Japanese stiltgrass is a nonnative invasive grass occupying a range of habitats in the eastern United States. Conventional management recommendations include hand-removal, mowing, or a nonselective herbicide application in autumn prior to flowering. However, no study has directly compared the ecological impacts of long-term management strategies on Japanese stiltgrass populations or recruitment and establishment of native flora. An experiment was initiated in 2002 and continued for three growing seasons in mixed pine-hardwood forests in central North Carolina. Conventional treatments included hand-removal, mowing, or an application of glyphosate (1.1 kg ai/ha) once in autumn, and selective removal by hand or fenoxaprop-P (0.19 kg ai/ha) season-long as needed. All treatments were compared to nontreated plots. Percent vegetation cover by species was recorded twice annually. Data were aggregated into five classes; Japanese stiltgrass, other exotic plants, native forbs, native monocots, and native woody plants. The soil seed bank of all species was estimated annually by extracting soil cores and documenting seedling emergence. All Japanese stiltgrass management treatments significantly reduced Japanese stiltgrass cover and seed bank over time compared to no management. However, recruitment and reestablishment of native plants and overall species richness were greater with selective Japanese stiltgrass management treatments including both hand-removal and fenoxaprop-P. Relative cover of other exotic plants decreased 2% to 49% after 3 yr with all Japanese stiltgrass management treatments except season-long hand-removal, which increased relative cover of other exotic plants 51%.

Medusahead is among the most invasive grasses in the western United States. Selective control of this noxious winter annual grass is difficult in California grasslands, as many other desirable annual grasses and both native and nonnative broadleaf forbs are also important components of the rangeland system. Intensive grazing management using sheep is one control option. This study was designed to determine the optimal timing for sheep grazing on heavily infested medusahead sites, and to evaluate the changes in species composition with different grazing regimes. Midspring (April/May) grazing reduced medusahead cover by 86 to 100% relative to ungrazed plots, regardless of whether it was used in combination with early spring or fall grazing. Early spring (March) or fall (October to November) grazing, alone or in combination, was ineffective for control of medusahead. In addition, midspring grazing increased forb cover, native forb species richness, and overall plant diversity. At the midspring grazing timing, medusahead was in the “boot” stage, just prior to exposure of the inflorescences. The success of this timely grazing system required high animal densities for short periods. Although this approach may be effective in some areas, the timing window is fairly narrow and the animal stocking rates are high. Thus, sheep grazing is unlikely to be a practical solution for management of large medusahead infestations

Aphthona spp. flea beetles have established and reduced the density of leafy spurge in much of the western United States. One way to measure the long-term impact and effectiveness of a weed control program is by monitoring the changes in the seedbank over time. The change in leafy spurge stem density and seed in the seedbank were evaluated 5 yr after Aphthona spp. were released to control this weed in the Little Missouri National Grasslands in southwestern North Dakota. Leafy spurge density and seed in both loamy overflow and loamy ecological sites decreased, whereas desirable (high-seral) forbs increased 5 yr after the biological control agents were released. Leafy spurge topgrowth was reduced from an average of over 200 stems/m2 to less than 8 stems/m2 in the most densely infested sites, and leafy spurge seed was reduced from an average of 68% of the seedbank to only 14% in both ecological sites. High-seral forb seed increased by over 300% in the seedbank, which indicated the floristic quality of the sites, was returning to a preinfestation state. Species with the largest increase included western rock jasmine and fringed sage, which increased at least three-fold in both sites. Less desirable low-seral forbs and grasses accounted for about 30% of the seedbank. In a greenhouse study, native grass production was reduced nearly 50% when grown in soil from Aphthona spp. release sites compared to nonrelease sites. Switchgrass production was reduced to a greater extent (66%) than green needlegrass, little bluestem, or western wheatgrass. The cause and extent of reduced native grass production in sites where Aphthona spp. were released has yet to be determined. The decrease in leafy spurge topgrowth and seed in the soil seedbank as desirable species seed increased, should lead to the long-term recovery of the plant community.

To determine the extent and nature of container-grown plant soil as a pathway for introduction of exotic plant species to Alaska, soil from container-grown ornamentals was obtained from vendors and was incubated in the greenhouse. Fifty-four plant species were identified growing in containers or germinating from the soil, and included Canada thistle—a prohibited weed in Alaska—and nine other species listed as invasive in Alaska. The number of species and estimated seed bank were very low for soil from vegetable starts/herbs and herbaceous bedding plants (< 2 seedlings/L soil), but was greater for soil from containers containing woody plants, especially balled and burlapped ornamentals (20 seedlings/L soil). Container alien plant seed bank size was strongly related to type of soil. Potting (soil-less) soil contained 1.2 germinating seeds/L, soil-based soil 5.5 seeds/L, and mineral soil 18.7 seeds/L. Growers and vendors were variables that also influenced the size of the container seed bank, suggesting that weed management practiced during production and at the point of sale can greatly influence seed banks of ornamental containers.

Giant reed is an invasive plant of riparian habitats throughout California and the United States. Glyphosate is approved for controlling giant reed in California. Sources indicate that 1.5% to 5% glyphosate solutions are effective at controlling giant reed. There is little published data on the relative efficacy of different concentrations of glyphosate which can be used to select an appropriate application regime for California habitats. We conducted two field experiments to test the hypothesis that glyphosate concentrations of 1.5%, 3%, and 5% applied as foliar sprays were equally effective at killing giant reed plants. Leaf chlorophyll content and the proportion of living stems declined significantly following treatment with 1.5% or greater solutions of glyphosate. New stems were observed the spring following treatment for plants treated with 1.5% glyphosate. No new stems were observed for plants treated with either 3% or 5% glyphosate. A treatment that included “bending and breaking” stems prior to treatment with 5% glyphosate provided control similar to 5% glyphosate alone. There was no there evidence that plants sprayed with only a mixture of the surfactant (Agridex), water, and a marking dye were affected beyond the short-term. These results indicate that 3% or 5% foliar applications with glyphosate were the most effective and consistent treatments for killing giant reed with a single late-season application. This result is especially important if the goal of the treatment program is to minimize the number of treatments, thus reducing labor costs and minimizing impacts on sensitive habitats by reducing the number of site visits.

Western juniper has been actively invading sagebrush plant communities for about 130 yr. Western juniper canopy cover generally increases as western juniper invades sagebrush steppe communities and succession progresses toward a western juniper woodland. Our goal was to estimate the impact of juniper invasion and canopy increase on understory vegetation structure and productivity on 101 sites in northeastern California. The primary objectives of this study were to: (1) examine the influence of increasing western juniper canopy cover on the composition and productivity of understory vegetation; and (2) assess the effects of western juniper removal on understory vegetation. Sites in early, mid-, and late successional stages and sites on the same soils that had not been invaded were selected. Sites where western juniper had been removed by prescribed fire, mechanical, or chemical methods were compared to adjacent untreated sites. Western juniper canopy cover, understory cover and species composition, productivity, and bare ground were determined at each site during May through July 2005 and 2006. Regression analysis was used to evaluate the relationship between western juniper canopy cover and understory vegetation parameters. Logistic regression was used to detect understory differences between treated (juniper removed) and untreated (juniper not removed) sites. A significant relationship was found between western juniper canopy cover and understory species richness, shrub cover, forb cover, total grass cover, cheatgrass cover, herbaceous productivity, and bare ground. Removal of western juniper increased total grass cover, cheatgrass cover, and productivity, and reduced bare ground. The results of this study support findings by researchers in other states that western juniper influences plant community structure and productivity, and removal of western juniper might reverse these changes in structure, but also might increase opportunities for invasion of cheatgrass.

Growth chamber studies were conducted using 55-L aquariums to evaluate the efficacy of the herbicide triclopyr and the fungal pathogen Mycoleptodiscus terrestris, applied alone and in combination against Eurasian watermilfoil. Treatments included 0.15, 0.40, and 1.50 mg acid equivalent (ae)/L triclopyr, 0.08, 0.16, and 0.32 ml/L M. terrestris, combinations of both agents at all rates, and an untreated control. Plants were exposed to all treatments for a 24-h contact time and plant biomass (shoot and roots) was recorded 6 wk after application. For both triclopyr and M. terrestris applied alone, plant control increased with treatment concentration. Compared with untreated plants, 1.50 mg/L triclopyr and 0.32 ml/L M. terrestris reduced Eurasian watermilfoil by 100 and 91%, respectively. Lower doses of herbicide or pathogen were less effective and plant recovery was observed from surviving plant tissues (stems and root crowns). Although M. terrestris at 0.08 ml/L did not significantly reduce shoot or root biomass and 0.15 mg/L triclopyr provided only 53% control of plants, combining both agents at these rates reduced Eurasian watermilfoil by 90%. Results demonstrated that integrating low doses of triclopyr with an indigenous pathogen, M. terrestris, can improve control of Eurasian watermilfoil. Lower use rates of triclopyr would minimize impacts to sensitive nontarget vegetation, reduce application costs, and may minimize impacts of label-imposed use restrictions.

Nine sites of cogongrass were included in a study of genotypic diversity and spread dynamics at the point of introduction and its adjacent areas in the southern United States. Clones evaluated with two primer pairs yielded a total of 137 amplified fragment length polymorphism (AFLP) loci of which 102 (74.4%) were polymorphic. Genetic diversity was measured as the percentage of polymorphic, Shannon's information index, Nei's gene diversity, and panmictic heterozygosity. Nei's gene diversity (HS) across all nine sites was estimated to be 0.11 and within site gene diversity ranged from 0.06 to 0.16. Bayesian estimate of gene diversity and Shannon's information index were higher (0.17 and 0.17, respectively). The samples from the point of introduction (Pi) had the lowest genetic diversity for all types of estimates. Within site variance accounted for 56% of the total variation and among site variance 44% (P < 0.05). Differentiation among sites was assessed using FST. The greatest difference was found between the Pi and the others. No relationship was found between genetic and geographic distances. Principal component analysis as well as cluster analysis separated individuals into three main clusters. The Pi formed a separate subcluster. Gene flow (Nm), inferred from Φ-statistics describing the genetic differentiation between pairs of sites ranged from 0.6 to 5.55. The lack of significant relationship between gene flow and geographic distance as well as genetic and geographic distances suggests that the invasion dynamics of cogongrass into the southern United States is primarily through anthropogenic activities and to the lesser extent through natural forces.

Chemical, cultural, and biological methods have been developed to control leafy spurge in a variety of environments. Aphthona spp. biological control agents have established throughout the northern Great Plains and Rocky Mountain region and successfully controlled leafy spurge in many areas, but notable exceptions include areas with sandy soils. Leafy spurge control can be improved by combining methods such as chemical, biological, or cultural treatments, compared to a single method used alone. The effects of Aphthona spp., imazapic herbicide, and interseeded native grass species alone or in combination for leafy spurge control were evaluated at two locations in southeastern North Dakota for 5 yr. Both the Sheyenne National Grassland (SNG) and Walcott, ND, study locations had greater than 80% sand soil. Leafy spurge stem density, canopy cover, and yield were reduced for 1 to 2 yr in all treatments that included imazapic, with no difference in control between single and combination treatments. Aphthona spp. and interseeded native grasses alone or combined did not reduce leafy spurge density or cover. Aphthona spp. population remained low throughout the study at both locations. Forb yield increased during the study at the SNG but not the Walcott location. Conversely, warm-season grass yield increased at Walcott but not at the SNG. Leafy spurge stem density declined from 92 to 50 stems/m2 in 5 yr at the SNG site. The decline could not be attributed to specific treatments applied in this study and may be due to self-limitation or soil pathogens.

Common reed is an invasive species that has overtaken wetland habitats in the eastern United States and can spread into roadsides, turf, and ornamental sites. The postemergence grass herbicides used in nursery crops and turf, clethodim, fenoxaprop, fluazifop, and sethoxydim, did not control common reed. Dithiopyr, MSMA, and quinclorac also did not control this weed. Glyphosate applied at 2.24 kg ai/ha (2.0 lb ai/ac) was more effective in preventing regrowth of common reed than glufosinate at 1.12 kg/ha (1.0 lb ai/ac). Mowing every 2, 4, or 8 wk controlled common reed 93, 81, and 69%, respectively, by the end of the growing season, but only reduced regrowth by approximately 55% the following May. Applying glyphosate at 2% v/v either 1 mo after a mowing or 2 wk prior to mowing reduced common reed regrowth the following May by approximately 90%. Applying glyphosate without mowing provided similar common reed control the following spring compared to glyphosate combined with a single mowing. Common reed regrew in all treated plots 1 yr after study initiation, indicating that control treatments must be repeated if common reed is to be eradicated from a site.

Perennial pepperweed is invasive throughout California. It thrives in a wide range of environments and is a common weed in floodplains, pastures, wetlands, and roadsides. In disturbed areas, perennial pepperweed rapidly forms monotypic stands with a thick litter layer. These infestations not only out-compete other vegetation, but prevent re-establishment of desirable species even after perennial pepperweed control. This experiment examined integrated management strategies with the goal of maximizing perennial pepperweed control and establishment of desirable native vegetation. The experiment was conducted at two sites in Lassen County, CA. Both sites were heavily infested with perennial pepperweed and lacked competing vegetation. The experimental design was a split-split-randomized block with four replications. Site preparation treatments included winter burning, summer and fall mowing, winter grazing, and fall disking. These treatments were designed to remove thatch to facilitate herbicide application and reseeding of desirable perennial grasses. Herbicide treatments included chlorsulfuron, 2,4-D, or glyphosate applied at the flower bud stage. Revegetation treatments included no seeding and no-till seeding of native perennial grasses. Most site preparation plus herbicide combinations reduced perennial pepperweed cover > 85% compared to the untreated control, although treatment efficacy was variable between sites and years. Burning, grazing, mowing, or disking in combination with herbicide treatment and no-till seeding was necessary for successful native perennial grass establishment. Burning or mowing with yearly 2,4-D applications for 3 yr gave the best combination of perennial pepperweed control and native grass establishment. Chlorsulfuron caused chlorosis and stunting to western wheatgrass, basin wildrye, and beardless wildrye at both sites when applied the spring before seeding. No treatment offered complete weed control, suggesting follow-up spot herbicide applications are needed for long-term perennial pepperweed suppression. These results provide several successful integrated strategies for control of perennial pepperweed and revegetation to a desired native perennial grass community.

The ability of arbuscular mycorrhizal fungi (AMF) to influence the performance of nonnative invasive plants in their introduced range has received increasing attention. The dependence of the invasive nonnative vine pale swallow-wort on AMF was studied in three greenhouse experiments. The aims of the present work were to (1) determine AMF colonization levels of field-collected pale swallow-wort plants and several co-occurring native and nonnative plant species, (2) evaluate the growth response of pale swallow-wort to different components of the soil microbial community from an infested site, and (3) determine the growth response of pale swallow-wort when grown with a nonlocal AMF species. AMF root colonization was greater in pale swallow-wort (85, 98, and 50% arbuscules, hyphae, and vesicles, respectively) than in leek (72, 80, and 25%), a species that has been frequently used as a predictor of AMF density in soil. Root colonization of pale swallow-wort in the field was also greater than root colonization of common milkweed, a native herbaceous species often co-occurring in the same habitats, as well as two other herbaceous species, Canada goldenrod and blueweed. Survival of pale swallow-wort plants was significantly greater in soil collected underneath dense monospecific stands of pale swallow-wort in a Henderson Harbor, NY, field site than in sterilized soil. After 12 wk, plants grown in sterilized soil had a 33% survival rate, whereas all plants grown in the unamended soil, with an intact microbial community, were alive. Moreover, plants grown in the unamended soil were 130% taller, had 50% more leaves, and had 83% greater total biomass compared with plants grown in sterile soil. Plants grown in soil containing a Glomus intraradices isolate collected in Troy, AL, were 50% shorter and had 15% lower total biomass than plants grown in the unamended New York field soil. These pale swallow-wort seedlings also had a high mycorrhizal dependency of 93%. Plants grown in a sterilized soil that was reamended with an AMF-free microbial wash had significantly lower belowground and total biomass than plants grown in the unamended soil with the resident AMF community. There was a trend of decreasing height and biomass for plants grown in sterile soil relative to the unamended controls treatment. Plants grown in sterilized soil had significantly (28%) greater total biomass than plants reamended with the AMF-free microbial wash. These findings suggest that AMF occurring in invaded habitats have beneficial effects on pale swallow-wort survival and growth.

Aminopyralid efficacy on Canada thistle and changes in species composition in both Canada thistle–infested and native plant communities were evaluated in Theodore Roosevelt National Park, ND. Aminopyralid at 120 g ae/ha was applied in September 2004 to one-half of 30 native or 30 Canada thistle–infested sites. Canada thistle density 10 mo after treatment (MAT) averaged 2 stems/m2 compared to 31 stems/m2 in the control and 22 MAT averaged 16 stems/m2 compared to 42 stems/m2. Although Canada thistle was reduced in treated sites, species richness was similar to the untreated control. Species evenness increased in treated subplots compared to the control as foliar cover of low- and high-seral forbs and high-seral monocot species increased. Species richness was reduced in native plant communities treated with aminopyralid from approximately 12 to 9 species but evenness was not affected. Plant species diversity was similar in Canada thistle–infested treated and nontreated subplots but was reduced by aminopyralid in native subplots. Despite the decrease in Canada thistle and increased foliar cover of native perennial plants, composition and structure 22 MAT in the Canada thistle–infested plant community did not resemble native plant communities. Native plant communities were composed of many perennial monocots and high-seral forbs, which had increased but were in lower relative abundance in the treated Canada thistle–infested subplots.

Prescribed grazing is an effective tool for controlling some rangeland weeds. Forage quality of eight nonnative rangeland weeds common to northern Idaho was determined. Five collection sites were established for each weed species: rush skeletonweed, meadow hawkweed, houndstongue, sulfur cinquefoil, yellow starthistle, Dalmatian toadflax, hoary cress, and tansy ragwort. Plants were collected at rosette, bolt, flower, and seed set stages; dried and separated into leaves, stems, flowers, or seed; and analyzed separately for crude protein (CP), neutral detergent fiber (NDF), and mineral content (ash). As expected, plants became more fibrous as they matured because stems increased in NDF value throughout the season and the leaf : stem ratio of most weeds we examined decreased throughout the season. In general, the weeds we examined expressed only moderate fiber levels, beginning the season with NDF values near 34% in the rosette stage and becoming near 52% NDF in the seed set stage. CP values were near 25% in the rosette stage for houndstongue, rush skeletonweed, and hoary cress whereas other weeds we examined had about 15% CP in the rosette. As the season advanced, all plants lost protein content and ended the season with CP values from 5 to 8% at seed set. Ash values declined for all species as the season progressed. Our results indicate that these weeds have forage values similar to many native plants and could be reasonable forage resources for livestock. We did not examine the content of secondary compounds in these weeds that may render them unpalatable and limit their forage value.

The importance of dispersal in the establishment and proliferation of exotic populations make this life history stage critical in the prediction and management of biological invasions. We observed the dispersal of seeds by patches of smooth brome invading northern fescue prairies and applied an inverse power function model to explore its potential invasion patterns. Based on our observations of two northern fescue prairies in Riding Mountain National Park, Manitoba, Canada, patterns of potential invasion were contingent upon the dispersal of seeds as individual florets or aggregated within spikelets and panicles. For example, although the majority of dispersed seeds were intercepted within one meter, inside and outside the margins of invading patches, slopes of the log–log plots of seed number against their dispersal distance were steeper for seeds dispersed as spikelets than individual florets. Despite the observed aggregation of seeds along the margins of invading patches, the number of dispersed seeds was poorly correlated with that germinated from the seed bank. The shallow dispersal gradient of individual florets and spikelets, combined with the steeper gradient of panicles suggest that smooth brome is capable of simultaneously invading along dense fronts and by establishing isolated foci. Although low correlations between the number of dispersed seeds and their recruitment from the seed bank might suggest postdispersal transport of seeds, other mechanisms, including seed predation and pathogens, remain unexplored. Conservation and restoration of northern fescue prairies must include efforts to control the dispersal of smooth brome seeds and reduce opportunities for their establishment.

The biological control agent Aphthona nigriscutis Foudras (Chrysomelidae) established in Fremont County, WY since 1992 on leafy spurge was released into a mixed stand a native plant Euphorbia robusta Engelm. During host range testing, E. robusta was a likely host for A. nigriscutis under laboratory conditions. In 1999, A. nigriscutis was observed feeding on both E. esula and 31 of 36 E. robusta plants present on about 2 ha (5 ac) where the visually estimated E. esula canopy cover was 50%. By August 2001, E. esula cover had declined to less than 5% and E. robusta plants had increased to 450 plants with 26 (5.8%) showing feeding damage. In 2006 Euphorbia esula ground cover was 2% and of 598 E. robusta plants originally marked, 391 could be located and four of these had damage consistent with A. nigriscutis feeding. For the 8-yr period, E. esula ground cover was inversely correlated to E. robusta density and positively correlated to A. nigriscutis feeding damage on E. robusta. This study shows that while also acceptable to A. nigriscutis in the field, feeding on E. robusta declined with declining densities of the target weed while E. robusta population densities increased. It seems that some risk in this regard is acceptable in light of the damage from the target weed and the generally high level of selectivity provided by biological control agents.

The invasive biotype of common reed has taken over wetlands and can spread into roadsides, turf, and ornamental sites. Additional control options are needed for managing this biotype in noncrop areas and other sites. In container trials, fosamine, glyphosate, and imazapyr all provided excellent control of common reed. Triclopyr suppressed common reed, with increasing rates improving the growth reduction. Chlorflurenol, primisulfuron, and sulfosulfuron did not suppress common reed growth. In field trials, glyphosate and imazapyr were more effective than fosamine, providing 82 and 93% control, respectively, the following April after either June or September applications. The optimum window of application for glyphosate may therefore include both summer and fall applications, wider than the commonly accepted window of only fall treatments. Fosamine provided greater suppression of common reed when applied in September (68% control) compared to June applications (43% control).

Field and garden experiments were performed to investigate if pathogens, generalist herbivores, or a specialist herbivore (the tortricid moth Pristerognatha fuligana) have the potential to control the invasive Himalayan balsam. Impacts of generalist herbivores and pathogenic fungi were excluded by using thiacloprid and trifloxystrobin. Results show no effect of any antagonist in the field experiment, though we found a modest impact of the combined influence of generalist herbivores and pathogenic fungi in the garden experiment, i.e., under suboptimal growing conditions. Limited information from the native area (the Himalayas) suggests a strong impact of specialist herbivores, thus confirming the enemy release hypothesis. We predict that impact studies of native specialized herbivores will clearly indicate antagonists from the native area with a strong impact on Himalayan balsam in its invaded area.