We report a detailed study of the pollen wall structure in Linum burkartii and L. usitatissimum (Linaceae). This study was carried out using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Results confirm that the exine of the species studied is formed by: the ectexine constituted by clavae, gemmae, and a loose basal layer, a thin endexine and a fibrilar intine. These observations constitute the first report on the ultrastructure of pollen in the Linaceae that characterizes the species here studied and expands the palynological knowledge of this family.

Lepraria normandinoides is currently circumscribed to include three chemotypes that differ in the presence/absence of two structurally similar depsidones: fumarprotocetraric acid and protocetraric acid. Populations of these chemotypes are sympatric and morphologically identical. A molecular phylogeny was inferred from ITS1, 5.8S, and ITS2 sequence data generated from a geographically broad sampling of populations of L. normandinoides. Two strongly-supported clades were recovered that correlate to the depsidone producing chemotypes. Consequently L. normandinoides is restricted to the protocetraric acid producing chemotype and the fumarprotocetraric acid chemotype is formally described as L. oxybapha. Acid deficient populations are tentatively retained in L. normandinoides.

Female flowers of gynodioecious species (plants with either female or hermaphroditic flowers) are typically smaller with no reward for pollinators. We proposed and tested that pollinators (bees and butterflies) would visit female plants less frequently and visit fewer flowers for a shorter duration in a native prairie species, Lobelia spicata. We observed and quantified floral visitors exhibiting pollinator behavior in a population of L. spicata in a remnant tallgrass prairie located in central Illinois. We also quantified the number of seeds per fruit and fruits per plant. We found female plants have fewer visitors than hermaphroditic plants. Bees and butterflies, once on female plants, visited fewer flowers and spent less time than on hermaphroditic plants. Despite the lower number and shorter duration of visits, the females set more seeds per fruit. However, females did not consistently produce more seeds per plant, which is required to maintain females in the population.

Members of the aquatic plant genus Utricularia have extremely modified vegetative morphology, which includes absence of roots and lack of a clear distinction between leaves and stems. In this study, the vegetative morphology, anatomy and growth pattern of Utricularia gibba (L.), a common species in the southern Florida Everglades, was analyzed to determine general developmental patterns and establish a basic model for growth of this species. Specimens were collected from different habitats in southern Florida during both the wet and dry seasons. Light and scanning electron microscopy were used to quantify anatomical and morphological characteristics. Utricularia gibba has a shoot apical meristem that produces repeating developmental units, each of which consists of a stolon bearing dichotomously branched leaf-like structure that supports the traps. These structures may subtend one or more axillary buds that can grow out to form additional stolons or inflorescences. The inflorescence produces aerial flowers, but also forms a modified stolon at its base that is covered with unique glandular papillae. Re-iteration of this model through branching creates the floating and benthic mats found in this species. The basic model can be expanded to other species of section Utricularia but not to all species of the genus.

Rhododendron maximum L. is an evergreen, clonal shrub that forms a dominant sub-canopy layer and is a key species in southern Appalachian forests. We investigated the age and distribution of R. maximum across the Coweeta Basin, a 1626 ha watershed in western North Carolina. We selected 16 perennial, second-order streams and used a Global Positioning System to establish site boundaries and map the coverage of R. maximum across the hillslopes from stream to ridge. In each site, three transects from stream edge to the ridge were used to measure diameters of overstory trees (≥ 2.5 cm dbh), tree saplings (< 2.5 cm dbh) and shrubs including R. maximum stems. Along each transect, we cut cross-sections of R. maximum ramets and extracted increment cores from nearest neighbor trees to determine ages. The 16 sites ranged in size from 0.3 to 1.9 ha depending on the distance from stream to ridge. Rhododendron maximum cover ranged from 25 to 100% and ages ranged from 6 to 120 years. Rhododendron maximum establishment year showed a skewed unimodal distribution with the peak establishment occurring between 1928 and 1940. Although the R. maximum age and distance-from-stream relationship was statistically significant, the relationship was not meaningful as distance-from-stream only explained 2.6% of the variation in R. maximum age (r²  =  0.026, P  =  0.0003, n  =  487). Distance from stream only explained 4.2% of the variation in overstory tree age (r²  =  0.042, P  =  0.0015, n = 237). It appears that R. maximum has not expanded upslope over the last 100 years; rather the ranges in sizes and ages suggest that ramets are recruiting under established R. maximum canopies particularly in the wetter, near stream locations.

Odocoileus virginianus Zimmerman (white-tailed deer) populations at Valley Forge National Historical Park in southeastern Pennsylvania have ranged from 70 to 93 deer per square km over the last decade. In 2010, we surveyed 30 fenced (to exclude deer) and unfenced paired plots established in 1992 to evaluate differences in woody and herbaceous vegetation. In addition, a broader collection and analysis of overstory and understory vegetation data were conducted in the two dry Quercus spp. forests and one mesic Liriodendron tulipifera L. forest that contain the fenced and unfenced plots. The understory of the Liriodendron stand had 72% cover of Microstegium vimineum, (Trin.) A. Camus, an exotic grass. Across all three stands, sapling and seedling densities were low to moderate (500 and 10,000 stems ha⁻¹, respectively) and were primarily shade tolerant Acer rubrum L. and Nyssa sylvatica Marsh. The numbers of Quercus seedlings and saplings were low to nonexistent in all stands. Similarly, the unfenced plots had low densities of seedlings and saplings, especially Quercus. The 18 year-old fenced plots contained 33,133 seedlings ha⁻¹, but lacked tree saplings. The fenced seedlings were dominated by A. rubrum, Sassafras albidum, (Nutt.) Nees, N. sylvatica and Fraxinus americana L. in two of the stands and Quercus montana L. in one of the Quercus stands. Shrub cover averaged 51% in the fenced plots (mainly Lindera, Viburnum, Ligustrum and Lonicera) and 5% in the unfenced plots. Moreover, Microstegium cover was very low in all fenced plots, possibly due to the high shrub cover. Fenced plots also had higher plant species diversity (Shannon Index and richness). The results of this study suggest that the long-term exclusion of O. virginianus resulted in increased tree seedling number, shrub cover and diversity, and low Microstegium cover, but was inconsistent in stimulating Quercus regeneration and sapling recruitment.

Analyzing the spatial pattern of individuals within a plant community is useful for inferring the underlying mechanisms and processes governing community organization. Aggregated patterns can be caused by habitat-specificity or dispersal limitation, while overdispersed patterns are often caused by competition or predatory host-specificity. However, few studies have looked at how abiotic influences might drive spatial patterning. Here, we address how topography might lead to differences in tree spatial patterning on a north- and south-facing slope at Dysart Woods. We also look at how species identity, canopy position, and sampling year affect spatial patterning. We found that the spatial distribution of trees was highly aggregated on a south-facing plot, but did not differ from random on a north-facing plot. We attribute these differences largely to the differential dominance of Acer saccharum and Fagus grandifolia on the north and south plots, respectively. Acer saccharum individuals were largely randomly distributed, while Fagus grandifolia individuals were highly aggregated. We also found differences between understory and overstory individuals, with smaller individuals being more aggregated, likely due to life-history or competitive self-thinning processes. Lastly, we observed a surprising change in the scale that aggregation occurred at from 1996 to 2011 on the south plot. In 1996, stems on the south plot were aggregated at distances >3 meters, but only at distances >10 meters in 2011. We found that this reduction in aggregation was largely due to the unusually high mortality rates of Cornus florida and Prunus serotina. We conclude that topography can greatly influence spatial patterning, and that this and other abiotic variables should be taken into consideration in future studies on community structure.

Diversity and compositional dynamics in deciduous forests of eastern North America are subject to an array of drivers, and these forests are thought to be undergoing a long-term dominance shift in which maples (Acer spp.) are replacing oaks (Quercus spp.). We examined dynamics of woody understory species in an old-growth central Appalachian forest to document baseline dynamics over a decade and describe the relationships between diversity and surrogate measures of productivity (density and cover). We also investigated the reaction of shrub- and ground-layer vegetation to a recent fire in one portion of the study site. Over a ten-year interval (2000–2010), shrub-layer densities of Acer spp. decreased significantly (P < 0.01) and Quercus spp. did not change. Ground-layer densities of Acer spp. increased significantly (P < 0.01) while mean cover decreased significantly (P < 0.01). For a few Quercus species, ground-layer densities increased significantly (P < 0.01), and mean cover of ground-layer Quercus alba increased significantly (P < 0.05). Significant unimodal relationships (P < 0.01) were exhibited between diversity/richness and density/cover in both understory strata. Plots burned by a wildfire in 2010 had significantly lower shrub-layer density and species richness (P < 0.01) compared to pre-disturbance data in 2000. In the same area, the ground-layer response to this fire disturbance was equivocal, with increased density, decreased cover, and unchanged diversity and richness. Decadal oak-maple dynamics in the woody understory of this old-growth stand did not fully support the hypothesized oak-to-maple dominance shift; changes in shrub- and ground-layer populations of Acer spp. were erratic while Quercus spp. populations appeared stable. Unimodal relationships between diversity and productivity surrogates were found in both woody understory layers and are likely driven by the range of environmental conditions found within the study site. Fire induced short-term changes in the shrub- and ground-layers and may have facilitated colonization of invasive species. Quantification of changes across all woody forest strata in areas of minimal anthropogenic disturbance (e.g., old-growth stands) is important for understanding baseline dynamics of deciduous forests in eastern North America.

An important concern of natural areas management is preserving rare taxa in the face of increasingly pervasive plant invasions. This is one study in an ongoing long-term botanical analysis of Big Everidge Hollow, a watershed containing old-growth forest within the Lilley Cornett Woods Appalachian Research Station on the Cumberland Plateau of eastern Kentucky. We report 23 new records from the site representing an array of life histories. Many of these species are scarce and difficult to identify; however, we also detected several non-native invasive species moving into the study site, which had previously been completely free of these species. The appearance of exotic species may have been facilitated by a recent wildfire that occurred in one portion of the watershed. This site remains one of the best examples of old-growth mixed mesophytic forest, and in our view, this nascent invasion poses a significant threat to the site's biological integrity.

A survey of Ozark chinquapin in northwest Arkansas showed the tree to be widespread but uncommon, locally concentrated on sandstone benches and upper slopes in karst terrain. The remains of pre-blight chinquapin trees were mapped and measured on two 200 ha sites in Benton and Washington counties. Tree density was estimated at about 1 tree per ha, consistent with published upland witness tree surveys. 108 relatively intact logs were documented, representing about half of the total large chinquapins estimated to have grown on the two sites. Most of these grew as single-stemmed trees up to 20 m in height and 40 cm in diameter. The predominance of single-stemmed trees may indicate the relative absence of fire in pre-blight forests. The lack of subsequent sprouts from the base of large trees and the abundance of living sprout clones not associated with old trees indicates most sprouts are old seedlings. Tree rings were used to reconstruct stand history on six 20 × 60 m plots. Release of adjacent trees showed that blight arrived in 1957. Stand recruitment was concentrated in an extended period from 1920 to 1940, with a few older trees dating to the mid 1800's. Estimated ages for chinquapin logs seemed to pre-date oak and hickory recruitment, suggesting release of suppressed understory stems. Tree rings were used to determine the diameter of trees on the plots in 1955. Height-diameter projections were used to compare the stature of the stand with reconstructed remains of chinquapin logs, demonstrating that Ozark chinquapin was growing as a canopy-dominant tree at the time when blight arrived.

Salt marsh islands are currently threatened by rates of erosion and sea level rise outpacing rates of sediment accretion. Past deposition of dredge materials was used to create spoil islands, which adds to total upland area but may result in different plant assemblages as found on natural islands. Plant diversity was compared between 11 spoil and 9 natural islands of the Timucuan Ecological and Historic Preserve, along a tributary of the St. Johns River, northeastern Florida. At each island, a 1 m² quadrat was used to estimate plant and unvegetated cover every 5 m along a transect that extended from the upland border to the approximate middle of each island. Three parallel transects were surveyed on each island. A total of 92 species was recorded with a mean of 21.8 species per spoil island and 16.7 species per natural island. Mean species richness was 3.0 ± 0.1 m⁻² and each species was present in < 3% of the sampled quadrats. Island area contributed to species richness in the preserve, describing 78.1% of the variability in species richness. Species diversity on spoil and natural islands was similar (ANOSIM; Clarke's r  =  0.069), indicating that the approximately 50-yr old spoil islands had sufficient time for physical conditions to be comparable to natural islands. The low species occurrences among islands were likely a response to local dynamics influencing species recruitment and survivorship. This study suggests that spoil material may serve to maintain upland habitat as rising sea level threatens coastal marshes.