Previous phylogenetic analyses of Lepidium included only a few accessions of L. montanum, L. flavum, and L. fremontii to represent western North American species. Two additional species endemic to southwest Idaho have posed both taxonomic and conservation questions regarding their species status. Lepidium papilliferum was originally described as a variety of L. montanum, is morphologically similar to L. montanum, and is found in small scattered populations in southwest Idaho. The plant is restricted to specific edaphic conditions known as slick spots where high clay content creates conditions amenable to L. papilliferum, but to few other species. Resolving whether the populations of L. papilliferum merit species status distinct from L. montanum is a vexing question and phylogenetic analyses can assist in resolving this issue. Like L. papilliferum, L. davisii has specific edaphic requirements and is found in playas, areas similar to slick spots, but larger and with deeper soils. Unlike L. papilliferum, L. davisii is morphologically distinct from L. montanum and has posed less of a taxonomic quandary. Previous phylogenetic studies have shown that American species of Lepidium are derived from an ancestral allopolyploid species. In this study we have expanded previous analyses to include L. papilliferum, L. davisii, and several accessions of L. montanum along with published sequences of ITS, cpDNA, and PISTILLATA first intron. The western North American species form a monophyletic group with L. davisii sister to the remainder of the clade. Within this clade, L. papilliferum and L. fremontii are each monophyletic and sister to each other, but are imbedded within a paraphyletic L. montanum.

A review of live, herbarium, and bibliographic material of Chelyorchis revealed considerable floral variation that appears to be geographically correlated. We assessed whether the genus included one or more taxonomic entities using a morphometric analysis that involved the study of principal component analysis across several floral characters in 70 specimens, representing both geographical range and morphological variation. To further establish the phylogenetic relationships of Chelyorchis within the Cuitlauzina-clade, and to analyze the evolution of relevant characters, we performed a combined analysis using nrITS, morphological, and gap-coding data, which included only 13 species. Both of our phylogenetic analyses strongly support the genus Chelyorchis as a monophyletic entity that is sister to a clade comprised of the genera Ticoglossum and Rossioglossum. The morphometric evidence strongly suggests that two taxa at the level of species should be recognized, differing primarily in the shape and proportions of the labellum but also in their geographical distribution. A new species, Chelyorchis pardoi, is described and illustrated to accommodate the South American morph. Full description of the new species and iconography of both taxa are provided, and diagnostic features are discussed. The conservation status of the new taxon is assessed against IUCN criteria.

Two new species of Croton L. (Euphorbiaceae) from campos rupestres of Minas Gerais State, Brazil. Two new species of Croton L. (Euphorbiaceae) are described and illustrated. Croton lenheirensis and C. pradensis occur in campo rupestre vegetation on quartzitic soils in the State of Minas Gerais, the former from Serra do Lenheiro in São João del Rei and the later from Serra de São José in Tiradentes. Both localities are within an important center of endemism for Croton sect. Medea (Klotzsch) Baill., which is represented by at least 30 species from South America. Taxonomic characters which circumscribe the Section are subshrub and shrub habit, laciniate or glandular stipules, stellate indumentum, distribution of glands at the margins of either laciniate or entire sepals of the pistillate flowers, and multifid styles. Croton lenheirensis morphologically resembles C. gnidiaceus Baill., but differs by triangular stipules, lack of petals in pistillate flowers and of apressed-stellate trichomes. Croton pradensis shares morphological characters with C. vestitus Spreng. and C. josephinus Müll. Arg., species also represented on Serra de São José flora, but it is easily distinguishable from both by its wooly sparse indumentum and pistillate flowers which present entire eglandular sepals and no petals.

Two species of Aletris occur in southern Florida: the yellow-flowered A. lutea and white-flowered A. bracteata. It is uncertain whether A. bracteata found in southern Florida is a distinct species or a variant of one of the other species found in the southeastern United States. Plants with intermediate flower colors are present in southern Florida populations, indicating environmental and/or genetic variation of A. bracteata or hybridization between A. bracteata and A. lutea. The morphology, phenology and habitat of flower color morphs of Aletris were examined in four populations of A. lutea and six populations of A. bracteata in southern Florida, as well as five populations of A. bracteata on Andros Island, Bahamas, the type locality for A. bracteata. Plants with intermediate flower color dominated five of the six A. bracteata populations in southern Florida. These plants were morphologically most similar to A. bracteata from the Bahamas. Although A. bracteata and A. lutea differed in flowering phenology in southern Florida, flowering times in some populations overlapped in March, allowing for hybridization. Florida A. bracteata plants were self-compatible and autogamous, but seed set for open-pollinated plants was higher than for self-pollinated or autogamously-pollinated plants. We hypothesize that A. bracteata and A. lutea are distinct species but that some hybridization between them has occurred in southern Florida.

This work reports on the fruit surface and anatomy of seven Neotropical species of Indigofera (I. campestris Bong. ex Benth., I. hirsuta L., I. lespedeziodes Kunth, I. microcarpa Desv., I. spicata Forssk., I. suffruticosa Mill., and I. truxillensis Kunth) to help species diagnosis and clarify the fruit type classification. Flowers and fruits at several stages of development were removed from living material, fixed, and examined with scanning electron (surface analyses) and light microscopies (histological analyses). Species showed differences in relation to the number of exocarp layers, secretory trichome morphology and distribution, presence of stomata, phenolic idioblast size and distribution in mesocarp, the number and arrangement of endocarp fibers, and the presence of a separation tissue. It is noteworthy that no separation tissue was observed in I. microcarpa and I. suffruticosa, although they have dehiscent fruits, which indicates a delayed dehiscence. The present work confirms that fruit anatomical characters can be utilized as a tool for fruit type classification, especially in Indigofera, the third largest genus of Leguminosae.

The non-native invasive deciduous shrub Lonicera maackii causes a reduction in plant growth and species diversity under its canopy. The mechanisms of these effects are not fully understood, but an apparent difference between L. maackii and native shrub species is its extended leaf duration. We tested the hypothesis that L. maackii has a longer leaf duration than native shrub species found in the same habitats. Leaf phenology of L. maackii and the native deciduous shrubs Asimina triloba and Lindera benzoin was observed at four sites in central Kentucky (USA) from March until December, 2007. Additionally, a late spring freeze allowed for examination of freeze tolerance among the three test species. Lonicera maackii leaf development was two to three weeks earlier than the natives in March and early April. A hard freeze in early April caused significant (P < 0.05) leaf mortality to both of the native species (60–100% leaf mortality at 3 of 4 sites) while L. maackii showed no observable damage. L. maackii had a later transition to fall color and leaf abscission than the native species, which were at a significantly later stage of development (closer to leaf abscission) for a period of four to six weeks. These data suggest two advantages for L. maackii over potential native competitors: 1) greater access to carbon via a longer leaf duration, and 2) a greater capacity to withstand freezing temperatures.

The conservation of Helonias bullata L. (swamp pink), a federally listed species, is closely associated with the maintenance of appropriate site hydrology. Although changes to site hydrology have been implicated in the degradation and extirpation of some Helonias bullata occurrences, quantitative hydrologic data with which to assess potential hydrologic impacts to the species are lacking. We characterized site hydrology, substrate, topography, tree-canopy cover, and hydrologic regimes associated with Helonias bullata at two colonies located along small streams in the New Jersey Pinelands and assessed the potential impact of simulated water-level reductions on the species and its habitat. Over the two-year study period, surface-water levels at the two colonies fluctuated by 11.9 and 27.9 cm. Sites were characterized by muck substrate and variable topography with steep-sided hummocks in and along stream channels. Tree-canopy cover above the two colonies was 36% and 9% less than canopy cover in the adjacent forests. The pronounced difference in water levels associated with Helonias bullata plants compared with water levels associated with points where the plant was not present suggested that Helonias bullata was not uniformly distributed at the sites in relation to water table. Helonias bullata clusters, composed of groups of individual plants, were typically associated with the emergent portions of hummocks in and along the stream channels. Based on measurements at 958 clusters, the two-year median water level at the two sites was 7.9 and 10.9 cm below the base of the clusters. More than 90% of the total cluster area at both sites was associated with water levels between 5.0 (submerged) and 19.9 cm (exposed). The greatest total cluster area was associated with water levels between 5.0 and 9.9 cm, which may be the optimal water-level range for Helonias bullata. A relatively small simulated water-level drawdown of 15 cm exposed more than 30% of the cluster area at both sites to extreme hydrologic conditions, which we defined as the water level beyond which < 10% of the total existing Helonias bullata cover occurred at our sites (i.e., ≥ 20 cm below the base of the clusters). A larger simulated water-level reduction of 30 cm exposed all or nearly all of the cluster area to extreme conditions. Simulated impacts on habitat were less pronounced for smaller drawdowns because losses of suitable habitat (i.e., habitat that occurs within the 10^th and 90^th percentiles of measured water levels associated with Helonias bullata) were countered by dewatering of habitat that was previously submerged. The extent to which this dewatered habitat can compensate for losses in suitable habitat depends upon the potential for Helonias bullata to colonize the dewatered habitat. The hydrologic relationships described in this study may inform restoration efforts for this species and provide the basis for assessing potential impacts to Helonias bullata sites that are subjected to hydrologic variation.

In xeric limestone prairies (XLPs, also known as “glades”), soil depth and fire are commonly posited to affect plant community composition and structure. Through mediating plant competition, these factors may explain plant distribution in XLPs, including mechanisms driving edaphic endemism. In a XLP in southwest Missouri, the abundance of Lesquerella filiformis Rollins (Missouri bladderpod), a rare winter annual, was higher in open sites, where competition was presumably low, than in Juniperus virginiana L. (eastern redcedar) or hardwood canopy cover types; L. filiformis density in the latter two cover types varied among years. Open, J. virginiana, and hardwood canopy cover types were distributed along an increasing soil depth gradient. During flowering of L. filiformis in April, photosynthetically active radiation was highest in open sites, intermediate below hardwood canopy (prior to leaf-out), and lowest below J. virginiana canopy. Removing J. virginiana was associated with greater increases in L. filiformis abundance between 2003 and 2005, and the basal area removed was positively associated with the magnitude of increase in plant abundance. These findings supported the hypothesis that interspecific competition, presumably at least partly for light, limited the distribution of L. filiformis in XLPs. While higher plant densities indicated that L. filiformis preferred open microhabitats, thinning J. virginiana provided a management option that increased L. filiformis density in marginal habitats subject to woody plant encroachment. Because factors affecting L. filiformis germination, establishment, growth, and survival likely vary along relatively fine-scale environmental gradients, (micro-) habitat specific management of XLP vegetation may assist in protecting rare edaphic endemic plant species, such as L. filiformis.

Forested bottomland conservation areas in the midwestern and southern United States are subject to an increasingly diverse range of demands for recreational use and other ecosystem services, many dependent upon the maintenance of specific plant communities. Historical land use and other disturbances have shaped present vegetation composition, but these impacts are poorly understood. This study examined historical land use records, dendrochronological evidence, and pre- and post-tornado vegetation, with and without salvage logging, to assess forest composition changes over approximately 125 years at Mermet Lake Conservation Area in southern Illinois. This site has land use history, vegetation cover, and a management mandate common to many large river bottomland forests in the Midwest and southern USA.

The vegetation of the area prior to Euro-American settlement was primarily a forest dominated by Taxodium distichum (L.) Rich. and Nyssa aquatica L. A period of drainage and conversion to agriculture began ca. 1900 and was followed by public ownership as a conservation area since 1950. Management during this latter period was characterized by partial hydrologic restoration and complete fire suppression. The post-agriculture forest was dominated by oaks (Quercus spp.) with a transition to mixed mesophytic and bottomland hardwood forests. Following a tornado, composition and diversity within the developing stand varied along a wind intensity gradient but tended toward increasing dominance of mixed mesophytic species at the expense of Quercus. Subsequent partial salvage logging further increased vegetation complexity in response to mineral soil exposure and creation of microtopographic variation. Grading and seeding of skid trails following salvage operations produced compositionally distinct vegetation communities. Increasing prevalence of the invasive exotic Microstegium vimineum (Trin.) A. Camus., especially on salvaged plots, is expected to continue to impact vegetation communities at Mermet Lake.

Our results suggest that historical alterations in site hydrologic conditions, from pre-drainage to drainage to partial hydrological restoration, in combination with associated land use changes, produced drastic changes in forest community composition over the last century. Managers of this and similar bottomland forest areas need to consider disturbance regime changes and appropriate silvicultural strategies needed to create or maintain the historical range of vegetation types associated with sometimes disparate conservation objectives

The objective of this study was to provide stratigraphic data that could augment historic and recent ecological information concerning the cause of the 20^th century vegetation shift from grassland to pine forest in a central Maryland serpentine barren, and potentially elucidate the role of fire, chromium mining, grazing, and other disturbances on serpentine vegetation. A 46-cm core was extracted from a shallow pond draining the watershed of Chimney Branch, underlain by serpentinite. Core chronostratigraphy was established by identifying two sharp spikes in Cr concentrations correlated with historic Cr mining activity peaks and an abrupt increase in pine needle abundance, when pine populations expanded around the coring site. Pine needle analysis reveals that Pinus virginiana (Virginia pine) has been present within a few meters of the coring site since ca. 1810, 120 years earlier than previously documented. A dearth of charcoal over the 200-yr period indicates there were no major fires at the coring site after 1820, although small ground fires within a few 10 s of meters cannot be precluded. Mean sedimentation rates between dated horizons reveal that sediment efflux was highest during the early to mid-19^th century (0.51–0.35 cm yr⁻¹), synchronous with active Cr mining activity, and lowest in the past 50 years (0.19 cm yr⁻¹) when the area fell under state management and pine forest expanded. Metal concentrations were highest for Cr and Ni, followed in decreasing order by V, Zn, Cu, and Pb. Results from this study suggest that 1) while Pinus virginiana was present in the early 1800s in sheltered lowland sites chromite-mining and its erosive effects on soil development may have been a factor in suppressing pine expansion in the Chimney Branch watershed between 1820 and 1920. Hence, pine expansion may be a natural stage in the succession of serpentine vegetation related directly to increasing soil depth, with the pace of expansion influenced by fire or mining disturbance; and 2) variation in Cr concentrations combined with mining history and macrofossils, in conjunction with aerial photos and recent studies, can provide useful stratigraphic dates to establish ecosystem development within the past 200 years in a serpentine environment.

Barrens were areas of scrub oak and prairie plants that were noted by Government Land Office (GLO) Surveys in Illinois. Surveyors described four vegetation types: prairie, barrens, scattering timber, and forest. Since settlement, these vegetation types have nearly disappeared. We used GIS and the GLO survey notes for Jersey and Greene Counties in southwestern Illinois to locate and estimate the area of barrens present at the time of the surveys. Using these estimates it was possible to determine that barrens covered about ten percent of each county. The larger barrens areas were surrounded by prairie and located in the more level eastern portions of the counties. Carya spp., Quercus marilandica, and Q. stellata were more abundant in barrens than in other vegetation types. Surveyor-designated prairies were located both on prairie soils (argiudolls) and forest soils (hapludalfs). Barrens, scattering timber, and forest were all found on forest soils and on steeper slopes than prairies. The distribution of vegetation types suggests that barrens represent an expansion of woody vegetation onto especially vulnerable prairie sites during the Little Ice Age (1450 to 1850 AD).

Taxonomic determinations of previously unidentified specimens collected in 1974, and of a few newly collected taxa, have enabled us to add seven genera and 29 species to the recorded flora of the Park. Further field and laboratory work have allowed us to delete one family, one genus, and four species. Diospyros virginiana (Ebenaceae) has been deleted from the flora after intensive efforts failed to verify a 1974 visual record. The new totals for the Caumsett flora are: 103 families, 340 genera, and 660 species. One addition, Rumex pulcher, is new to New York State.