In the southern Appalachians, Rhododendron maximum L. (Ericaceae) is a key evergreen understory species, often forming a subcanopy in forest stands. Little is known about the significance of R. maximum cover in relation to other forest structural variables. Only recently have studies used Global Positioning System (GPS) technology as a field-based method to map the perimeter of shrub patches as a means of estimating canopy cover. We assessed the viability of using GPS technology to accurately measure R. maximum canopy cover in mountainous terrain; and we compared canopy cover to other R. maximum abundance variables, forest structural attributes, and environmental factors. We selected forty 20 × 40 m permanent plots at Coweeta Hydrologic Laboratory in western North Carolina to employ a variety of methods (visual estimates, GPS, and x-y coordinate measurements) to estimate canopy cover of R. maximum within each plot. We found a positive relationship between the GPS method and the more accurate x-y coordinate measurements (r = 0.967, p < 0.001). We compared the GPS-derived estimates to other measures of R. maximum abundance and found positive relationships between cover and density (r² = 0.800, p < 0.001), basal area (r² = 0.747, p < 0.001), total biomass (r² = 0.761, p < 0.001), and leaf area index (r² = 0.761, p < 0.001). The GPS method is a reliable field-based technology to estimate evergreen canopy cover and it could be used to estimate more difficult to measure parameters of R. maximum, given the significant relationships found in this study.

Glossy buckthorn (Rhamnus frangula L. or Frangula alnus Mill.) is a tall shrub to small tree, native to Eurasia and North Africa, which can invade both wetland and upland sites in northeastern and midwestern North America. Invasion by R. frangula is often patchy in riparian savannas of the Allegheny High Plateau of northwestern Pennsylvania: Some sites are heavily invaded, whereas others nearby are sparsely invaded or not invaded at all.

Because microtopography in riparian savannas can be diverse, with slightly elevated dry hummocks and wet depressions providing habitats for a range of upland and wetland plant species, we tested the hypothesis that the patchy nature of R. frangula invasion of riparian savannas is determined in part by soil drainage, which may ultimately affect shrub establishment and growth. We used depth to mottling by iron oxides as a primary, indirect measure of soil drainage and inundation and also measured the depth of leaf litter, humus, and A and B horizons in belt transects that were either heavily or sparsely to noninvaded by R. frangula at three riparian savanna study sites in the Allegheny National Forest. Sites with relatively well-drained soils, indicated by greater mottling and B horizon depths, supported significantly greater densities of R. frangula seedlings and shrubs than did sites with apparently wetter soils having shallower mottling and B horizon depths. We suggest that physiological stress associated with wet soils may restrict establishment of R. frangula, explaining in part its patchy invasion patterns in Allegheny High Plateau riparian savannas.

The high marsh in southeast Atlantic coast salt marshes forms a relatively small but ecologically important transition zone between low marsh and the terrestrial shoreline. However, long-term trend data from high marshes are limited to a few studies. Permanent plots established in a high marsh near Waties Island in northeast South Carolina were measured for plant coverage from 2002–2010. At the beginning and at the end of the study, four groups of plots were identified: mixed indicated by Borrichia frutescens, Distichlis spicata, and Fimbristylis castanea; Juncus indicated by Juncus roemerianus; Salicornia indicated by Salicornia virginica; and Spartina indicated by Spartina patens. Ordination of the 2010 plot data and soil analyses produced clear separation of the groups along a single dominant axis with Salicornia and Juncus groups at the high end of a salinity gradient and the Spartina group at the high end of soil organic matter gradient. Comparison of plots classified in 2002 to the same plots in 2010 suggested both stability and change, depending on community classification. Salicornia and Spartina groups were stable. The mixed group experienced a switch in dominance from Distichlis spicata to Borrichia frutescens, whereas the Juncus group had gradually declining importance of Juncus roemerianus and gradually increasing importance of Borrichia frutescens. These data represent patterns and trends in a system and time period not affected by development or influenced by high intensity disturbance, and can be used as a reference for other high marshes in the immediate area experiencing environmental change.

Within a successional context, the vegetation associated with the cedar glade ecosystem in middle Tennessee develops from bare limestone bedrock to subclimax redcedar, preclimax oak-hickory, and climax mixed hardwood forests. Studies on the composition and structure of forests associated with cedar glade–forming limestones (Lebanon, Ridley) are rare. We sampled the canopy and understory of six forest stands in middle Tennessee on these limestones. Observed number of canopy species was 14–24 across stands; estimated richness was greater by 1–3 species (bootstrap) or 3–6 species (first-order jackknife) than observed richness. With the exception of Ailanthus altissima in one stand, all other canopy species were native. Juniperus virginiana, Fraxinus americana, Carya ovata, and Quercus muehlenbergii were primary canopy components in 4 or 6 stands, and C. glabra, Q. shumardii, Ulmus alata, F. quadrangulata, Q. alba, and Q. velutina in 2–3 stands. When we included stands from a previously published study (most on the non-glade Carters Limestone) with our data, a principal components analysis identified three groups with the axes approximating a moisture-bedrock gradient and a time-successional gradient. An examination of regeneration in our stands predicts that (1) mesophytes and/or fire-sensitive species (Acer saccharum, Fraxinus spp., Celtis spp.) will increase and (2) xerophytes and/or fire-adapted species (Quercus spp., Carya spp.) will decrease. Altogether, our results strongly suggest that the oak-hickory stage shown in successional outlines of vegetation development associated with the cedar glade ecosystem may not occur in its current state in the future.

Forest fragmentation continues to increase throughout the United States and many studies show a trend toward taxonomic homogenization in fragments. Characterizing the species composition and diversity of remaining forest fragments and identifying factors influencing composition and diversity can help to inform management practices and aid in setting conservation priorities. Because influencing factors tend to be site-specific, studies of local and regional fragment community composition are needed. We studied forest community composition and diversity in four forest fragments of the Shenandoah Valley, Virginia. We found significant (p < 0.05) differences in tree species diversity (H') as well as abundance of exotic species across forest fragments. In addition, results of nonmetric multidimensional scaling (NMS) showed that there were significant differences in species composition among fragments. Soil properties and stand age, in addition to presence of exotic species, were the most important factors influencing differences in tree species composition. Older fragments on high quality soils, with low abundances of exotic species and low levels of disturbance, were more diverse and had greater abundances of economically and ecologically valuable overstory species such as Quercus spp., Carya spp., Nyssa sylvatica and Liriodendron tulipifera. Younger fragments on clay/sandy soils with higher disturbance levels had higher abundances of exotic invasive species, such as Ailanthus altissima and Microstegium vimineum, and lower abundances of late-successional native hardwood tree species. Based on seedling and sapling species composition, we predict that these forest fragments are likely to change in species composition with mortality of overstory trees.

The primary objectives of this project were to determine which species of Crotalaria (Fabaceae) occur in Alabama and the county distribution of each species. Crotalaria, known commonly as rattlebox, is recognized as consisting of seven species in Alabama. The most common species are Crotalaria sagittalis, C. rotundifolia, and C. spectabilis. The less common species are C. purshii, C. lanceolata, C. pallida, and C. ochroleuca. In Alabama, introduced species of Crotalaria (C. lanceolata, C. ochroleuca, C. pallida, and C. spectabilis) generally have showier inflorescences and reach greater maximum heights than native species (C. purshii, C. rotundifolia, and C. sagittalis). The dichotomous key and descriptions we present are modifications from earlier authors; however, all measurements are based on morphological features of the vegetative and reproductive structures of the more than 460 specimens studied during this project. Data for the county-level distribution maps were compiled entirely from herbarium vouchers.