Early detection of an invading nonindigenous plant species (NIS) may be critical for efficient and effective management. Adaptive survey sampling methods may provide unbiased sampling for best estimates of distribution of rare and spatially clustered populations of plants in the early stages of invasion. However, there are few examples of these methods being used for nonnative plant surveys in which travelling distances away from an initial or source patch, or away from a road or trail, can be time consuming due to the topography and vegetation. Nor is there guidance as to which of the many adaptive methods would be most appropriate as a basis for invasive plant mapping and subsequent management. Here we used an empirical complete census of four invader species in early to middle stages of invasion in a management area to assess the effectiveness and efficiency of three nonadaptive methods, four adaptive cluster methods, and four adaptive web sampling methods that all originated from transects. The adaptive methods generally sampled more NIS-occupied cells and patches than standard transect approaches. Sampling along roads only was time-efficient and effective, but only for species with restricted distribution along the roads. When populations were more patchy and dispersed over the landscape the adaptive cluster starting at the road generally proved to be the most time-efficient and effective NIS detection method.

Because of its ability to transform ecosystems by increasing the prevalence of fire, the invasive annual red brome is a priority exotic species for management in arid lands of the southwestern United States. By sampling red brome presence and 97 environmental (climatic, topographic, and soil) and native vegetation (e.g., perennial species richness) variables on 126 sites, we assessed biophysical correlates with red brome distribution on a 755,000-ha (1.9 million ac) Mojave Desert landscape. Brome occupied 55 of 126 (44%) 0.09-ha plots. The simplest models (i.e., those containing the fewest or most easily obtained variables) in multivariate (classification trees and nonparametric multiplicative regression) and univariate (χ2) models often portrayed red brome distribution as well, or nearly as well, as more complicated models containing more variables harder to obtain. The models varied, however, in their abilities for describing brome presence compared with absence. For example, a simple classification tree using only elevation, soil great group, parent material, and vegetation type improved estimates of brome presence for 55% of sites, absences for 87%, and overall for 73% of sites compared with a naïve model containing the observed frequency of brome in the data. Conversely, a more complicated model, including soil boron and sulfur, performed better for presences (96%) than for absences (73%; 83% overall). Results also showed variable support for two general postulates in invasive species science. Red brome distribution was not correlated with soil N, which is inconsistent with the supposition that nutrient-rich soils are more prone to invasion. Brome was correlated with native perennial species richness to support the postulate that exotic species abundance is correlated with species-rich habitats, but the correlation was weak (r = 0.38) and similar in strength to correlations with many other environmental variables. On this relatively low-elevation landscape, the areas currently most invaded by red brome include the higher elevations (> 777 m [2,549 ft]), limestone–sandstone soils, and burrobush and mixed perennial communities. Areas least inhabited by brome are the lowest elevations (< 491 m), gypsum soils, and creosotebush and saltbush communities.

Water hyacinth is among the most widespread invasive plants worldwide; however, its effects on waterbirds are largely undocumented. We monitored site use by waterbirds at Lake Chapala, the largest lake in Mexico and recently designated Ramsar site, to evaluate the potential influence of water hyacinth cover on species composition and aggregate measures of the waterbird community, including waterbird density, species richness, and Simpson's index of diversity. We examined the response of waterbirds to changes in percent water hyacinth cover at 22 independent sites around the lake during six study seasons from May 2006 to February 2008. We found little evidence to suggest that percent water hyacinth cover affected aggregate community measures; however, multivariate analysis of relative species composition suggested that water hyacinth cover corresponded with seasonal species composition (Canonical Correspondence r = 0.66, P = 0.007) when seasonal site cover averaged 17.7 ± 4.67% (winter 2007). Several migratory species were not observed during this season, which could suggest that some small-bodied migratory species avoided Lake Chapala during the winter of high water hyacinth cover. We suspect that observed changes in the waterbird community are in response to species-specific tolerances for water hyacinth and indirect abiotic and biotic effects of its presence (e.g., invertebrate and fish composition).

Producers facing infestations of invasive annual grasses regularly voice the need for practical revegetation strategies that can be applied across broad landscapes. Our objective was to determine the potential for scaling up the single-entry approach for revegetating medusahead-infested rangeland to broader, more heterogeneous landscape-scale revegetation of winter annual grass–infested rangeland. We hypothesized, when applied on a highly variable landscape scale, the combination of imazapic and seeding would provide highest abundance of perennial grasses and lowest amount of annual grasses. Treatments included a control, seeding of crested wheatgrass (‘Hycrest’) and Sandberg's bluegrass, spraying (60 g ai ha−1 imazapic), and a simultaneously applied combination of spraying and seeding. The HyCrest and Sandberg's bluegrass seeding rates were 19 and 3.4 kg ha−1, respectively. The treatments were applied to large plots (1.4 to 8 ha) and replicated five times, with each replication located in different watersheds throughout southeastern Oregon. This study shows that the single-entry approach can be scaled up to larger landscapes, but variation within establishment areas will likely be high. This procedure should reduce the costs over multientry treatment applications and make revegetating annual grass–infested rangeland across landscapes more affordable.

Conservation and land management decisions often are based primarily on natural science, but could be more successful if human influences were effectively integrated into decision making. This is especially true for efforts to manage invasive plants, whose arrival is usually the product of deliberate human introduction. Risk-assessment models that predict the probability that a nonnative plant will naturalize or invade are useful tools for managing invasive plants. However, decisions based on such models could affect stakeholders differently. Careful assessment of risk-analysis methodologies should consider the importance of stakeholder participation. We surveyed the perceptions of four stakeholder groups (conservation professionals, master gardeners, professional horticulturists, and woodland landowners) in Iowa about invasive plants, general management approaches, and risk-assessment models. We also examined whether or not a stakeholder's nature relatedness plays a role in shaping his or her responses. Stakeholder perceptions varied less than expected across all four groups. Eighty-seven percent of respondents agreed invasive plants are a problem, and 88.4% agreed that we have a responsibility to manage them to protect natural areas. Support for the use of risk-assessment models also was high, with 78.7% of respondents agreeing that their use has potential to prevent plant invasions. Nature relatedness scores for all groups were correlated with respondent perspectives on invasive plants. Respondents believed biologically significant error rates (errors that might introduce a new invasive plant) should not exceed 5 to 10%. Respondents were more tolerant of horticulturally limiting errors (errors that restrict sale/use of a plant that would not have become invasive), reporting rates of 10 to 20% as acceptable. Researchers developing risk-assessment models might wish to aim for error rates within these bounds. General agreement among these stakeholder groups suggests potential support for future risk-management efforts related to invasive plants.

Understanding the ecological characteristics of areas invaded and not invaded by exotic plants is a priority for invasive plant science and management. Buffelgrass is an invasive perennial species that managers view as a major threat to indigenous ecosystems of conservation lands in Australia, Mexico, the United States, and other locations where the species is not native. At 14 sites in Saguaro National Park in the Arizona Uplands of the Sonoran Desert, we compared the soil, vegetation, and soil seed bank of patches invaded and not invaded by buffelgrass. Abiotic variables, such as slope aspect and soil texture, did not differ between buffelgrass patches and patches without buffelgrass. In contrast, variables under primarily biotic control differed between patch types. Soil nutrients, such as organic C and NO3–N, were approximately twofold greater in buffelgrass compared with nonbuffelgrass patches. Average native species richness was identical (14 species 100 m−2) between patch types, but native plant cover was 43% lower in buffelgrass patches. Unexpectedly, native seed-bank densities did not differ significantly between patch types and were 40% greater than buffelgrass seed density below buffelgrass canopies. Results suggest that (1) soil nutrient status should not be unfavorable for native plant colonization at buffelgrass sites if buffelgrass is treated; (2) at least in the early stages of buffelgrass patch formation (studied patches were about 10 yr old), native vegetation species were not excluded, but rather, their cover was reduced; and (3) native soil seed banks were not reduced in buffelgrass patches.

The Asian grass Miscanthus sinensis (Poaceae) is being considered for use as a bioenergy crop in the U.S. Corn Belt. Originally introduced to the United States for ornamental plantings, it escaped, forming invasive populations. The concern is that naturalized M. sinensis populations have evolved shade tolerance. We tested the hypothesis that seedlings from within the invasive U.S. range of M. sinensis would display traits associated with shade tolerance, namely increased area for light capture and phenotypic plasticity, compared with seedlings from the native Japanese populations. In a common garden experiment, seedlings of 80 half-sib maternal lines were grown from the native range (Japan) and 60 half-sib maternal lines from the invasive range (U.S.) under four light levels. Seedling leaf area, leaf size, growth, and biomass allocation were measured on the resulting seedlings after 12 wk. Seedlings from both regions responded strongly to the light gradient. High light conditions resulted in seedlings with greater leaf area, larger leaves, and a shift to greater belowground biomass investment, compared with shaded seedlings. Japanese seedlings produced more biomass and total leaf area than U.S. seedlings across all light levels. Generally, U.S. and Japanese seedlings allocated a similar amount of biomass to foliage and equal leaf area per leaf mass. Subtle differences in light response by region were observed for total leaf area, mass, growth, and leaf size. U.S. seedlings had slightly higher plasticity for total mass and leaf area but lower plasticity for measures of biomass allocation and leaf traits compared with Japanese seedlings. Our results do not provide general support for the hypothesis of increased M. sinensis shade tolerance within its introduced U.S. range compared with native Japanese populations.

Perennial pepperweed is an invasive plant that is expanding rapidly in several plant communities in the western United States. In California, perennial pepperweed has aggressively invaded seasonal wetlands, resulting in degradation of habitat quality. We evaluated the rate and dynamics of population spread, assessed the effect of disturbance on spread, and determined the biotic and abiotic factors influencing the likelihood of invasion. The study was conducted at eight sites within two wetland regions of California. Results indicate that in undisturbed sites, spread was almost exclusively through vegetative expansion, and the average rate of spread was 0.85 m yr−1 from the leading edge. Spread in sites that were disked was more than three times greater than in undisturbed sites. While smaller infestations increased at a faster rate compared with larger populations, larger infestations accumulated more newly infested areas than smaller infestations from year to year. Stem density was consistently higher in the center of the infestations, with about 2.4 times more stems per square meter compared with the leading edge at the perimeter of the population. The invasion by perennial pepperweed was positively correlated with increased water availability but was negatively correlated with the cover of perennial and annual species. Thus, high cover of resident vegetation can have a suppressive effect on the rate of invasion, even in wetland ecosystems. On the basis of these results, we recommend that resident plant cover not be disturbed, especially in wet areas adjacent to areas currently infested with perennial pepperweed. For infested areas, management efforts should be prioritized to focus on controlling satellite populations as well as the leading edge of larger infestations first. This strategy could reduce the need for costly active restoration efforts by maximizing the probability of successful re-establishment of resident vegetation from the adjacent seedbank.

Japanese stiltgrass is one of the most aggressive, rapidly spreading invasive plants in the eastern United States. Management guidelines state that mowing can help manage Japanese stiltgrass but that mowing is most effective when done late in the season after the plants begin to flower and before they set seed. In this study, I tested the effectiveness of mowing at three different times between mid-June and early September in 2009 and 2010, as well as mowing twice in 1 yr and for two consecutive years. The effectiveness of mowing Japanese stiltgrass was determined by measuring percentage of cover, biomass, seed production, and the number of stems in the summer following mowing. All mowing treatments significantly reduced percentage of cover, biomass, seed production, and the number of Japanese stiltgrass stems the following year. In 2009, all of the mowing treatments significantly reduced biomass, percentage of cover, and seed production. The latest mow, at the end of August, resulted in a slightly greater reduction of cleistogamous seeds. In 2010, the earliest mowing treatment, in mid-June, did not reduce cover and biomass as much as the other mowing treatments. Overall, these results suggest that mowing can be an effective control method for Japanese stiltgrass and that mowing any time after June should be effectively equivalent, although later mowing may provide some marginal advantage.

Chinese and European privets are among the most aggressive invasive shrubs in forestlands of the southern United States. We analyzed extensive field data collected by the U.S. Forest Service covering 12 states to identify potential determinants of invasion and to predict likelihood of further invasion under a variety of possible management strategies. Results of multiple logistic regression, which classified 75% of the field plots correctly with regard to species presence and absence, indicated probability of invasion is correlated positively with elevation, adjacency (within 300 m) to waterbodies, mean extreme maximum temperature, site productivity, species diversity, natural regeneration, wind disturbance, animal disturbance, and private land ownership and is correlated negatively with slope, stand age, site preparation, artificial regeneration, distance to the nearest road, fire disturbance, and public land ownership. Habitats most at risk to further invasion (likelihood of invasion > 10%) under current conditions occur throughout Mississippi, with a band stretching eastward across south-central Alabama, and in eastern Texas and western Louisiana. Invasion likelihoods could be reduced most by conversion to public land ownership, followed by site preparation, fire disturbance, artificial regeneration, and elimination of animal disturbance. While conversion of land ownership may be neither feasible nor desirable, this result emphasizes the opportunity for reducing the likelihood of invasions on private lands via increased use of selected management practices.

We investigated nonindustrial private forest (NIPF) owners' invasive plant risk perceptions and mitigation practices using statistical analysis of mail survey data and qualitative analysis of interview data collected in Oregon's ponderosa pine zone. We found that 52% of the survey sample was aware of invasive plant species considered problematic by local natural resource professionals; 70% was concerned about these species; and 46% had treated invasive plants on their parcels. Owners' perceptions of invasive plant risks fell along a spectrum ranging from a lack of awareness or concern, to the view that invasive plant infestations have discrete causes and controllable consequences, to the perception that incursions by invasive plants have diffuse causes and uncontrollable effects. Being aware or concerned about invasive plant species were predictors (p ≤ 0.001) of whether owners treat their parcels to control invasive plants. Holding wildlife habitat and/or biodiversity as an important forest management goal was also a predictor (p ≤ 0.08) of whether owners treated their parcels to control invasive plants. Some owners were sensitive to the risks of invasive plant infestations from nearby properties, and a surprisingly high percentage of respondents had cooperated with others in forest management activities previously. Our findings suggest three approaches to increasing the frequency of invasive plant mitigation by NIPF owners that hold promise: (1) raising awareness and concern about invasive plants and their impacts on forest management goals that owners care about, such as wildlife habitat and/or biodiversity; (2) providing assistance to help owners mitigate invasive plants they feel unable to control; and (3) engaging owners in coordinated efforts across ownership boundaries to address invasive plant risks.

Integrated weed management strategies (IWM) are being advocated and employed to control invasive plants species. In this study, we compared three management strategies (biological control alone [BC], BC with fire [BC + F], and BC with mowing [BC + M]) to determine if physical controls reduce seed production by Scotch broom and interfere with the action of the biological control agent—the Scotch broom seed weevil. We measured seed production and seed predation by the weevil at both pod and plant scale, and seed bank density over two field seasons. We found no difference in the number of seeds per pod among management strategies. However, combining management strategies (BC + M and BC + F) resulted in significant reductions in pods per plant, mature seeds per plant, and seed bank density relative to biological control alone. We did not find differences among management strategies in number of weevils per pod or proportion of seeds predated by the weevil at either pod or whole-plant scale. However, combining management strategies (BC + M and BC + F) resulted in a significant reduction in healthy mature seeds per plant relative to biological control alone. Although both integrated strategies outperformed biological control alone in reducing seed production and the seed bank, with no statistical difference between them, we propose that short-rotation prescribed fire could prove to be a more effective strategy for long-term management of Scotch broom due to its potential for slightly greater depletion of the seed bank.

Invasive species researchers often ask: Why do some species invade certain habitats while others do not? Ecological theories predict that taxonomically related species may invade similar habitats, but some related species exhibit contrasting invasion patterns. Brassica nigra, Brassica tournefortii, and Hirschfeldia incana are dominant, closely related nonnative species that have overlapping, but dissimilar, distributions. Brassica tournefortii is rapidly spreading in warm deserts of the southwestern United States, whereas B. nigra and H. incana are primarily limited to semiarid and mesic regions. We compared traits of B. tournefortii that might confer invasiveness in deserts with those of related species that have not invaded desert ecosystems. Brassica tournefortii, B. nigra and H. incana were compared in controlled experiments conducted outdoors in a mesic site (Riverside, CA) and a desert site (Blue Diamond, NV), and in greenhouses, over 3 yr. Desert and mesic B. tournefortii populations were also compared to determine whether locally adapted ecotypes contribute to desert invasion. Experimental variables included common garden sites and soil water availability. Response variables included emergence, growth, phenology, and reproduction. There was no evidence for B. tournefortii ecotypes, but B. tournefortii had a more rapid phenology than B. nigra or H. incana. Brassica tournefortii was less affected by site and water availability than B. nigra and H. incana, but was smaller and less fecund regardless of experimental conditions. Rapid phenology allows B. tournefortii to reproduce consistently under variable, stressful conditions such as those found in Southwestern deserts. Although more successful in milder, mesic ecosystems, B. nigra and H. incana may be limited by their ability to reproduce under desert conditions. Rapid phenology and drought response partition invasion patterns of nonnative mustards along a gradient of aridity in the southwestern United States, and may serve as a predictive trait for other potential invaders of arid and highly variable ecosystems.

Where the nonnative annual grass downy brome proliferates, it has changed ecosystem processes, such as nutrient, energy, and water cycles; successional pathways; and fire regimes. The objective of this study was to develop a model that predicts the presence of downy brome in Central Oregon and to test whether high presence correlates with greater cover. Understory data from the U.S. Department of Agriculture (USDA) Forest Service's Current Vegetation Survey (CVS) database for the Deschutes National Forest, the Ochoco National Forest, and the Crooked River National Grassland were compiled, and the presence of downy brome was determined for 1,092 systematically located plots. Logistic regression techniques were used to develop models for predicting downy brome populations. For the landscape including the eastside of the Cascade Mountains to the northwestern edge of the Great Basin, the following were selected as the best predictors of downy brome: low average March precipitation, warm minimum May temperature, few total trees per acre, many western junipers per acre, and a short distance to nearest road. The concordance index = 0.92. Using the equation from logistic regression, a probability for downy brome infestation was calculated for each CVS plot. The plots were assigned to a plant association group (PAG), and the average probability was calculated for the PAGs in which the CVS plots were located. This method could be duplicated in other areas where vegetation inventories take place.

Control options for the nonnative common mullein are of increasing interest to land managers in the west. Common mullein is a prolific seed producer, with a single plant able to produce well over 100,000 seeds. We found that mechanical control of common mullein before mature seed capsules developed along the raceme significantly reduced viable seed production. Seeds from immature capsules had very low viability (early reproductive stage = 0.08%, 95% CI = 0.06%, 0.67%; mid reproductive stage = 1.52%, 95% CI = 0.49%, 3.11%). This information allows managers to time their management efforts so that they can reduce the amount of plant material that must be disposed of in order to control the spread of common mullein seeds.

Japanese barberry is an important landscape shrub that has a demonstrated potential to be invasive in ∼30 states across the central and northern United States. Forty-six cultivars were evaluated for their potential to produce fruits and seeds in a randomized, replicated field planting. Seeds from a subset of cultivars were evaluated for their ability to germinate and survive as seedlings in a deciduous woodland. Seed production for cultivars varied from no seeds to more than 12,000 seeds plant−1 and the number of seeds per fruit ranged from 0.1 to 1.8. Five cultivars produced fewer than 100 seeds plant−1, and two cultivars failed to produce fruit. When plants were allowed to mature for 4 to 5 yr beyond the first evaluation time, cultivars exhibited significant increases in fruits per plant, producing as much as 35,000 fruits plant−1 (‘Sparkle’). ‘Golden Devine’ and ‘Red Chief’, fruitless cultivars at the first evaluation, produced 165 and 20 fruit plant−1, respectively, at the follow-up evaluation, demonstrating that long-term evaluation of cultivars is necessary to accurately assess sterility. Between 12.5 and 31% cultivar seed sown in a deciduous woodland germinated, and seedlings survived at rates between 5.6 and 29.3%. Coupling cultivar seed-production data with germination and survival data in a deciduous woodland suggests that even cultivars producing as few as 100 seeds annually have the potential to contribute a few seedlings each year to a natural area.

Roads function as prime habitats and corridors for invasive plant species, and can contribute significantly to the spread and establishment of weeds inside protected areas. Smooth bedstraw (Galium mollugo) populations have recently expanded in the abandoned agricultural fields of Bic National Park (Quebec, Canada), and may represent a threat to the preservation of plant diversity and to the quality of visitor experience. The main objective of this study was to map the distribution of the species in the park and to delineate factors influencing the abundance of the plant in fallow fields. We hypothesized that road proximity was the main factor explaining the presence and abundance (frequency of recording) of smooth bedstraw in fields. Vegetation surveys were conducted in abandoned agricultural fields, and two logistic regression models were built to examine the relationship between the presence and abundance of smooth bedstraw and environmental and historical variables. Smooth bedstraw populations were also mapped along transportation corridors. The abundance of smooth bedstraw significantly increases within 125 m (410 ft) of a paved road. The recent proliferation of smooth bedstraw in Bic National Park is likely associated with the construction of road embankments during the paving of gravel roads. The paved road network has probably acted as a dispersal corridor for smooth bedstraw, while the abandoned agricultural fields located close to the paved roads provided suitable habitats facilitating population establishment over large areas. We recommended to the park's authorities to cease paving all remaining gravel roads in the park, not only to stop smooth bedstraw invasion, but also the spread of other invasive species of concern.

Invasive saltcedar species have replaced native riparian trees on numerous river systems throughout the western United States, raising concerns about how this habitat conversion may affect wildlife. For periods ranging from 2 to 11 yr, I used live-trapping to monitor small mammal populations in paired saltcedar and native riparian woodlands at four sites in western Nevada and eastern California. Heteromyid rodents, such as Merriam's and Ord's kangaroo rats, were more likely to occur in saltcedar habitats, but other rodent species, particularly the montane vole and western harvest mouse, occurred more often in native habitats, and this balanced species richness in habitat comparisons. The most common species at all sites, the deer mouse, did not show any consistent differences in abundance or in mean body mass between the two habitat types. However, the ratio of captured male to female deer mice was higher in saltcedar than native habitats at two sites. Deer mice as well as Ord's kangaroo rats also had higher rates of being recaptured following initial capture in native habitats, which may have been due to fewer transient individuals occurring in these habitats. By contrast, Merriam's kangaroo rats may have been more transient in native habitats because they were more likely to be recaptured in saltcedar. Individuals of two species, pinyon mouse and white-tailed antelope ground squirrel, had greater mean body mass in native habitats than they did in saltcedar, implying that they may have maintained superior condition in native habitats.

The effects of annual, aerial and ground, herbicide treatments with glyphosate and metsulfuron were evaluated for control of Old World climbing fern (OWCF) and effects on native plants on tree islands in Arthur R. Marshall Loxahatchee National Wildlife Refuge during 2006 to 2009. Initial aerial herbicide treatments reduced OWCF cover by greater than 98% on metsulfuron-treated islands and greater than 88% on glyphosate-treated islands, but there was a concomitant decrease in native ground cover with both herbicides. Follow-up ground treatments, during years two and three of the study, were effective at maintaining low levels of OWCF. OWCF cover at the end of the study was 1 to 2% of pretreatment cover on metsulfuron-treated islands and 8 to 10% on glyphosate-treated islands. At the end of the study (3 yr after treatment), species richness was dominated by ruderal native species not typically found on tree islands. The survival rate of tree and shrubs was 65 to 93% on islands treated with metsulfuron and 6 to 20% on islands treated with glyphosate. Minimum effects were recorded for canopy cover on tree islands treated with metsulfuron compared with glyphosate. These data indicate that the aerial application of metsulfuron can be used for control of OWCF on tree islands. Follow-up ground treatments will be required for OWCF regrowth and new sporelings and should be conducted within 1 yr of the aerial application.