A ramp/wild leek (Allium tricoccum) is a perennial forest herb and popular non-timber forest product native to North America. Ramps are of conservation concern due to human harvest pressure but they also face a variety of other threats affecting forest understory plants overall. We investigated the possible influence of four such biotic stressors (invasive herbs, understory woody plants, and earthworms, and white-tailed deer pressure) on ramp genetic diversity and the composition of the associated understory floristic community at 10 sites with robust ramp populations (Pennsylvania, USA). We characterized deer browse pressure in a new manner, using cluster analysis based on four predictors. We saw no evidence of deer browse on ramps but detected several negative associations between biotic stressors and understory metrics: (1) fewer palatable summer Liliaceous taxa (Maianthemum spp., Medeola virginiana, Polygonatum spp., Streptopus lanceolatus var. lanceolatus, Trillium spp., and Uvularia spp.) as deer pressure increased; (2) lower native herbaceous species richness with more invasive herb abundance; and (3) lower ramp clonal richness with greater invasive shrub abundance. Our results support the premise that land managers and ramp stewards should consider the influence of biotic stressors, particularly invasive species, when attempting to conserve ramp populations and their associated forest understory communities.
Identifying a region's full range of remnant plant communities is necessary for conserving its biodiversity and associated ecosystem services. The eastern North American Coastal Plain, known for its historical dominance by longleaf pine-wiregrass (Pinus palustris-Aristida beyrichiana) communities, appears to have had embedded landscapes that supported shortleaf pine-oak-hickory-bluestem (Pinus echinata, Quercus spp., Carya spp., Schizachyrium scoparium) communities that preexist European colonization. We investigated this hypothesis in a local landscape by comparing plant community composition and soil chemical and physical characteristics between remnant longleaf pine and shortleaf pine communities and between shortleaf pine and old-field (post-tillage) pine communities. Ordination showed overall plant species composition to be distinct among the three communities, although with considerable overlap of generalist species. Most indicator species for the shortleaf pine community were identified in previous studies as associated with remnant pine communities and having life-history characteristics associated with “old-growth” grasslands, unlike those for the old-field pine community. They also had more northeastern continental distributions than indicator species for the longleaf pine community. The shortleaf pine community generally had higher soil pH and nutrient concentrations than both longleaf pine and old-field pine communities, and it differed from longleaf pine communities in most physical soil characteristics, consistent with broader distributions of the dominant tree species. These results indicate that shortleaf pine-oak-hickory-bluestem is a remnant community distributed according to local physiographic characteristics. We suggest that this community be identified and prioritized for conservation, including minimizing soil disturbance and maintaining appropriate fire regimes.
Evaluating vegetation management outcomes is challenging because plant community datasets can be summarized in multiple ways, and the chosen metric strongly shapes conclusions. The Floristic Quality Index (FQI), which combines mean species conservatism values (mean Cnat) with measures of species richness or nativity, provides a useful framework for monitoring management outcomes. However, averaging C-values across species reduces the metric's sensitivity and limits its ability to detect change. Using plot data from 23 South Sound prairie sites in Washington State, I assessed the efficacy of a modified FQI equation (FQI(cmax)) that used the maximum C-value (max C) instead of the mean C-nat, weighted by conservatism-weighted cover-based nativity (nativityc-val). First, I evaluated whether species with high C-values tended to co-occur more often than expected by chance and whether max C correlated positively with the cover and richness of high-fidelity prairie taxa. Second, I evaluated the correlation between max C and mean C-nat, native richness, nativity(c-val), and three formulations of FQI. Mean C-nat was not correlated with max C, could distinguish only two community states, and was heavily influenced by species richness and the regional C-value distribution. In contrast, FQI(cmax )identified seven community quality states, was independent of species richness and nativity(c-val), and was less affected by the regional C-value distribution. Lastly, FQI(cmax )was better able to detect directional responses to four management interventions than the three other FQI equations. The components of FQIcmax also revealed a trade-off among species conservatism, native richness, and native cover in relation to herbicide use. These findings suggest that a simple change to the FQI formula can significantly enhance its sensitivity and clarity when assessing management impacts.
Hybridization between rare and common congeners poses a recurring challenge in ecological restoration, particularly where multiple conservation priorities overlap. In Pacific Northwest prairies, restoration efforts have brought the federally listed plant Castilleja levisecta into close proximity with C. hispida, a more common congener widely planted to support the endangered Taylor's checkerspot butterfly (Euphydryas editha taylori). Concerns have emerged that hybridization could compromise the genetic integrity of C. levisecta, yet the strength of reproductive barriers-particularly across cytotypes of C. hispida-has not been quantified. We conducted a controlled greenhouse crossing experiment to evaluate post-pollination reproductive barriers between diploid C. levisecta and diploid, tetraploid, and hexaploid cytotypes of C. hispida. We quantified fruit set, seed set, and germination across reciprocal F1 crosses, assessed fertility of early-generation hybrids through backcrosses and conspecific hybrid crosses, and integrated these components into estimates of overall postzygotic reproductive isolation. Homoploid crosses between diploid C. levisecta and diploid C. hispida produced viable, partially fertile offspring with low reproductive isolation. In contrast, interploid crosses involving tetraploid or hexaploid C. hispida exhibited sharply reduced seed set and germination, resulting in strong to nearly complete isolation. Backcrosses involving diploid hybrids showed little isolation, whereas those involving triploid hybrids were largely infertile. These results indicate that ploidy is a key determinant of postzygotic isolation in this system and suggest that polyploid C. hispida cytotypes represent lower-risk options where co-planting is being evaluated. However, field studies assessing premating barriers and hybrid establishment remain essential for site-specific restoration decisions.
Low native plant establishment is a common restoration challenge for land managers across many ecosystems. Determining the factors contributing to low establishment rates can lead to more efficient use of limited seed resources. This study considers seed predation as a potential factor for low establishment rates. Using exclusion cages, we tested whether small mammals and invertebrates prey on seed from five functionally important plants with less than 5% establishment rates in Pacific Northwest prairie restoration sites: Balsamorhiza deltoidea, Dichanthelium oligosanthes, Armeria maritima, Viola adunca, and Erigeron speciosus. Three other factors were tested to assess their influence on predation: seed size, edge effect, and timing. Five recently burned prairies were studied, each containing three 50 m transects with three plots along a gradient from the unburned edge to the burned interior. Each plot contained four subplots (treatments): insect entry, small mammal entry, complete exclusion (control), and no exclusion. Each treatment contained a seed dish with 20 seeds of each species. Seeds remaining were counted after 24 hr, 48 hr, and 1 wk. A preference by small mammals for larger-sized seeds (B. deltoidea, D. oligosanthes, and V. adunca) was found. No seed predation by insects was observed under the specific conditions of this study. Furthermore, no evidence of an edge effect influencing predation was found, and the majority of seed predation by small mammals was observed at the 1 wk count. These findings suggest that attempting to mitigate seed losses from small mammals could positively influence seedling establishment rates for species with larger seeds.
