Kentucky Bluegrass Savanna

This system represents deep soil savannas and woodlands of the Inner Bluegrass Basin of Kentucky (EPA Ecoregion 71l and "S. Fork Licking River arm" of EPA Ecoregion 71d). Only remnants or stands undergoing restoration are currently extant. The original woodland-savanna aspect, especially on drier uplands, is believed to have been dominated by fire-resistant oaks, especially Quercus muehlenbergii and Quercus macrocarpa, but also with a variety of other species such as Fraxinus quadrangulata, Robinia pseudoacacia, Gleditsia triacanthos, Acer saccharum, Fraxinus americana, Fraxinus pennsylvanica, Carya cordiformis, Juglans nigra, and the rare Gymnocladus dioicus. The understory is composed of cool-season grasses, as far as known (e.g., Elymus, Dichanthelium) with Arundinaria gigantea (extensive canebrakes). Settlers referred to a "buffalo grass" of unknown identity (possibly Dichanthelium clandestinum or Dichanthelium scoparium). The fire regime is unknown. Characteristic remnant trees (e.g., Fraxinus quadrangulata, Quercus macrocarpa) are fire-tolerant.

The original woodland-savanna aspect, especially on drier uplands, is believed to have been dominated by fire-resistant oaks, especially Quercus muehlenbergii and Quercus macrocarpa, but also with a variety of other species such as Fraxinus quadrangulata, Robinia pseudoacacia, Gleditsia triacanthos, Acer saccharum, Fraxinus americana, Fraxinus pennsylvanica, Carya cordiformis, Juglans nigra, and the rare Gymnocladus dioicus. The understory is composed of cool-season grasses, as far as known (e.g., Elymus, Dichanthelium) with Arundinaria gigantea (extensive canebrakes). Settlers referred to a "buffalo grass" of unknown identity (possibly Dichanthelium clandestinum or Dichanthelium scoparium). Historical descriptions also mention "pea vine," two or three species of nettles, Vernonia species, Ageratina altissima, and Trifolium stoloniferum.

These savannas or woodlands occur on deep fertile soils of the Inner Bluegrass Basin of Kentucky (Ecoregion 71l and "S. Fork Licking River arm" of Ecoregion 71d of EPA (EPA 2004) and Woods et al. (2002)).

Central Kentucky grasslands were maintained by a combination of grazing, periodic drought, and fire, but one would expect that woody succession was also retarded by the heavy, clayey soils originating from the limestone substrate (Landfire 2007a). The first approximation map of presettlement fire regimes of the U.S. indicated fire regimes of 4-6 and 7-12 years in the model area (Frost 1998). Anthropogenic burning by Native Americans was probably an important part of the presettlement fire regime. This would have resulted in an expansion of these savannas and woodlands into otherwise forested areas. Native American population decline after the 1500s may have led to a decrease in the amount and frequency of burning in these savannas. Characteristic remnant trees (e.g., Fraxinus quadrangulata, Quercus macrocarpa) are fire-tolerant. Grazing by native herbivores (white-tailed deer and bison) may have been an important factor in maintaining the open character of these savannas.

In the gently rolling limestone regions, large expanses of land without significant firebreaks lie between the major firebreak streams. The large size of fire compartments in these areas suggests that fire frequency should have been high, perhaps 4-6 years where understory species were conducive to fire spread. Areas dominated by Elymus species may have experienced lower fire frequency because of the reduced capacity of this fuel type to carry fire (J. Campbell pers. comm.). Some writers think that these cool-season grasses (e.g., Elymus) predominated in preference to warm-season grasses such as Schizachyrium. In contrast, a fire frequency of 7-12 years could be expected in areas with broken topography such as the more rugged parts of the Outer Bluegrass and other limestone margin regions (Landfire 2007a). Lightning and Native Americans likely provided roughly equal influence as ignition sources in presettlement Kentucky, with Indian influence being the dominant factor locally near population concentrations and around fall and winter hunting camps. U.S. Weather Service lightning ground flash monitoring stations indicate a lightning strike density of 4-8 strikes per square kilometer per year in the limestone regions. While only a tiny fraction of strikes result in ignitions, this rate would have produced a fire regime sufficient to support canebrakes and woodlands even in the absence of man (C. Frost pers. comm.).

The most critical anthropogenic threat to native grasslands, savannas and barrens is their conversion to human-created land uses, including residential development, quarries, industrial development, infrastructure development, and others (TNC 1996c). Rocky glade areas, if present, may be the last areas to be converted to development and housing due to the unsuitability of the soil to septic tanks. Other common threats and stressors include both the removal of disturbance (e.g., fire, grazing) and the effects of inappropriate or too intensive or constant disturbance. These areas often attract off-road-vehicle use.

Fire plays a critical role in the maintenance of most native grasslands. Without it, Juniperus species, Quercus species and other hardwoods quickly regenerate, shading out the herbaceous plants, and leading to a shift in species diversity from the ground layer to the upper woody strata, resulting in a net loss of species diversity (Taft et al. 1995). At sites with intermediate levels of woody encroachment, a signal of restoration potential is an inverse relationship between woody stem density and ground layer species richness (Taft 2009). In landscapes where open grassland or savanna vegetation is part of the matrix, and where woody plants have taken over areas once occupied by open grassland and savanna vegetation, the light-dependent species may only persist on the open edges (roadsides, powerlines) of forested patches (Taft 1997). In southeastern grasslands, complete transition to forest dominated vegetation can occur in one or two decades (Wiens and Dyer 1975). More information is needed about the particular appropriate ranges of fire-return times and intensities in the various systems, along with factors other than fire (e.g., soil/substrate, aspect, herbivory, hydroperiod and flooding) that help maintain grasslands and related communities. Occasional surface fire will retard woody plant encroachment and help maintain herbaceous diversity, as will, to an extent, grazing or mowing. Too intensive or frequent application of these disturbances will have deleterious effects on stand structure and species diversity. In general, mosaics of scrub and grassland, produced by light to moderate grazing (or occasional fire) will support the greatest diversity (Duffey et al. 1974). Cutting or mowing is not as favorable to plant diversity as is grazing because it is nonselective and does not result in the same kind of soil disturbance and compaction as do the hooves of grazing animals (DeSelm and Murdock 1993). Fire is a critical disturbance factor for southeastern native grasslands, but the intensity, duration, and timing of the fires are all important in their effect on the vegetation (DeSelm and Murdock 1993). In addition to occasional fire, periodic drought may also be important in regulating woody plant encroachment in native grasslands. It is believed that native grasslands have evolved under a combined system of grazing, drought, and periodic fire (Duffey et al. 1974, Estes et al. 1979, Noss 2013).

Fragmentation of native grasslands, barrens, and savannas occurs with the development of housing and industrial sites, as well as the construction of roads, which not only function as firebreaks, limiting the areas that can be burned with one ignition event, but which make it more difficult to mitigate the effects of smoke on human populations and their activities. A small isolated patch has a low probability of receiving a lightning strike frequently enough to maintain a grassland condition. In many cases, grassland systems were once extensive on the landscape, but have now been reduced to scattered and isolated remnant patches, presenting conservation and management challenges. These disturbances have had damaging effects on fragile soil profiles and plant and animal species. These combined impacts also foster a trend toward biotic homogenization, which results in the gradual replacement of ecologically distinct natural communities by those dominated by weedy generalists (McKinney and Lockwood 1999). In other cases, the grassland or glade system naturally occurs in small isolated patches occurring within an otherwise forested matrix.

Many native grassland sites, particularly the more productive ones, have been converted to plantations of exotic grasses and legumes (DeSelm and Murdock 1993). Even if not completely converted, the extirpation of native species and the concomitant spread of invasive exotic plants (particularly Ligustrum species and Lonicera species shrubs, as well as Ailanthus altissima, Albizia julibrissin, Alliaria petiolata, Lespedeza cuneata, Microstegium vimineum, and Miscanthus sinensis) will fundamentally alter the character of native grasslands, barrens, savannas, and glades. Some of these exotics are allelopathic, thereby presenting a greater threat to native species (N. Murdock pers. comm.). Opportunistic native increaser plant species (e.g., Juniperus virginiana) can also shade out light-requiring herbaceous plants (TNC 1996c).

The most significant potential climate change effects over the next 50 years include shifts to dramatically drier or moister climate regimes. A cooler and wetter regime would most likely accelerate the trend toward woody plant encroachment, removing drought as a factor in its inhibition. A moderately drier regime during the growing season could favor the characteristic native grasses and forbs, which are adapted to these conditions better than the generalists. An extremely drier regime for an extended period of time could ultimately have negative effects.

This system is restricted to the Inner Bluegrass Basin of Kentucky (Ecoregion 71l and "S. Fork Licking River arm" of Ecoregion 71d of EPA (2004) and Woods et al. (2002)). Only remnants or stands undergoing restoration are currently extant.