Herbaceous Flora of Blackland Prairie Remnants in Mississippi and Western Alabama

Castanea, Dec 2007 by Barone, John A, Hill, Jovonn G

ABSTRACT

Blackland prairies are found in two regions of Mississippi and Alabama, the Black Belt and the Jackson Prairie Belt. As the vegetation of remnant prairies in these belts has received limited attention, we collected and identified the grasses and forbs from 19 sites in Mississippi and western Alabama over a year. A total of 196 species were collected, including 168 native species. When compared with prairie species lists for these regions from the early 20th century, our results suggest that 18 species present then may now be rare or absent. Habitat data for these species indicate that most favor more mesic or sandier soils. Their apparent absence implies that existing prairies are a subset of historic prairies that were not useful for agriculture because of drier and/or more clayey soils. To preserve the biodiversity of this ecosystem, conservation efforts should include more mesic or sandy sites, including prairie-forest ecotones.

INTRODUCTION Grasslands were a scattered but ecologically distinctive part of the historic landscape in the Southeast (DeSelm and Murdock 1993). One type of grassland, blackland prairies, occurred in areas with calcareous bedrock (Peacock and Schauwecker 2003). In Mississippi and Alabama, blackland prairies were found in two regions, the Black Belt and the Jackson Prairie Belt (DeSelm and Murdock 1993, Peacock and Schauwecker 2003). These regions were not continuous prairie in the 19th century, but instead consisted of numerous prairie "islands" surrounded by forest (Rostlund 1957, Jones and Patton 1966, Barone 2005a, b). General Land Office surveys of the regions performed in the 1830's and 1840's indicate that about 145,000 ha of prairie were present in the Black Belt (Barone 2005a) and about 19,500 ha in the Jackson Prairie Belt (Barone 2005b). Due to agricultural development and urban growth, less than 1% of the historic area of prairie remains, making this an endangered ecosystem (Noss et al. 1995). Moreover, many of the remaining prairie remnants are deteriorating because of stock grazing, erosion and the increasing density of woody species, such as, Juniperus virginiana L., due to fire suppression.

The vegetation of these blackland prairies has not been well-studied (Foti 1989, DeSelm and Murdock 1993, Leidolf and McDaniel 1998). For the Black Belt, only three floristic inventories covering one or two sites have been published (Schuster and McDaniel 1973, Leidolf and McDaniel 1998, Forbes 1999). The flora of the Jackson Prairie Belt has received more attention, though these studies have focused on sites within the Bienville National Forest (Jones 1971, Gordon and Wiseman 1989, Wieland et al. 1991; summarized in Moran et al. 2003).

In the present study, we censused the grasses and forbs of 19 remnant prairies in the Black Belt and the Jackson Prairie Belt from sites in Mississippi and western Alabama. We had two main goals. The first was to assess the frequency of species across sites, in order to better charaderize the composition of these prairies. We focused on non-woody species since these are the priority in on-going prairie restoration efforts and are the defining vegetation of prairies. The second goal was to compare the current spedes composition of relid prairies with historical sources that list species found on these prairies in the early 20th century. While no comprehensive list exists for any site until Schuster and McDaniel (1973), partial lists of prairie species are provided by Lowe (1911, 1921) for Mississippi and Mohr (1901) for Alabama. By comparing the composition of extant prairies with the historic observations from these sources, we can better understand how the composition of native species may have changed on these prairies over the last century.

METHODS Five study sites were located in the Jackson Prairie Belt in Mississippi, and 14 were in the Black Belt in Mississippi and western Alabama (Figure 1; Table 1). Sites were chosen based on their accessibility and their relative lack of human disturbance. A detailed history of the sites is not known, but in general most appear to have remained prairie because of a benign neglect resulting, for example, from their inaccessibility to cattle. The sites have no obvious on-going human disturbance, except for roadside prairies that are occasionally mowed and the site in the Tombigbee National Forest, which is burned annually in the early spring (Table 1). Almost all the sites are within or near the boundaries of known historic prairies, as seen on General Land Office plat maps from the 1830's and 1840's (Figure 1).

The sites were visited monthly from June through October 2005 and again during April and May 2006. Areas of each site (Table 1) were measured using a Trimble Geo XT GPS receiver (Trimble, Inc., Sunnyvale, CA) except for the Harrell Hill site where the area provided by Moran et al. (2003) was used. During each visit we recorded and collected voucher specimens of the grasses and forbs in flower. Threatened species were not collected but photographed for documentation. For roadside prairies, collections were only made from areas more than two meters from the road or from a roadside ditch, if present. We also excluded species found growing exclusively beneath the canopy of Juniperus virginiana trees. Herbarium specimens are on deposit in the herbarium at Columbus State University (COLG), in Columbus, Georgia. We used Radford et al. (1968), Smith (1994) and Jones (2005) for identification. Nomenclature follows the Integrated Taxonomic Information System (retrieved June 15, 2007, from http://www.itis.gov).