Volume 66 – Issue 1/2 (Mar/Jun 2001)

Plant taxa endemic (22) or nearly endemic (22) to the Coastal Plain of North and South Carolina show a distributional pattern strongly associated with the Cape Fear Arch geological uplift. The great majority (41 of 44) are concentrated in a region extending along the coast from Carteret County, North Carolina to Georgetown County, South Carolina, and northwestward along ancient Cape Fear River terraces to the Sandhills region, where the endemics and near-endemics are concentrated from Harnett County, North Carolina to Darlington County, South Carolina. Among the 44 taxa, 41 are angiosperms with 33 dicots and eight monocots. Among the angiosperms, 40 are biennial or perennial, 32 are herbaceous, 30 are wetland obligates or wetland facultative, and 34 are associated with fire-maintained southern pine and evergreen shrub communities. Ultisols and spodosols are the soil orders with which the majority of the endemic/near endemic taxa are associated, but some of the rarest taxa are associated with entisols and alfisols. Biological, habitat, and distributional characteristics of the endemic/near-endemic taxa suggest the Cape Fear Arch region contained migration corridors, refugia, and isolated habitats favorable for speciation during the last glacial episode.

A database of the county distribution of rare vascular plant species endemic to the southeastern United States was compiled in order to review geographic trends. A total of 482 species representing 246 genera in 95 families were identified as rare southeastern endemics. The histogram of total county occurrences for these endemic species is significantly skewed to the right, indicating that those taxa that are endemic to the southeastern United States are typically rare within the region. Six centers of endemism were identified: Central Peninsular Florida, Florida Panhandle, Southern Appalachian Mountains, Tennessee Central Basin, Mid-Atlantic Coastal Plain, and West Gulf Coastal Plain. This identification of geographic trends will assist in the proactive conservation of areas harboring large numbers of endemic taxa, and these data provide the first step towards an integrative understanding of the evolutionary ecology of endemic floras.

Genetic variation is often low in narrowly endemic species, and may be further depleted by habitat loss and fragmentation. Few studies have tested predictions about the distribution of genetic variation among co-occurring endemic plants species. We describe genetic variation and its relationship to life history traits for seven narrowly endemic, federally endangered Florida scrub species: Dicerandra christmanii, D. frutescens, Eryngium cuneifolium, Hypericum cumulicola, Liatris ohlingerae, Nolina brittoniana, Warea carteri. These species have varying life histories, degrees of habitat specialization, and geographic distributions. Measures of genetic (allozyme) diversity (mean number of alleles/locus, percentage of loci polymorphic and expected heterozygosity) varied among species. However, genetic variation was generally lower than published means for plants and also generally lower for means for comparable groups (endemics, short-lived herbs, species with mixed mating systems, species with gravity dispersed seeds). The chief exception was L. ohlingerae, which had relatively high genetic variation. All three measures of genetic variation produced the same ranking among species: L. ohlingerae > D. christmanii > N. brittoniana > E. cuneifolium > D. frutescens > W. carteri > H. cumulicola. For six of these species, we compared genetic variation with rankings of eight life history factors. Genetic variation was highest in long-lived, demographically stable, outcrossing species with long pollinator dispersal distances. Attributes such as median population size, habitat specificity, geographic range, and estimated primary seed dispersal distances were not related to rankings for genetic variation. The studied species varied widely in genetic differentiation among populations (0.02 < F_st < 0.72). The most differentiated species, H. cumulicola, is pollinated by specialized bees that may move short distances, therefore limiting gene flow among isolated patches. These comparisons emphasize that co-occurring narrowly endemic species can have a diversity of genetic patterns and that many factors can influence the amount and distribution of genetic variation. Further loss of genetic variation due to habitat loss and fragmentation will impact the genetic variation of these species differently. A single conservation strategy for this suite of species is therefore unlikely to achieve genetic conservation goals.

Here we submit that mathematical tools used in population viability analysis can be used in conjunction with floristic and faunistic surveys to predict changes in biogeographic range. We illustrate our point by summarizing the results of a demographic study of Lobelia boykinii. In this study we used deter ministic and stochastic matrix models to estimate the growth rate and to predict the time to extinction for three populations growing in the Carolina bays. The stochastic model better discriminated among the fates of the three populations. It predicted extinction for two populations in the next 25 years but no extinction of the third population for at least 50 years. Probability of extinction is likely correlated with hydrologic regime and fire frequency of the bay in which a population is found. The stochastic model could be combined with information about the geographic distribution of L. boykinii habitats to predict short-term biogeographic change.

Spiraea virginiana Britton is a Federally listed rhizomatous shrub endemic to the southern Blue Ridge and Appalachian Plateau physiographic provinces. Populations of S. virginiana are found restricted to scoured sections of high gradient streams within the Ohio River drainage. Present evidence indicates the species is reproducing asexually, most probably through the deposition of rhizomes from upstream populations forming new downstream ramets. Phenotypic variation was examined through a morphometric evaluation of 25 leaf measurements and analyzed using Principal Components Analysis (PCA) and Discriminant Function Analysis. Identity and structure at the molecular level were examined with Randomly Amplified Polymorphic DNA (RAPDs) and band patterns were used to construct a cluster analysis. Combining cluster analysis and biogeographic data identified past gene flow. Patterns of variation found within S. virginiana indicate that there is some degree of relatedness along short reaches of a single river and that within a secondary drainage basin a downstream distribution of propagules from multiple tributaries results in a mix of phenotypes. Results place the S. virginiana ancestral population in the southern part of its range suggesting a southward migration followed by recolonization northward, concordant with the work of Delcourt and Delcourt (1981, 1984). Biogeographical patterns of variation within S. virginiana identify the Cumberland Plateau as a migratory route. In addition, evidence suggests that that the deeply dissected Cumberland Plateau is the probable site of a Pleistocene refugium.

Five terete-leaved Talinum species occur in the Southeast on rock outcroppings. All are summer-flowering perennials with succulent leaves and ephemeral, rose-colored flowers. Talinum parviflorum (2n = 24) has tiny flowers that open late in the day and self-pollinate. This species is distributed in the central United States from Arkansas westward, but an outlier population, formerly named T. appalachianum, occurs in north Alabama. Talinum mengesii populations are predominantly diploid (2n = 24). This species, mainly distributed in northern Alabama, western Georgia and southeastern Tennessee, occurs on both granite and sandstone. Two enclaves of T. mengesii populations are isolated from the main distribution of the species and intervening outcrops have only T. teretifolium. Talinum teretifolium (2n = 48), distributed from Georgia to Pennsylvania, is an allopolyploid derived from T. mengesii and T. parviflorum. This conclusion is based on morphological, reproductive and biochemical evidence, and the recreation of the species from parental species. Talinum calcaricum (2n = 48) is endemic to limestone outcrops in Tennessee and Alabama. Evidence presented suggests a derivation from T. calycinum, which is distributed from Arkansas westward and has both diploid and tetraploid populations that grow on shale, sandstone or limestone. These two species differ in floral characteristics, but are 100% interfertile.

Three taxonomically recognized, allopatric, and ecologically differentiated varieties comprise Phacelia dubia (L.) Trel. (Hydrophyllaceae). Although widespread in the southeastern United States, the distribution and taxonomic affinities of P. dubia have been overlooked in South Carolina. Populations of P. dubia occurring on Piedmont granite outcrops in seven South Carolina counties have been referred to as putative variety “imitator.” Artificial hybridizations with P. dubia var. dubia and var. georgiana McVaugh produced partially sterile F1 hybrids. Hybrids with var. interior Fern. were mostly fertile except for occasional unidirectional sterility. The fertility distributions of backcross progenies of both var. dubia-“imitator” and var. georgiana-“imitator” F1 hybrids to “imitator” parents were bimodal, and fertility was associated with one allozyme and two DNA markers, suggesting a major segregating factor controls hybrid fertility. In contrast, “imitator” is not morphologically distinct relative to vars. dubia and georgiana. Conflicting morphological, ecological and reproductive data highlight the problem of taxonomic recognition of a morphologically cryptic entity.

Studies comparing rare and widespread congeners are particularly valuable when examining hypotheses regarding the condition of rarity. Comparison of close relatives can minimize confounding effects of phylogenetic history and, optimally, eliminate ecological factors not directly relevant. Among various hypotheses regarding rarity, this paper focuses on genetic correlates and potential ecophysiological restrictions, using a comparison from the coneflower genus Echinacea as an example. Echinacea tennesseensis is an endangered species endemic to limestone “cedar glades” of Tennessee. Echinacea angustifolia is a widespread prairie species. Genetic comparison of these close relatives supports the hypothesis that rare species have lower levels of genetic variability within populations than do widespread species. However, an ecophysiological study measuring photosynthesis following different light and soil moisture preconditioning regimes provides no evidence that narrow tolerances for these factors are responsible for the endemic species’ narrow geographical range. These results for the Echinacea congeners are compared with other relevant studies.

Table mountain pine (Pinus pungens Lamb.) communities of the southern Appalachian Mountains have been maintained historically by lightning- and human-caused fires. Characteristic stands have a table mountain pine overstory, a chestnut oak (Quercus prinus L.), scarlet oak (Q. coccinea Muenchh.) and black gum (Nyssa sylvatica Marshall) understory, and a mountain laurel (Kalmia latifolia L.) shrub layer. Following more than sixty years of fire suppression, most stands have increased densities of oaks and mountain laurel as well as fire-intolerant species such as red maple (Acer rubrum L.) and white pine (P. strobus L.). Previous research suggests that restoration of these communities can only be accomplished with high intensity fires that open the forest canopy and expose mineral soil. Opportunities to conduct such bums, however, are limited under current prescribed burning guidelines. Two recent studies examined community response to prescribed burning. Fires of low and medium-low intensity gave rise to abundant regeneration but may not have killed enough of the overstory to prevent shading. High-intensity fires killed almost all overstory trees but may have destroyed some of the seed. Fires of medium-high intensity may have been most successful; they killed overstory trees and allowed abundant regeneration. Large numbers of these seedlings survived the first growing season as their roots penetrated duff to reach mineral soil. Hardwood rootstocks resprouted after all fire intensities and may out-compete pine seedlings for available resources. Fires of lower intensity than previously recommended may best provide conditions for table mountain pine regeneration but additional research is needed. Prescriptions calling for lower intensity fires may widen the burning window defined by current guidelines.

The tremendous diversity of the plants of southeastern North America has long been recognized in the botanical community. Boufford and Spongberg (1983) summarized the history of awareness of the phytogeographical relationships of eastern North America and eastern Asia, tracing the origin of the ideas back to Linnaeus. Core (1970) recounted the botanical explorations of pioneer botanists in the 1700’s and 1800’s such as John and William Bartram, Andre and F.A. Michaux, and C.S. Rafinesque, who opened the eyes of the world to the natural wonders of the Southeast. Boufford and Spongberg (1983) pointed out that Nuttall made a trip through the southeastern United States, retracing the paths of Bartram and Michaux, and subsequently discussed the regional affinities of plants in a seminal 1818 publication that laid the foundation for the work of Asa Gray.