and depth ("depth of lithologic contact") were included, efficacy of the model increased considerably. Incorporation of additional soils (presence of flagstone on surface) and geologic (formation/member) data further increased predictive capability of the Mann et al. (1999) model. Presence of a mollic epipedon was determined to be the most diagnostic characteristic of "threatened calcareous habitat" in Cedar Creek Preserve. In contrast to the Mann et al. (1999) model, Lowell and Astroth (1989) found soil series (Gasconade) to be the best predictor of glade distribution in Hercules Glade Wilderness Area. Elevation and aspect of slope also were strong "controlling factors" of distribution and quality (high quality = limited redcedar invasion; low quality = extensive redcedar invasion) of XLPs in this region.
Several studies of aerial photographs obtained over relatively short chronological sequences (<50 years) suggest XLPs are quite susceptible to woody plant invasion. Laughlin (2004) reported areal reductions of 78% and 92% in Great Plains and Westfall Ridge "prairies," respectively, in the Ridge and Valley of Pennsylvania from 1949 to 1994. In limestone glades in the Ozarks of south-central Missouri, Kimmel and Probasco (1980) reported a 34% decrease in glade area with 0% to 15% woody plant cover and a 31% increase in glade area with 50% to 100% woody plant cover from 1938 to 1975. The authors attributed this phenomenon to decreased burning and grazing over the previous 30 years. Ver Hoef et al. (1993) documented average reductions in "glade" area of 14.4% and 32.4% from 1955 to 1984 on sites developed on Eminence Dolomite and Gasconade Dolomite, respectively. Rates of forest encroachment in XLPs in Hercules Glade Wilderness Area from 1938 to 1986 generally decreased when sites were burned (Lowell & Astroth, 1989). However, fire significantly increased the area of some low-quality sites (i.e., sites with significant encroachment of woody species) but failed to retard encroachment and corresponding decline in other low-quality sites.
Referring to limestone glades in Harrison County, Indiana, Aldrich et al. (1982) stated, "Aerial photographs document the continuing shrinkage of these glades, as they were nearly double their present size in the 1940's." Similarly, Bacone et al. (1983) reported, "Aerial photographs [of glades and barrens in Crawford and Perry counties, Indiana] show a remarkable decrease in size over the last forty years due to encroachment by woody species." In two "prairies" developed on Cedarville Dolomite in Adams County, Ohio, Annala et al. (1983) documented a 31% reduction in prairie areas between 1938 and 1971. The authors concluded, "The edaphic conditions of the prairie remnants [Lynx Prairie and Bohl Property] in Adams County [Ohio; Edge of Appalachia Preserve] undoubtedly have helped to retain them. Nevertheless, the successional process clearly is converting these areas to forest." In other sites in the Edge of Appalachia Preserve (The Wilderness, Buzzardroost Rock, Hanging Prairie, and Cave Hollow), Annala and Kaputska (1983) reported "... some degree of forest encroachment in all sites" from 1938 to 1971. Particularly interesting were the authors' observations that, "Many of the present-day prairie remnants have been established (or reestablished) on cultivated or pasture land," and "other areas were prairie in 1938 and indicated no evidence of agricultural disturbance."
Despite the susceptibility of many XLPs to encroachment by woody species, some sites apparently are quite stable. Although large perimeter to area ratios and patches of woody vegetation prevented accurate measurement of the area of XLPs in Kentucky (Lawless et al., 2004), comparison of aerial photographs taken of Muldraugh's Barren and Scudder Glade (Hardin County, Kentucky) in 1958 and 1988 does not show dramatic changes in these two sites. However, proximity of old homesteads (circa 18501900) to the latter and to Spalding Glade (Lame County, Kentucky) (Lane Linnenkohl, Kentucky State Nature Preserves Commission, pers. commun.) does not rule out anthropogenic disturbance (e.g., timber harvesting and grazing) as a possible cause of origin for these and other sites in Kentucky. It also is not known whether disturbance is the reason why succession is not evident in these sites between 1958 and 1988.
Because endogenous factors such as bedrock type, soil depth, and degree and aspect of slope are relatively static over geologic time periods, some researchers have focused on the frequency of disturbance events, particularly fire, as a potential source of variability in the area of XLPs in recent times. Juniperus virginiana individuals often are the source of dendrochronological data in XLPs owing to their abundance and relatively long life spans. In a study of seven live and 14 dead redcedars (J. virginiana) in Burtram Hollow "glades" (Ava Ranger District, Missouri Ozarks), Guyette and McGinnes (1982) reported, "Prior to 1870, fires, as marked by tree scars, were present somewhere on the study area (2.59 [km.sup.2]) every 3.2 years. After 1870, the frequency of scarring drops to one every 22 years." In conclusion, the authors made the following statement.
Much of this change in fire frequency is due to fire suppression, possibly by early settlers, and after 1940 by the U.S. Forest Service. Other possible reasons for the change in fire frequency are the removal of a major ignition source (Osage Native Americans), the building of roads (firebreaks), severe overgrazing and erosion (reduced fuel loads).
Dendrochronological data also led Jenkins et al. (1997) to conclude that displacement of Native Americans (Cherokee) and subsequent fire suppression were largely responsible for increased woody plant density in a "savanna-glade-woodland mosaic" on Turkey Mountain in the Arkansas Ozarks. Variable fire-return intervals also were detected in the dendrochronological records obtained by Batek et al. (1999) in the Current River watershed in the Missouri Ozarks. Batek et al. made the following statement regarding the influence of Native Americans on historical fire frequencies and corresponding vegetation patterns.
Fires exerted strong constraints on vegetation composition and patterns. Historical patterns of Native American occupancy in the region are consistent with the reconstructed vegetation and fire histories suggest that anthropogenic fire regimes played an overriding role in the development of Ozark vegetation in the 1800's.
Although Beilmann and Brenner (1951) did not study dendrochronological records, their examination of numerous historical documents (particularly accounts of early travelers) led them to conclude, "Fire, perhaps more than any other factor, maintained the prairie and park-like aspect of the Ozarks." In a "barren-forest mosaic" in southern Indiana, Guyette et al. (2003) obtained dendrochronological results that contrast strongly with those of both Guyette and McGinnes (1982) and Batek et al. (1999). Guyette et al. (2003) summarized their findings as follows: "Fires were more frequent and occurred more regularly in the latter half of the fire chronology (1821 to 1999) where the mean fire interval was 5.3 years [as opposed to 23.0 years from 1650 to 1820] and fire intervals ranged from 1 to 40 years." Results of Guyette et al. (2003) suggest historic fire frequencies vary significantly between areas, and furthermore, that fuel loads, vegetation patterns, weather and climate, topography, and "pyro-cultural influence" profoundly affect fire-return intervals.
Boettcher and Kalisz (1991) used the mass and shape of opal phytoliths in dolomite-derived soils in an attempt to delimit the historical extent of XLPs in the E. Lucy Braun Preserve in Adams County, Ohio. Dumbbell-shaped phytoliths are characteristic of grasses, and therefore large concentrations of these structures are indicative of past and/or present grassland vegetation. Soil samples collected from XLPs and forests both had high concentrations of opal. However, dumbbell-shaped phytoliths were "infrequently encountered" in samples collected from both vegetation types, thus suggesting, "... phytolith-rich forbs as well as grasses may have been important constituents of these forest openings in the past." These data lead the authors to conclude, "... prairie and forest vegetation ... naturally alternated over time on these areas," and therefore, "... distinction between primary prairies [natural prairies] and secondary prairies [prairies formed by human disturbance of forests] has little meaning...."
Considerable reduction in the size of openings at many XLP sites over relatively short time periods (<50 years) support the conclusion that the overwhelming majority of them are quite susceptible to encroachment by woody species and thus are not primary. Dendrochronological studies conducted in this vegetation type suggest anthropogenically elevated fire frequencies associated with Native American cultures probably are responsible for maintenance (and possibly for the origin) of this vegetation type throughout much of the eastern United States (Fig. 3). However, Native American burning practices most likely resulted in establishment of XLPs only in areas with extremely shallow ([less than or equal to] 25 cm), rocky soils and high solar heat loads (moderate to steep slopes with south to west aspects). Prior to periodic anthropogenic disturbance, these areas would have supported relatively open-canopied, xerophytic forests that contained heliophytic herbaceous taxa characteristic of XLPs. Introduction of fire into these habitats would have decreased woody plant cover, resulting in increased insolation and thus in dominance by these heliophytic graminoids and forbs (see Jenkins & Jenkins, 1999).
In more mesic habitats, XLPs probably originated from clearing of forests for agricultural purposes (grazing, row crop cultivation, or timber harvesting). Such practices likely promoted soil erosion, especially on moderate to steep slopes, resulting in abandonment of these marginal agricultural lands. The more xeric conditions of this degraded habitat provided a suitable environment for colonization by the XLP flora (Baskin et al., 1994, 1997), which undoubtedly occurred at a rapid pace in areas adjacent to XLPs or open-canopied xerophytic forests.
In the absence of management (prescribed fire, mowing, cutting, grazing), succession to forested community types occurs in XLPs. Succession converts XLPs on deeper soils to hardwood forests in periods of a century or less. In more xeric sites, Juniperus virginiana/hardwood forests are the likely outcome of succession over prolonged time periods. The rate at which XLPs revert to forested community types via succession may differ between regions as a result of climate. For instance, in habitats of comparable soil depths and plant-available water holding capacities, succession may proceed much faster in the southern Interior Low Plateaus and Ridge and Valley than in the Ozark Plateaus owing to lower growing-season precipitation in the latter.
XLPs in the Cahaba River valley in Bibb County, Alabama, and perhaps those in some sites in the Ozark Plateaus in Missouri and Arkansas may be the only examples of this vegetation type that are edaphic climaxes. Although researchers have not studied aerial photographs or dendrochronological records in the Ketona Dolomite sites, their open nature at present suggests that they are able to persist in the absence of fire. Conversely, aerial photographic and dendrochronological data available for sites in the Ozark Plateaus suggest that these sites are relatively susceptible to encroachment by woody species in the absence of periodic fire. The heliophytic taxa endemic to XLPs in both of these regions [nine of 13 differentiated at the specific (versus subspecific or varietal) level] suggests that they could have persisted over geologic time spans.
Affinities of Xeric Limestone Prairies to Other Vegetation Types
EASTERN NORTH AMERICA
A number of other vegetation types in eastern North America are remarkably similar to XLPs because of their association with calcareous rock outcrops and dominance by [C.sub.4] perennial grasses. Dry lime prairies (sensu Anderson, 1954) and hill prairies (sensu Ugarte, 1987) in upper midwestern United States and "cedar glades" (sensu Curtis, 1959) in Wisconsin are developed on shallow, rocky soils derived from limestone and dolomite (calcareous gravels in some glaciated regions). Furthermore, these vegetation types typically occur on moderate to steep slopes with south to west aspects and are susceptible to invasion by woody species, particularly Juniperus virginiana.
Dry lime prairies are concentrated in the "rugged topography of the Driftless Area of southwestern Wisconsin" and extend into northwestern Illinois, northeastern Iowa, and southeastern Minnesota (Anderson, 1954). In 41 stands sampled by Anderson, Schizachyrium scoparium had the highest average frequency (71.3%), followed by Bouteloua curtipendula (62.4%) and Andropogon gerardii (50.2%). Additional taxa with mean frequency values >25% were Symphyotrichum sericeum, Euphorbia corollata, Amorpha canescens, Solidago nemoralis, Panicum perlongum, and Dalea purpurea. Rhus glabra and Juniperus virginiana were dominant woody invaders in dry lime prairies, along with Juniperus communis.
Schizachyrium scoparium, Andropogon gerardii, Sporobolus heterolepis, and Bouteloua curtipendula had the highest total cover values in 32 hill prairies in northeast Iowa sampled by Ugarte (1987). Solidago nemoralis, Aster azureus, Amorpha canescens, and Coreopsis palmata were characteristic forbs. Interestingly, Ugarte described a Bouteloua curtipendula-Sporobolus aft. vaginiflorus community type that developed on gravelly, heavily grazed sites. Furthermore, total cover of Juniperus virginiana exceeded that of all other woody species, with the exception of Rhus glabra. Ugarte attributed "encroachment of hill prairies by woody species" to "suppression of fire and overgrazing." Portions of some hill prairies described by Evers (1955) (e.g., Sunset Trail, Bielema, and Clendenny) apparently are shallow and rocky and are dominated by C4 perennial grasses, and thus they resemble XLPs.
According to Curtis (1959), "... a steep hillside of thin loess over limestone.., or a gravelly glacial moraine" are typical habitats for "cedar glades," which he and Bray (1955) considered a type of savanna. Schizachyrium scoparium had the highest average frequency (38.0%) in this vegetation type, followed by Euphorbia corollata (27.1%), Andropogon gerardii (26.6%), Arenaria stricta (23.0%), Solidago nemoralis (20.8%), Amorpha canescens (20.8%), and Anemone cylindrica (20.1%). Juniperus virginiana is the only dominant woody taxon in this vegetation type (data of J.R. Bray from Curtis, 1959).
"Calcareous prairies" in the Kisatchie National Forest in Louisiana, described by MacRoberts and MacRoberts (1993), are difficult to characterize owing to the complex geology in this region and lack of information on vegetation and soil depth. "Calcareous concretions" were reported in the soil surface in both sites. However, rock outcrops in proximity to these sites are primarily sandstone and support glade communities. The calcareous prairies "... are on or near the summits of hills ...," where soil depth is considerably greater than in adjacent sandstone glade communities (Michael MacRoberts, pers. commun.). Composites, grasses, and legumes are dominant taxa in these prairies, and both sites currently are being invaded by woody species (Crataegus spp., Rhus copallina, Viburnum dentatum, Comus drummondii, Diospyros virginiana, and Prunus spp.) from calcareous forests in the region. Both the calcareous prairies and sandstone glades are embedded within locally dominant longleaf pine savanna vegetation, and MacRoberts and MacRoberts (1993) attribute woody encroachment in the calcareous prairies to decreased fire frequency. Because of the uncertainty about whether these areas fit the definition of XLP, we have not included them on the distribution map (Fig. 2) of XLPs in eastern United States.
In eastern North America, alvars are restricted to the Great Lakes Region, where they are developed on Ordovician limestones and dolomites (Catling & Brownell, 1999). Catling and Brownell describe two principal alvar types: shoreline and plateau. They acknowledge disparate origins for these two community types in the following statement: "Lack of tree cover on shoreline alvars may be largely a result of flooding and erosion, whereas fire may be the most important factor in limiting woody cover on plateau alvars." Alvar grasslands most closely resemble XLPs and occur in both of these habitats. In contrast to alvar pavement community types, alvar grasslands typically occur in soils with depth >2 cm. Xeric grasslands are dominated by C3 (e.g., Carex scirpoidea, Danthonia spicata, and Poa compressa) perennial graminoids and are extremely variable with regard to forb composition (Catling & Brownell, 1999).
WESTERN NORTH AMERICA
In west-central Kansas, "very dry," "dry," and "dry mesic" prairies on rocky shallow soils (Mollisols) over limestone are floristically (at species or genus level) and vegetationally similar to the XLPs in eastern United States (Hulett & Tomanek, 1969; Heitschmidt et al., 1970; Hladek et al., 1972). Schizachyrium scoparium was the dominant species in remnant stands on very dry limy sites and was nearly as important as Andropogon gerardii in stands on dry sites (Hladek et al., 1972). Order of importance on dry mesic sites was Andropogon gerardii > Bouteloua curtipendula > Schizachyrium scoparium. Important forbs (all perennials) in these limestone prairies include Ambrosia psilostachya, Aster oblongifolius, Dalea purpurea, Echinacea angustifolia, Houstonia angustifolia, Lesquerella ovalifolia, Malvastrum coccineum, Oenothera serrulata, Paronychia jamesii, Psoralidium tenuifolium, Scutellaria resinosa, Solidago rigida, Tetraneuris stenophylla, Thelesperma gracile, and Tragia ramosa (Hulett & Tomanek, 1969; Heitschmidt et al., 1970; Hladek et al., 1972). Plant taxa endemic, or nearly so, to dry limestone prairies of the Mixed Grass Association of the southern Great Plains (southern Nebraska to northern Texas) include Aster fendleri, Clematis freemontii, Oenothera macrocarpa var. oklahomensis, O. freemontii, Phlox oklahomensis, Scutellaria resinosa, and Tomanthera densiflora (Great Plains Flora Association, 1977; 1986).
Mean annual precipitation in XLPs in eastern United States ranges from approximately 100 cm in Missouri (Missouri Climate Center, 2004) to approximately 150 cm in Alabama (Southeast Regional Climate Center, 2004). These precipitation regimes generally support forest vegetation in eastern United States (Braun, 1950), and thus the existence of nonforested herbaceous communities in this region (including XLPs) is not the result of low precipitation. Rather, these grasslands and rock outcrop communities are either edaphic climaxes (probably sites on Ketona Dolomite in the Ridge and Valley in Alabama and perhaps some sites in the Ozark Plateaus in Missouri and Arkansas in the case of XLPs), or they are the result of disturbance, either anthropogenic (vast majority of XLPs) or natural (e.g., scouring action of water bodies in riparian communities such as shoreline alvars and cobble bars).
Borchert (1950) attributed both the origin and maintenance of grasslands in the Central Plains and Prairies Formation to low mean annual precipitation and high seasonality in precipitation. Other researchers (e.g., Saner, 1950; Anderson, 1982; Axelrod, 1985) present cogent arguments for the combined effects of climate, fire, and grazing in establishment and maintenance of these grasslands. However, the rocky, shallow-soil grasslands in western and west-central Kansas (to which XLPs bear the closest resemblance) are an extraordinary case. In these sites, unlike the majority of prairies in central North America developed on deep, fertile Mollisols (Weaver, 1954), edaphic conditions are an important factor limiting recruitment, survival, and reproduction of woody taxa. The combination of these edaphic conditions with the low mean annual precipitation in this region (55 to 65 cm; Hladek et al., 1972; Heitschmidt et al., 1970) suggests that climatic and edaphic factors interact to produce the xeric conditions responsible for the absence or limited occurrence of woody taxa in these grasslands. Furthermore, low total plant cover (-10-25%; Heitschmidt et al., 1970; Hladek et al., 1972) in these rocky prairies makes it unlikely that they are even capable of supporting a fire.
Directions for Future Research
Additional studies are needed to further our understanding of the flora, vegetation, physical environmental characteristics, origin, and maintenance of XLPs in eastern United States. Phytogeographical analysis of the collective flora of this vegetation type in eastern United States would provide insight into the origins of the XLP flora and afford an opportunity for floristic comparisons among regions. Vegetation data are lacking for many regions, including the Central Lowland in Missouri and Indiana, the Interior Low Plateaus in Indiana, the Appalachian Plateaus in Ohio, and the Ridge and Valley in Bibb County, Alabama. Additional dendrochronological studies are necessary for a more complete evaluation of the historical role of fire in the origin and maintenance of XLPs. Additional soil chemical and physical analyses in XLPs and adjacent habitats could provide insight into the relative importance of edaphic conditions in curbing encroachment of woody taxa into XLPs. Analyses of soil profiles in XLPs may allow inferences about the factors (i.e., erosion and lack of profile development) responsible for current edaphic conditions. Aerial photographic studies, particularly for sites in southeastern United States, would allow further assessment of XLP stability under a variety of topographic, edaphic, and climatic conditions. Further comparisons (via floristic, vegetation, and physical environmental data) of XLPs, "cedar glades" in Wisconsin (sensu Curtis, 1959), and dry lime prairies (sensu Anderson, 1954) and loess hill prairies (sensu Ugarte, 1987) in the upper Midwest would facilitate development of a comprehensive classification system for xeric grasslands of eastern United States.
Ackerman, E. A. 1941. The Koppen classification of climates in North America. Geogr. Rev. 31: 105-111 + map.
Adams, C. A., J. M. Baskin & C.C. Baskin. 2003. Comparison of the morphology, flowering phenology, and life cycle type in plants of Grindelia lanceolata (Asteraceae) from cedar glade communities in middle Tennessee and northern Alabama: a common garden study. Sida 20: 1059-1071.
Aldrich, J. R., J. A. Bacone & M.D. Hutchison. 1982. Limestone glades of Harrison County, Indiana. Proc. Indiana Acad. Sci. 91: 48(~485.
Allison, J. R. & T. E. Stevens. 2001. Vascular flora of Ketona Dolomite outcrops in Bibb County, Alabama. Castanea 66: 154-205.
Anderson, O. 1954. The phytosociology of dry lime prairies of Wisconsin. Ph.D. diss., Univ. of Wisconsin, Madison.
Anderson, R. C. 1982. An evolutionary model summarizing the roles of fire, climate, and grazing animals in the origin and maintenance of grasslands: an end paper. Pp. 297-308 in J. R. Estes, R.J. Tyrl, and J. N. Brunken (eds.), Grasses and grasslands: systematics and ecology. Univ. of Oklahoma Press, Norman.
Annala, A. E. & L. A. Kaputska. 1983. Photographic history of forest encroachment in several relict prairies of the Edge of Appalachia preserve system, Adams County, Ohio. Ohio J. Sci. 83:109-114.
--, J. D. Dubois & L. A. Kaputska. 1983. Prairies lost to forest: a 33-year history of two sites in Adams County, Ohio. Ohio J. Sci. 83: 22-27.
Arkansas Natural Heritage Commission [ANHC]. 2004. Rare species search engine (http://www .naturalheritage.com/program/element-search/). Arkansas Natural Heritage Commission, Little Rock.
Axelrod, D. I. 1985. Rise of the grassland biome, central North America. Bot. Rev. 51: 163-201.
Bacone, J. A. & L.A. Casebere & M.D. Hutchison. 1983. Glades and barrens of Crawford and Perry counties, Indiana. Proc. Indiana Acad. Sci. 92: 367-373.
Barbour, M. G. & D. W. Billings. 2000. North American terrestrial vegetation. Ed. 2, Cambridge Univ. Press, Cambridge.
Bartgis, R. L. 1993. The limestone glades and barrens of West Virginia. Castanea 58: 69-89.
Baskin, C. C. & J. M. Baskin. 1975. The cedar glade flora of Bullitt County, Kentucky. Castanea 40: 184-190.
Baskin, J. M. & C.C. Baskin. 1977. An undescribed cedar glade community in middle Tennessee. Castanea 42: 140-145.
-- & --. 1978. On the occurrence of the cedar glade endemic Viola egglestonii in Kentucky. Trans. Kentucky Acad. Sci. 39: 74-75.
-- & --. 1986. Distribution and geographical/evolutionary relationships of cedar glade endemics in southeastern United States. A. S. B. Bull. 33:138-154.
-- & --. 1989. Cedar glade endemics in Tennessee, and a review of their autecology. J. Tennessee Acad. Sci. 64: 63-74.
--- & --. 1999. Cedar glades of the southeastern United States. Pp. 206-219 in R.C. Anderson, J. S. Fralish & J. M. Baskin (eds.), Savannas, barrens, and rock outcrop plant communities of North America. Cambridge Univ. Press, Cambridge.
-- & ---. 2000. Vegetation of limestone and dolomite glades in the Ozarks and Midwest regions of the United States. Ann. Missouri Bot. Gard. 87: 286-294.
-- & --. 2003. The vascular flora of cedar glades of the southeastern United States and its phytogeographical relationships. J. Torrey Bot. Soc. 130:101-118.
-- & --. 2004. History of the use of "'cedar glades" and other descriptive terms for vegetation on rocky limestone softs in the Central Basin of Tennessee. Bot. Rev. 70: 403-424.
--,-- & E. W. Chester. 1994. The Big Barrens Region of Kentucky and Tennessee: further observations and considerations. Castanea 59: 226-254.
--, K. M. Snyder & C.C. Baskin. 1993. Nomenclatural history and taxonomic status of Echinacea angustifolia, E. pallida, and E. tennesseensis (Asteraceae). Sida 15: 597-604. --, E. W. Chester & C. C. Baskin. 1997. Forest vegetation of the Kentucky Karst Plain (Kentucky and Tennessee): review and synthesis. J. Torrey Bot. Soc. 124: 322-335.
Batek, M. J., A. J. Rebertus, W. A. Sehroeder, T. L. Haithcoat, E. Compass & R. P. Guyette. 1999. Reconstruction of early nineteenth-century vegetation and fire regimes in the Missouri Ozarks. J. Biogeogr. 26: 397-412.
Beilmann, A. P. & L. G. Brenner. 1951. The recent intrusion of forests in the Ozarks. Ann. Missouri Bot. Gard. 38:261-282.
Belden, A. Jr., G. R. Fleming, J.C. Ludwig, J. F. Townsend, N.E. Van Alstine & T. F. Wideboldt. 2004. Noteworthy collections: Virginia. Castanea 69: 144-153.
Berg, T. M., J. H. Barnes, W. D. Sevon, V. W. Skema, J. P. Wilshusen & D. S. Yannacci. 1989. Physiographic provinces of Pennsylvania. Pennsylvania Geologic Survey, Ser. 4, Map 13. Department of Environmental Resources, Harrisburg, Pennsylvania.
Binns, S. E., B. R. Baum & J. T. Arnason. 2002. A taxonomic revision of Echinacea (Asteraceae: Heliantheae). Syst. Bot. 27: 610-632.
Boettcher, S. E. & P. J. Kalisz. 1991. The prairies of the E. Lucy Braun Preserve, Adams County, Ohio: a soil study. Ohio J. Sci. 91: 122-128.
Borchert, J. R. 1950. The climate of the central North American grassland. Ann. Assoc. Amer. Geogr. 40: 1-39.
Braun, E. L. 1928. The vegetation of the Mineral Springs Region of Adams County, Ohio. Ohio Biol. Surv. Bull. 15 (Vol. 3, No. 5): 375-517.
--. 1950. Deciduous forests of eastern North America. Blakiston, Philadelphia.
Bray, J. R. 1955. The savanna vegetation of Wisconsin and an application of the concepts of order and complexity to the field of ecology. Ph.D. diss., Univ. of Wisconsin, Madison.
Brockman, C. S. 1998. Physiographic regions of Ohio [Map]. Ohio geology (Summer edition, pp. 4-5). Ohio Department of Natural Resources, Columbus.
Catling, P. M. & V. R. Brownell. 1999. Alvars of the Great Lakes Region. Pp. 375-405 in R. C. Anderson, J. S. Fralish & J. M. Baskin (eds.), Savannas, barrens, and rock outcrop plant communities of North America. Cambridge Univ. Press, Cambridge.
Catling, P. M., A. A. Reznicek & W. J. Crins. 1993. Carex juniperorum (Cyperaceae), a new species from northeastern North America, with a key to Carex sect. Phyllostachys. Sys. Bot. 18: 496-501.
Clark, W. C. & A. C. Zisa. 1976. Physiographic map of Georgia. Georgia Department of Natural Resources, Atlanta.
Collins, J. L. & T. F. Wieboldt. 1992. Trifolium calcaricum (Fabaceae), a new clover from limestone barrens of eastern United States. Castanea 57: 282-286.
Curtis, J. T. 1959. The vegetation of Wisconsin. Univ. of Wisconsin Press, Madison.
DeSelm, H. R. 1988. The barrens of the Western Highland Rim of Tennessee. Pp. 199-219 in D. H. Snyder (editor), Proceedings of the first annual symposium on the natural history of lower Tennessee and Cumberland fiver valleys. The Center for Field Biology, Austin Peay State Univ., Clarksville, Tennessee.
--. 1989. The barrens of Tennessee. J. Tennessee Acad. Sci. 64: 89-95.
--. 1991. Barrens of the Central Basin of Tennessee. Pp. 1-26 in D. H. Snyder (ed.), Proceedings of the fourth annual symposium on the natural history of lower Tennessee and Cumberland river valleys. The Center for Field Biology, Austin Peay State Univ., Clarksville, Tennessee.
--. 1993. Barrens and glades of the southern Ridge and Valley. Pp. 81-136 in S. W. Hamilton, E. W. Chester & A. F. Scott (eds.), Proceedings of the fifth annual symposium on the natural history of lower Tennessee and Cumberland river valleys. The Center for Field Biology, Austin Peay State Univ., Clarksville, Tennessee.
-- & E. W. Chester. 1993. Further studies on the barrens of the northern and western Highland Rims of Tennessee. Pp. 137-160 in S.W. Hamilton, E. A. Chester & A. F. Scott (eds.), Proceedings of the fifth annual symposium on the natural history of lower Tennessee and Cumberland river valleys. The Center for Field Biology, Austin Peay State Univ., Clarksville, Tennessee.
-- & N. Murdock. 1993. Grass-dominated communities. Pp. 87-141 in W. H. Martin, S. G. Boyce & A. C. Echternacht (eds.), Biodiversity of the southeastern United States: upland terrestrial communities. John Wiley and Sons, New York.
-- & D. H. Webb. 1997. Vegetation studies of barrens in northwestern Alabama. Pp. 204-213 in A. F. Scott, S.W. Hamilton, E. W. Chester & D. S. White (eds.), Proceedings of the seventh annual symposium on the natural history of lower Tennessee and Cumberland fiver valleys. The Center for Field Biology, Austin Peay State Univ., Clarksville, Tennessee.
--, P. B. Whitford & J. S. Olson. 1969. The barrens of the Oak Ridge area, Tennessee. Amer. Midl. Naturalist 81 : 315-330.
Erickson, R. O., L. G. Brenner & J. Wraight. 1942. Dolomitic glades of east-central Missouri. Ann. Missouri Bot. Gard. 29: 89-101.
Evers, R. A. 1955. Hill prairies of Illinois. Bull. Illinois Nat. Hist. Survey 26: 367-446.
Fenneman, N. M. 1938. Physiography of eastern United States. McGraw-Hill, New York.
--. 1946 . Physical divisions of the United States [Map]. Prepared by Nevin M. Fenneman in cooperation with the Physiographic Committee of the Geological Survey, Washington, D.C.
Gardner, R. L. & D. Minnie. 2004. The vascular flora of the Richard and Lucile Durell Edge of Appalachia Preserve System. Ohio Chapter of The Nature Conservancy, Dublin.
George, J. R. 1996. A vegetative analysis of two central Missouri limestone glades. M.S. thesis, Central Missouri State Univ., Warrensburg.
Glenn, S. M. & S. L. Collins. 1990. Patch structure in tallgrass prairies: dynamics of satellite species. Oikos 57: 229-236.
Great Plains Flora Association [GPFA]. 1977. Atlas of the flora of the Great Plains. Iowa State Univ. Press, Ames.
--. 1986. Flora of the Great Plains. Univ. Press of Kansas, Lawrence.
Guyette, R. P. & E. A. McGinnes Jr. 1982. Fire history of an Ozark glade in Missouri. Trans. Missouri Acad. Sci. 16: 85-93.
--, D.C. Dey & M.C. Stambugh. 2003. Fire and human history of a barren-forest mosaic in southern Indiana. Amer. Midl. Naturalist 149:21-34.
Hall, M. T. 1955. Comparison of Juniper populations on an Ozark glade and old fields. Ann. Missouri Bot. Gard. 42: 171-194.
Harper, R. M. 1920. The limestone prairies of Wilcox County, Alabama. Ecology l: 198-203.
Heikens, A. L. 1991. Classification of the natural forest openings in southern Illinois. Ph.D. diss., Southern Illinois Univ., Carbondale.
--. 1999. Savanna, barrens, and glade communities of the Ozark Plateau province. Pp. 220-230 in R. C. Anderson, J. S. Fralish & J. M. Baskin (eds.), Savannas, barrens, and rock outcrop plant communities of North America. Cambridge Univ. Press, Cambridge.
-- & P.A. Robertson. 1995. Classification of barrens and other natural xeric forest openings in southern Illinois. Bull. Torrey Bot. Club 122: 203-214.
Heitschmidt, R. K., G. K. Hulett & G. W. Tomanek. 1970. Vegetation map and community structure of a west central Kansas prairie. S.W. Naturalist 14: 337-350.
Hicks, J. L. 1981. A vegetative analysis of Hercules Glade Wilderness. M.S. thesis, Southwest Missouri State Univ., Springfield.
Hladek, K. L., G. K. Hulett & G. W. Tomanek. 1972. The vegetation of remnant shale-limestone prairies in western Kansas. S.W. Naturalist 17: 1-10.
Hoagland, B. W., A. K. Buthod, I. H. Butler, P. H. C. Crawford, A. H. Udasi, W. J. Elisens & R. J. Tyrl. 2004. Oklahoma Vascular Plants Database (http://geo.ou.edu/botanical). Oklahoma Biological Survey, Univ. of Oklahoma, Norman.
Homoya, M. A. 1987. A floristic survey of a limestone glade in Versailles State Park, Ripley County, Indiana. Proc. Indiana Acad. Sci. 96: 407-412.
--. 1994. Indiana barrens: classification and description. Castanea 59:204-213.
--, D. B. Abrell, J. R. Aldrich & T. W. Post. 1984. The natural regions of Indiana. Proc. Indiana Acad. Sci. 94:245-268 + colored map of the natural regions of Indiana by M.A. Homoya.
Hulett, G. K. & G. W. Tomanek. 1969. Remnant prairies on the shallow limy range site in north central Kansas. J. Range Managem. 22: 19-23.
Jenkins, M. A. & S. E. Jenkins. 1999. Savanna and glade vegetation of Turkey Mountain, Arkansas: Effects of a single prescribed burn. Pp. 127-134 in C. Warwick (ed.), Proceedings of the fifteenth North American prairie conference. Natural Areas Association, Bend, Oregon.
Jenkins, S. E., R. P. Guyette & A. J. Rebertus. 1997. Vegetation-site relationships and fire history of a savanna-glade woodland mosaic in the Ozarks. Pp. 184-199 in S. G. Pallardy, R. A. Celich, H. G. Garrett & P. S. Johnson (eds.), Proceedings of the eleventh central hardwood conference, Univ. of Missouri, Columbia. USFS NCES GTR-NC 188.
Keeland, B. D. 1978. Vegetation and soils in calcareous glades of northwest Arkansas. M.S. thesis, Univ. of Arkansas, Fayetteville.
Keener, C. 1983. Distribution and biohistory of the endemic flora of the mid-Appalachian shale barrens. Bot. Rev. 49: 65-115.
Kim, D. H., D. Heber & D. W. Still. 2004. Genetic diversity of Echinacea species based upon amplified fragment length polymorphism markers. Genome 47:102-111.
Kimball, R. T., D. J. Crawford, J. R. Page & P. J. Harmon. 2001. Inter-simple sequence repeat (ISSR) diversity within Monarda fistulosa var. brevis (Lamiaceae) and var. fistulosa in West Virginia. Brittonia 53:511-518.
Kimmel, V. L. & G. E. Probasco. 1980. Change in woody cover on limestone glades between 1938 and 1975. Trans. Missouri Acad. Sci. 14:69-74.
Kucera, C. L. & S.C. Martin. 1957. Vegetation and soil relationships in the glade region of the southwestern Missouri Ozarks. Ecology 38: 285-291.
Kurz, D. R. 1981. Flora of limestone glades in Illinois. Pp. 183-185 in R. L. Stuckey & K. J. Reese (eds.), The Prairie Peninsula--In the "shadow" of Transeau. Proceedings of the sixth North American prairie conference, The Ohio State Univ., Columbus. Ohio Biol. Surv. Biol. Notes No. 15.
Ladd, D. & P. Nelson. 1982. Ecological synopsis of Missouri glades. Pp. 1-20 in E.A. McGiness, Jr. (ed.), Proceedings of the cedar glade symposium. Occas. Pap. Missouri Acad. Sci. 7.
Laughlin, D. C. 2002. Limestone prairies and the threatened grass, Bouteloua curtipendula, in Pennsylvania. M.S. thesis, The Pennsylvania State Univ., University Park.
--. 2004. Woody plant invasion and the importance of anthropogenic disturbance within xeric limestone prairies. J. Pennsylvania Acad. Sci. 78: 12-28.
-- & C.F. Uhl. 2003. The xeric limestone prairies of Pennsylvania. Castanea 68:300-316. Lawless, P. J. 2005. Xeric limestone prairies of eastern United States. Ph.D. diss., Univ. of Kentucky, Lexington.
--, J. M. Baskin & C.C. Buskin. 2004. The floristic ecology of xeric limestone prairies in Kentucky, and a comparison to limestone cedar glades and deep-soil barrens. Sida 21: 1055-1079.
--, -- & --. In press. Scale-dependent classification of xeric limestone prairies: annual or perennial grasslands? Ann. Missouri Bot. Gard.
Logan, J. M. 1992. The glades of the Buffalo National River, Arkansas. M.S. thesis, Iowa State Univ., Ames.
Lowell, K. E. & J. H. Astroth. 1989. Vegetative succession and controlled fire in a glades ecosystem. Int. J. Geogr. Inform. Syst. 3: 69-81.
Ludwig, J.C. 1999. The flora of dolomite and limestone barrens in southwestern Virginia. Castanea 64: 209-230.
MacRoberts, B. R. & M. H. MacRoberts. 1993. Floristics of two Louisiana sandstone glades. Phytologia 74: 431-437.
Mann, L. K., A. W. King, V. H. Dale, W. H. Hargrove, R. Washington-Allen, L. R. Pounds & T. L. Ashwood. 1999. The role of soil classification in geographic information system modeling of habitat pattern: threatened calcareous ecosystems. Ecosystems 2: 524-538.
Markham, C. G. 1970. Seasonality of precipitation in the United States. Ann. Assoc. Amer. Geogr. 60: 593-597.
Maxwell, R. H. 1987. Natural area remnants within the Indiana army ammunition plant, Charlestown, Indiana: the little bluestem glades. Proc. Indiana Acad. Sci. 96:413-419.
McClain, W. E. & J. E. Ebinger. 2002. A comparison of the vegetation of three limestone glades in Calhoun County, Illinois. S.E. Naturalist (Steuben) l: 179-188.
Missouri Climate Center [MCC]. 2004. http://www.hprcc.unl.edu/wrcc/states/mo.html. Missouri Climate Center, Univ. of Missouri-Columbia.
National Climatic Data Center [NCDC]. 2001. Mean annual precipitation (inches) for the United States based on normal period 1961-1990. National Climatic Data Center, Asheville, North Carolina.
Nelson, P. & D. Ladd. 1982a. Missouri glades: part I. Missouriensis 3(3): 5-9.
-- & --. 1982b. Missouri glades: part II. Missouriensis 3(4): 6-11.
-- & --. 1983. Preliminary report on the identification, distribution and classification of Missouri glades. Pp 59-76 in C.L. Kucera (ed.), Proceedings of the seventh North American prairie conference. Southwest Missouri State Univ., Springfield.
Quarterman, E. & R. L. Powell. 1978. Potential ecological/geological natural landmarks of the Interior Low Plateaus. U.S. Department of the Interior, Washington, D.C.
Rhoades, C. C., S. P. Miller, & M. M. Shea. 2004. Soil properties and soil nitrogen dynamics of prairie-like forest openings and surrounding forests in Kentucky's Knobs region. Amer. Midl. Naturalist 152: 1-11.
--, -- & D. L. Skinner. 2005. Forest vegetation and soil patterns across glade-forest ecotones in the Knobs Region of northeastern Kentucky, USA. Amer. Midl. Naturalist 154: 1-10.
Sapp, C.D. & J. Emplaincourt. 1975. Physiographic regions of Alabama. Map #168. Alabama Geological Survey, Tuscaloosa.
Sauer, C. 1950. Grassland climax, fire and man. J. Range Managem. 3: 16-20.
Schwegman, J. E. 1973. Comprehensive plan for the Illinois nature preserves system. Part 2. The natural divisions of Illinois. Illinois Nature Preserves Commission, Rockford. 32 p + map.
Skinner, B. R. 1979. Location and site requirements of three endangered/threatened Missouri glade plant species. M.S. thesis, Univ. of Missouri, Columbia.
--, G. E. Probasco & F. B. Samson. 1983. Environmental requirements of three threatened plants on limestone glades in southern Missouri. Biol. Conservation 25: 63-73.
Southeast Regional Climate Center [SERCC]. 2004. http://www.dnr.state.sc.us/climate/sercc/climate info/historical/historical.html. Southeast Regional Climate Center, Columbia, South Carolina.
Steyermark, J. A. 1940. Studies on the vegetation of Missouri. I. Natural plant associations and succession in the Ozarks of Missouri. Publ. Field Mus. Nat. Hist., Bot. Ser. 9: 348-475.
Swinford, E. M. 1985. Geology of the Peebles quadrangle, Adams County, Ohio. Ohio J. Sci. 85: 218-230.
Thom, R .H. & J. H. Wilson. 1980. The natural divisions of Missouri: an introduction to the natural history of the state. Pp. 1-11 in C. L. Kucera (ed.), Proceedings of the seventh North American prairie conference, Southwest Missouri State Univ., Springfield.
Trammell, T. L. E., C.C. Rhoades, & P.A. Bukaveckas. 2004. Effects of prescribed fire on nutrient pools and losses from glades occurring within oak-hickory forests of central Kentucky. Restorat. Ecol. 12: 597-604.
Trewartha, G. T. 1968. An introduction to climate. Ed. 4. McGraw-Hill, New York.
Ugarte, E. A. 1987. Hill prairies of northeast Iowa: vegetation and dynamics. Ph.D. diss., Iowa State Univ., Ames.
United States Department of Agriculture, Natural Resources Conservation Service [USDA, NRCS]. 2004. The PLANTS database, Version 3.5 (http://plants.usda.gov). National Plant Data Center, Baton Rouge, Louisiana.
United States Fish and Wildlife Service [USFWS]. 1988. Lesquerella filiformis recovery plan. United States Fish and Wildlife Service, Twin Cities, Minnesota.
Ver Hoef, J. M., S.R. Reiter & D.C. Glenn-Lewin. 1993. Glades of the Ozark Scenic Riverways, Missouri. A report to the National Park Service, U.S. Department of the Interior, Midwest Region, Omaha, Nebraska. NPS Cooperative Agreement CA 6640-8-8008.
Ware, S. W. 2002. Rock outcrop plant communities (glades) in the Ozarks: a synthesis. S.W. Naturalist 47: 585-597.
Weaver, J. E. 1954. North American prairie. Johnsen Publishing, Lincoln, Nebraska. & F. E. Clements. 1938. Plant ecology. Ed. 2. McGraw-Hill, New York.
Webb, D. H., H. R. DeSelm & M. Dennis. 1997. Studies of prairie barrens of northwestern Alabama. Castanea 62: 173-184.
White, D., B. Palmer-Ball Jr. & R. R. Cicerello. 1994. An endangered species survey of Fort Knox Military Reservation, Kentucky [technical report]. Kentucky State Nature Preserves Commission, Frankfort.
Willman, H. B., J.C. Frye, J. A. Simon, K. E. Clegg, D. H. Swann, E. Atherton, C. Collinson, J.A. Lineback & T. C. Buschbach. 1967. Geologic map of Illinois. Illinois State Geological Survey, Urbana.
Woodward, S. L. & R. L. Hoffman. 1991. The nature of Virginia. Pp. 23-48 + plates 1-24 in Virginia's endangered species, Coord. K. Terwillinger. The McDonald and Woodward Publ. Co., Blacksburg, Virginia. Figure 3 is a map of the physiographic provinces of Virginia drafted by S.L. Woodward (pets. commun, with J. M. Baskin).
PATRICK J. LAWLESS, (1) JERRY M. BASKIN, (1) AND CAROL C. BASKIN (2)
(1) Department of Biology University of Kentucky Lexington, Kentucky 40506-0225, U.S.A.
(2) Department of Plant and Soil Sciences University of Kentucky Lexington, Kentucky 40546-0312, U.S.A.
Table I Summary of the physiographic distribution of xeric limestone prairies in eastern United States organized by state and physiographic province (sensu Fenneman, 1938,  1949) Physiographic/natural State Physiographic province region Alabama Interior Low Plateaus Little Mountain Interior Low Plateaus Moulton Valley Interior Low Plateaus Southern Highland Rim Ridge and Valley Cahaba Valley Coastal Plain Fall Line Hills Coastal Plain Southern Red Hills Arkansas Ozark Plateaus Springfield Plateau Ozark Plateaus Salem Plateau Ozark Plateaus Boston Mountains Georgia Ridge and Valley Chickamauga Valley Illinois Central Lowland Chicago Lake Plain Central Lowland Western Grand Prairie Central Lowland Glaciated Middle Mississippi Border Ozark Plateaus Northern Section of Ozark Division Ozark Plateaus Southern Section of Ozark Division Interior Low Plateaus Lesser Shawnee Hills Indiana Central Lowland Muscatatuck Flats and Canyons Central Lowland Switzerland Hills Interior Low Plateaus Mitchell Plain Interior Low Plateaus Mammoth Cave Plateau Kentucky Interior Low Plateaus Marion [Area] Interior Low Plateaus Mammoth Cave Plateau Interior Low Plateaus Ohio River Hills and Lowlands Interior Low Plateaus Pennyroyal Plain Interior Low Plateaus Knobstone Escarpment & Knobs Interior Low Plateaus Outer Blue Grass Missouri Central Lowland Osage Plains Central Lowland Eastern Glaciated Plains Central Lowland Lincoln Hills Ozark Plateaus Springfield Plateau Ozark Plateaus Upper Ozark Ozark Plateaus St. Francois Mountains Ozark Plateaus Elk River Ozark Plateaus White River Ozark Plateaus Lower Ozark Ozark Plateaus Ozark Border Ohio Interior Low Plateaus Northeastern Blue Grass Appalachian Plateaus Marietta Plateau Pennsylvania Ridge and Valley Appalachian Mountain Tennessee Interior Low Plateaus Western Higland Rim Interior Low Plateaus Cumberland River Basin Ridge and Valley -- (a) Virginia Ridge and Valley Allegheny Mountains West Virginia Ridge and Valley -- (a) State Physiographic reference Site reference(s) Alabama Quarterman & Powell, 1978 DeSelm & Webb, 1997; Webb et al., 1997 Quarterman & Powell, 1978 DeSelm, 1997; Webb et al., 1997 Quarterman & Powell, 1978 Webb et al., 1997 Sapp & Emplaincourt, 1975 Allison & Stevens, 2001 Sapp & Emplaincourt, 1975 Webb et al., 1997 Sapp & Emplaincourt, 1975 Harper, 1920 Arkansas Fenneman, 1938, Keeland, 1978  1949 Fenneman, 1938, Keeland, 1978;  1949 Logan, 1992 Fenneman, 1938, Keeland, 1978  1949 Georgia Clark & Zisa, 1976 DeSelm, 1993 Illinois Schwegman, 1973 Kurz, 1981 Schwegman, 1973 Kurz, 1981; McClain & Ebinger, 2002 Schwegman, 1973 Kurz, 1981 Schwegman, 1973 Kurz, 1981; Heikens & Robertson, 1995 Schwegman, 1973 Kurz, 1981; Heikens & Robertson, 1995 Schwegman, 1973 Kurz, 1981; Heikens & Robertson, 1995 Indiana Homoya et al., 1984 Maxwell, 1987 Homoya et al., 1984 Homoya, 1987 Quarterman & Powell, 1978 Aldrich et al., 1982 Quarterman & Powell, 1978 Bacone et al., 1983 Kentucky Quarterman & Powell, 1978 Lawless et al., 2004 Quarterman & Powell, 1978 Lawless et al., 2004 Quarterman & Powell, 1978 Lawless et al., 2004 Fenneman, 1938, Lawless et al., 2004  1949 Quarterman & Powell, 1978 Lawless et al., 2004; Rhoades et al., 2004 Quarterman & Powell, 1978 Lawless et al., 2004; Rhoades et al., 2005 Missouri Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Thom & Wilson, 1980 Ladd & Nelson, 1982 Ohio Quarterman & Powell, 1978 Braun, 1928 Brockman, 1998 Rick Gardner, pers. commun. Pennsylvania Berg et al., 1989 Laughlin & Uhl, 2003 Tennessee Quarterman & Powell, 1978 DeSelm & Chester, 1993 Quarterman & Powell, 1978 Baskin & Baskin, 1977; -- (a) DeSelm, 1993 Virginia Woodward & Hoffman, 1991 Ludwig, 1999 West Virginia -- (a) Bartgis, 1993 (a) Not reported. Table II Summary of the geologic substrates upon which xeric limestone prairies occur in eastern United States. Formations and members are organized by physiographic province and geologic system(s), respectively Geologic formation/member System(s) State(s) Reference(s) Ozark Plateaus St. Louis Limestone Mississippian Illinois Heikens, 1991 Salem Limestone Mississippian Illinois Heikens, 1991 Burlington Limestone Mississippian Missouri Nelson & Ladd, 1983 Keokuk Limestone Mississippian Missouri Nelson & Ladd, 1983 Boone Formation Mississippian Arkansas Theo Witsell, (limestone) pers. commun. Pitkin Limestone Mississippian Arkansas Theo Witsell, pers. commun. Kimswick Formation Ordovician Missouri George, 1996; (limestone) Arkansas Theo Witsell, pers. commun. Plattin Formation Ordovician Missouri George, 1996; (limestone) Arkansas Logan, 1992 Joachim Dolomite Ordovician Missouri Erickson et al., 1942; Arkansas Theo Witsell, pers. commun. Everton Formation Ordovician Missouri Erickson et (dolomite) al., 1942; Arkansas Theo Witsell, pers. commun. Powell Dolomite Ordovician Missouri Erickson et al., 1942; Arkansas Theo Witsell, pers. commun. Cotter Dolomite Ordovician Missouri Erickson et al., 1942; Arkansas Theo Witsell, pers. commun. Jefferson City Ordovician Missouri Erickson et Dolomite al., 1942; Arkansas Theo Witsell, pers. commun. Gasconade Dolomite Ordovician Missouri Erickson et al., 1942; Ver Hoef et al., 1993 Eminence Dolomite Cambrian Missouri Ver Hoef et al., 1993 Potosi Dolomite Cambrian Missouri Ver Hoef et al., 1993 Derby-Doerun Dolomite Cambrian Missouri Ladd & Nelson, 1982 Bonneterre Formation Cambrian Missouri Ladd & Nelson, (dolomite) 1982 Central Lowland Keokuk Limestone Mississippian Missouri Ladd & Nelson, 1982 Burlington Limestone Mississippian Missouri Ladd & Nelson, 1982 Chouteau Limestone Mississippian Illinois Willman et al., 1967 Salem Limestone Mississippian Illinois Willman et al., 1967 St. Louis Limestone Mississippian Illinois Willman et al., 1967 Ste. Genevieve Mississippian Illinois Willman et al., Limestone 1967 Kimswick Formation Ordovician Missouri Ladd & Nelson, (limestone) 1982 Plattin Formation Ordovician Missouri Ladd & Nelson, (limestone) 1982 Dillsboro Formation Ordovician Indiana Homoya, 1987 (limestone and shale) Whitewater Formation Ordovician Indiana Homoya, 1987 (limestone and shale) Interior Low Plateaus Frailey Shale Mississippian Illinois Heikens, 1991 Haney Limestone Mississippian Illinois Heikens, 1991 Renault Limestone Mississippian Illinois Heikens, 1991 Bethel Limestone Mississippian Illinois Heikens, 1991 Ridenhoen Formation Mississippian Illinois Heikens, 1991 (limestone) Salem Limestone Mississippian Kentucky Lawless et al., 2004 Glen Dean Limestone Mississippian Kentucky Lawless et al., 2004 Reelsville Limestone Mississippian Kentucky Lawless et al., 2004 Beech Creek Limestone Mississippian Kentucky Lawless et al., 2004 Girkin Limestone Mississippian Kentucky Lawless et al., 2004 Paint Creek Limestone Mississippian Kentucky Lawless et al., 2004 Warsaw Limestone Mississippian Tennessee DeSelm & Chester, 1993 Fort Payne Formation Mississippian Tennessee DeSelm & (limestone) Chester, 1993 Tuscumbia Limestone Mississippian Alabama Webb et al., 1997 Bangor Limestone Mississippian Alabama Webb et al., 1997 Louisville Limestone Silurian Kentucky Lawless et al., 2004 Peebles Dolomite Silurian Ohio Swinford, 1985; Gardner & Minnie, 2004 Lilley Formation Silurian Ohio Swinford, 1985; (dolomite) Gardner & Minnie, 2004 Bisher Formation Silurian Ohio Swinford, 1985; (dolomite) Gardner & Minnie, 2004 Estill Shale Silurian Ohio Swinford, 1985; Gardner & Minnie, 2004 Upper Part of Crab Silurian Kentucky Lawless et al., Orchard Formation/ 2004 Lower Part of Crab Orchard and Brassfield Formation (clay shale, limestone, and dolomite) Laurel Formation Silurian Kentucky Rhoades et al., (shale and 2004 dolomite) Lebanon Limestone Ordovician Tennessee DeSelm 1988, 1991 Ridley Limestone Ordovician Tennessee DeSelm 1988, 1991 Ridge and Valley Keyser Formation/ Devonian, Pennsylvania Steve Grund & Tonoloway Formation Silurian John Kunsman, (undivided) pers. commun. (limestone) Onondoga Formation/ Devonian Pennsylvania John Kunsman, Old Fort Formation pers. commun. (undivided) (limestone) Chambersburg Ordovician Pennsylvania John Kunsman, Formation pers. (limestone) Coburn Formation Ordovician Pennsylvania John Kunsman, through Nealmont pers. commun. Formation (undivided) (limestone) Reedsville Formation Ordovician Pennsylvania John Kunsman, (limestone) pers. commun. Tytoona Cave Ordovician Pennsylvania Steve Grand, (limestone) pers. commun. Benner Formation/ Ordovician Pennsylvania Steve Grund & Loysburg Formation John Kunsman, (undivided) pers. commun. (limestone) Bellefonte Formation/ Ordovician Pennsylvania Steve Grund, Axemann Formation pers. commun (undivided) (limestone and dolomite) Tonoloway Limestone Ordovician West Virginia Bartgis, 1993 Helderberg Limestone Ordovician West Virginia Bartgis, 1993 Ben Hur Limestone Ordovician Virginia Ludwig, 1999 Knox Group Dolomite Ordovician Virginia DeSelm, 1993 Chickamauga Limestone Ordovician Alabama, DeSelm, 1993; Georgia, DeSelm et Tennessee, al., 1969; DeSelm, 1989 Virginia Ludwig, 1999 Conasauga Formation Ordovician Alabama DeSelm, 1993 (limestone) Mines Limestone Cambrian Pennsylvania Steve Grund, pers. Commun. Ketona Formation Cambrian Alabama Allison & (dolomite) Stevens, 2001 Rome Formation Cambrian Virginia Ludwig, 1999 (limestone) Honaker Formation Cambrian Virginia Ludwig, 1999 (limestone and dolomite) Coastal Plain Midway Limestone Tertiary Alabama Harper, 1920 (Eocene) Appalachian Plateaus Conemaugh Formation Pennsylvanian Ohio Rick Gardner, (limestone) pers. commun. Table III Summary of the soil series upon which xeric limestone prairies occur in eastern United States. Series are organized alphabetically by order. Physiographic province(s) and state(s) in which each series occurs are provided along with reference(s) Soil series Subgroup Physiographic province Alfisols Beasley Typic Hapludalfs Interior Low Plateaus Bratton Typic Hapludalfs Interior Low Plateaus Caneyville Typic Hapludalfs Interior Low Plateaus Colbert Vertic Hapludalfs Ridge and Valley Interior Low Plateaus Interior Low Plateaus Conasauga Oxyaquic Hapludalfs Ridge and Valley Crider Typic Paleudalfs Central Lowland Cumberland Rhodic Paleudalfs Interior Low Plateaus Eden Typic Hapludalfs Central Lowland Elba Typic Hapludalfs Appalachian Plateaus Fredonia Typic Hapludalfs Interior Low Plateaus Hagerstown Typic Hapludalfs Interior Low Plateaus Interior Low Plateaus Ridge and Valley Lenberg Ultic Hapludalfs Interior Low Plateaus Lowell Typic Hapludalfs Appalachian Plateaus Opequon Lithic Hapludalfs Interior Low Plateaus Ridge and Valley Rosine Ultic Hapludalfs Interior Low Plateaus Seaton Typic Hapludalfs Central Lowland Shrouts Typic Hapludalfs Interior Low Plateaus Stookey Typic Hapludalfs Interior Low Plateaus Ozark Plateaus Talbott Typic Hapludalfs Interior Low Plateaus Vertrees Typic Paleudalfs Interior Low Plateaus Wellston Ultic Hapludalfs Interior Low Plateaus Zanesville Oxyaquic Fragiudalfs Interior Low Plateaus Ultisols Bodine Typic Paleudults Interior Low Plateaus Clarksville Typic Paleudults Ozark Plateaus Elliber Typic Hapludults Ridge and Valley Gilpin Typic Hapludults Interior Low Plateaus Lebanon Typic Fragiudults Ozark Plateaus Nixa Glossic Fragiudults Ozark Plateaus Mollisols Barfield Lithic Hapludolls Interior Low Plateaus Corydon Lithic Argiudolls Interior Low Plateaus Gasconade Lithic Hapludolls Ozark Plateaus Ozark Plateaus Gladeville Lithic Rendolls Interior Low Plateaus Ridge and Valley Sogn Lithic Haplustolls Central Lowland Inceptisols Garmon Dystric Eutrudepts Interior Low Plateaus Muskingum Typic Dystrudepts Interior Low Plateaus Sulphura Typic Dystrudepts Interior Low Plateaus Vertisols Houston Oxyaquic Hapluderts Coastal Plain Soil series State Reference Alfisols Beasley Kentucky Lawless et al., 2004; Rhoades et al., 2004 Bratton Ohio Gardner & Minnie, 2004 Caneyville Kentucky Lawless et al., 2004; Rhoades et al., 2004; Mann et al., 1999 Colbert Alabama, Georgia, DeSelm, 1993 Tennessee, Virginia Alabama Webb et al., 1997 Kentucky Lawless et al., 2004 Conasauga Alabama, Georgia, DeSelm, 1993 Tennessee, Virginia Crider Indiana Maxwell, 1987 Cumberland Kentucky Lawless et al., 2004 Eden Indiana Homoya, 1987 Elba Ohio Rick Gardner, pers. commun. Fredonia Kentucky Lawless et al., 2004 Hagerstown Indiana Bacone et al., 1983 Kentucky Lawless et al., 2004; Mann et al., 1999 Pennsylvania Laughlin & Uhl, 2003 Lenberg Kentucky Lawless et al., 2004 Lowell Ohio Rick Gardner, pers. commun. Opequon Ohio Gardner & Minnie, 2004 Pennsylvania Laughlin & Uhl, 2003 Rosine Kentucky Lawless et al., 2004 Seaton Illinois McClain & Ebinger, 2002 Shrouts Kentucky Lawless et al., 2004 Stookey Illinois Heikens, 1991 Illinois Heikens, 1991 Talbott Tennessee DeSelm, 1991 Vertrees Kentucky Lawless et al., 2004 Wellston Indiana Bacone et al., 1983 Zanesville Indiana Bacone et al., 1983 Ultisols Bodine Tennessee DeSelm & Chester, 1993 Clarksville Missouri Skinner et al., 1983 Elliber Pennsylvania Laughlin & Uhl, 2003 Gilpin Indiana Bacone et al., 1983 Kentucky Lawless et al., 2004 Lebanon Missouri Skinner et al., 1983 Nixa Missouri Skinner et al., 1983 Mollisols Barfield Tennessee DeSelm, 1991 Corydon Indiana Homoya, 1994 Kentucky Lawless et al., 2004; Mann et al., 1999 Gasconade Missouri Hicks, 1981; George, 1996 Arkansas Keeland, 1978 Gladeville Tennessee DeSelm, 1991 Tennessee Mann et al., 1999 Sogn Illinois McClain & Ebinger, 2002 Inceptisols Garmon Kentucky Lawless et al., 2004; Mann et al., 1999 Muskingum Indiana Bacone et al., 1983 Sulphura Tennessee DeSelm & Chester, 1993 Vertisols Houston Alabama Harper, 1920 Table IV Summary of dominant taxa in xeric limestone prairies of eastern United States. Data (quantitative and qualitative) are organized by physiographic province (sensu Fenneman 1938. 1946 ) Physiographic province Dominant taxon Reference Data Ozark Plateaus Sporobolus Hall, 1955 frequency neglectus Schizachyrium Kucera & mean cover scoparium Martin, 1957 Schizachyrium Keeland, 1978 mean cover scoparium Schizachyrium Skinner, 1979 mean frequency scoparium (a) Sporobolus Skinner, 1979 mean frequency neglectus (b) Schizachyrium Hicks, 1981 mean importance scoparium value Schizachyrium Logan, 1992 semi- scoparium quantitative Schizachyrium Ver Hoef et geometric scoparium al., 1993 mean of cover-class midpoint Panicum virgatum George, 1996 mean cover Schizachyrium George, 1996 mean cover scoparium Central Lowland Schizachyrium Maxwell, 1987 qualitative scoparium Schizachyrium Heikens, 1991 mean cover scoparium Schizachyrium McClain & mean cover scoparium Ebinger, 2002 Bouteloua McClain & mean cover curtipenduda Ebinger, 2002 Hedyotis McClain & mean cover nigricans Ebinger, 2002 Interior Low Plateaus Schizachyrium Braun, 1928 mean frequency scoparium Andropogon Braun, 1928 mean frequency gerardii Schizachyrium Baskin & frequency scoparium Baskin, 1977 Schizachyrium Kurz, 1981 mean frequency scoparium (c) Schizachyrium DeSelm, 1988 importance scoparium value Schizachyrium Heikens, 1991 mean cover scoparium Silphium Heikens, 1991 mean cover terebinthinaceum Schizachyrium DeSelm, 1991 mean cover scoparium Schizachyrium DeSelm & Webb, mean cover scoparium 1997 Sporobolus DeSelm & Webb, mean cover clandestinus 1997 Schizachyrium Lawless et al., mean cover scoparium in press Andropogon Lawless et al., mean cover gerardii in press Silphium Lawless et al., mean cover terebinthinaceum in press Ridge and Valley Bouteloua Bartgis, 1993 mean cover curtipendula Solidago arguta Bartgis, 1993 qualitative var. harrisii Monarda fistulosa Bartgis, 1993 qualitative var. brevis Paronychia Bartgis, 1993 qualitative virginica Schizachyrium DeSelm, 1993 mean cover scoparium Andropogon DeSelm, 1993 mean cover gerardii Schizachyrium Ludwig, 1999 mean cover scoparium Andropogon Ludwig, 1999 mean cover gerardii Schizachyrium Allison & qualitative scoparium Stevens, 2001 Bouteloua Laughlin & Uhl, semi- curtipendula 2003 quantitative Coastal Plain Schizachyrium Harper, 1920 qualitative scoparium (a) Based on frequency data collected in 0.01- and 0.1-[m.sup.2] quadrats centered around the rare focal species Stenosiphon linifolius. (b) Based on frequency data collected in 0.01- and 0.1-[m.sup.2] quadrats centered around the rare focal species Penstemon cobaea var. purpureus and Centaurium texense. (c) Tweny-three of 32 sample sites were located in the Interior Low Plateaus and nine in the Ozark Plateaus; however, data were not stratified by site. Table V Summary of the geographic distribution of 13 taxa endemic, or nearly so, to xeric limestone prairies of the eastern United States Kev: OZMO = Ozark Plateaus of Missouri; OZAR = Ozark Plateaus of Arkansas; RVWV = Ridge and Valley of West Virginia; RVVA = Ridge and Valley of Virginia; RVAL = Ridge and Valley of Alabama Endemic taxa OZMO OZAR RVWV RVVA RVAL Castilleja kraliana x Coreopsis grandiflora var. inclinata x Dalea cahaba x Delphinium treleasei x x Echinacea paradoxa var. paradoxa x x Erigeron strigosus var. dolomiticola x Liatris oligocephala x Monarda fistulosa subsp. brevis x x Onosmodium decipiens x Scutellaria bushii x x Silphium glutinosum x Spigelia gentianoides x tialerianella ozarkana x x Total 5 5 1 1 8 Table VI Distribution of limestone cedar glade endemic/near-endemic taxa in xeric limestone prairies of eastern United States Limestone Cedar Glade XLP endemic taxa distribution Reference Astragalus tennesseensis * Tennessee DeSelm, 1991 Dalea foliosa * Tennessee DeSelm, 1991 Dalea gattingeri Alabama DeSelm, 1993 Georgia DeSelm, 1993 Tennessee DeSelm, 1993 Echinacea tennesseensis Tennessee DeSelm, 1991 Leavenworthia alabamica Alabama Webb et al., 1997 Leavenworthia exigua var. Georgia DeSelm, 1993 exigua Tennessee DeSelm & Chester, 1993 Leavenworthia exigea var. Kentucky Lawless et al., 2004 laciniata Leavenworthia exigua var. Alabama Allison & Stevens, 2001 lutea Leavenworthia stylosa Tennessee Baskin & Baskin, 1977 Lobelia appendiculata var. Tennessee Baskin & Baskin, 1977; gattingeri DeSelm, 1991 Onosmodium molle Tennessee DeSelm, 1991 Pediomelum subacaule Alabama Allison & Stevens, 2001 Georgia DeSelm, 1993 Tennessee DeSelm, 1991; 1993 Talinum calcaricum Kentucky Lawless et al., 2004 Trifolium calcaricum Virginia Collins & Wieboldt, 1992; Ludwig, 1999 * Near endemic.
|Printer friendly Cite/link Email Feedback|
|Author:||Lawless, Patrick J.; Baskin, Jerry M.; Baskin, Carol C.|
|Publication:||The Botanical Review|
|Date:||Jul 1, 2006|
|Previous Article:||Research in the garden: averting the collections crisis.|
|Next Article:||Morphology and anatomy of palm seedlings.|