Increasing the effectiveness of reed canary-grass Phalaris arundinacea control by burning and glyphosate application in wet meadow restorations, North St.Paul and University of Minnesota Landscape Arboretum, Minnesota, USA

  • Published source details Reinhardt Adams C. & Galatowitsch S.M. (2006) Increasing the effectiveness of reed canary grass (Phalaris arundinacea L.) control in wet meadow restorations. Restoration Ecology, 14, 441-451


In wet meadow restorations in Midwest USA, an invasive grass, reed canary-grass Phalaris arundinacea, is usually controlled by spring burning and spring glyphosate application but relative effectiveness had not been assessed. A survey of managers who control it by these methods revealed that 95% reported only short-term control, with some reporting little control beyond a few weeks. An experiment was thus undertaken to evaluate effects of burning and herbicide application timings.

Study sites: Two study sites in Minnesota were selected for uniform topography, hydrology and P. arundinacea cover: an abandoned sod farm in North St.Paul (NSPSF); and an area within the University of Minnesota Landscape Arboretum in Chanhassen (60 km west of the NSPSF). Both had been drained several decades earlier, with P.arundinacea present for at least 25 years and cover of 75-100%. P. arundinacea seed bank density at NSPSF was nearly double that of the Arboretum (872 vs. 475 seeds/m² in the top 5 cm layer).

Experimental design: Twenty 12 × 12 m plots were established at each site. Around each plot, a 1 m strip was covered with a plastic liner as a buffer resulting in 10 x 10 m plots. Treatments were burning vs. no burn (control) and herbicide application (no herbicide; mid-May; late August; and late September), i.e. eight treatment combinations. In year 2, half of the plots at each site were randomly selected for repeat burning and herbicide treatment, the other half received only first round treatment, i.e. 16 treatment combinations (burn treatments) × 4 (herbicide treatments) × 2 (number of rounds).

Implementation of treatments: Herbicide applications were of Roundup at manufacturer's recommended dose (2% solution applied 187 L/ha). Initially, the two later season herbicide applications were timed to compare pre- and post-frost effectiveness. When no difference was found, applications in 2001 and 2002 were timed to include the period of rhizome carbohydrate accumulation (autumn) and stagnation.

After treatment, plots were seeded in early May with seed mixes of grasses (10 species, including prairie cord grass Spartina pectinata, blue-joint grass Calamagrostis canadensis and switch-grass Panicum virgatum, seeded at 967 seeds/m² - 15 kg/ha) and forbs (18 species, including thoroughwort Eupatorium perfoliatum, common sneezeweed Helenium autumnale and wild bergamot Monarda fistulosa, seeded at 165 seeds/m² - 5 kg/ha). After burning, grass seed was raked into the soil and forb seed spread onto the surface.

All NSPSF plots received burning and herbicide treatments during the 2000 growing season. In May 2001, half of the plots were seeded and monitored in 2001 and 2002. The other half received a second round of burning and herbicide treatments in 2001, were seeded in May 2002, and were monitored in 2002 and 2003. The same schedule was implemented a year later at the Arboretum.

Monitoring: Live shoot density was measured in April prior to treatments and in May post-burn at permanent sampling points. In late August each year, P. arundinacea and other aboveground biomass and cover were assessed. To invetigate seed bank response, Arboretum plots were sampled in October 2000 (before treatment), October 2001 (1 year after treatment) and October 2002. Three samples per plot were taken with a soil corer (7.5 cm diameter to 10 cm depth) and mixed. A 250 cm³ subsample was taken and spread in a plastic tray (19.5 × 19.5 × 6 cm) filled with soil. Samples were grown in a greenhouse for 6 months. Seedlings were counted and identified. P.arundinacea seed bank density was similar between plots.

Site differences in P.arundinacea response were not significant, therefore data were combined for analyses.

Response to burning: Burning initially increased P.arundinacea shoot density (4 weeks post-burn in unburned plots 520 shoots/m² and 1,180/m² in burned plots). However, 12 weeks post-burn, biomass was similar in burned and unburned plots (200-300 g/m²). Plots burned in two consecutive years had a P. arundinacea biomass similar to plots that received no-burning. Burning did not influence the effect of herbicide.

Response to herbicide: After one round, early spring P.arundinacea shoot density was reduced in mid-May herbicide plots to 141/m² (22% of control density); late August and late September herbicide plots, however, had virtually no shoots (2-3/m², <1% of control density). After two rounds, early spring P. arundinacea shoot density was further reduced (late August and late September applications = 0-1/m², mid-May applications = 26/m² = 3% of control levels). No herbicide treatment was significantly more effective than any other. After both one and two rounds of herbicide applications, growth in August and September herbicide plots was from seedlings, but in May plots was primarily from rhizomes.

One round of herbicide application in late August and late September was more effective (90% biomass reduction compared to control) than mid-May applications (75% reduction). Two rounds further reduced biomass. Two rounds of mid-May herbicide applications resulted in 49 g/m², equivalent to the reduction achieved by one round of application in late August or late September.

P.arundinacea seed bank: Burned plots had lower seed bank density than controls after 1 year (50 vs. 275 seeds/m²) and after 2 years of treatments (75 vs. 175 seeds/m²). Application of one round of herbicide treatments did not affect the seed bank, but plots after 2 years of herbicide treatments had a lower seed bank density (60-120 seeds/m²) than no herbicide plots (280 seeds/m²).

 Recolonization: Despite effective removal of P.arundinacea by one round of late-season herbicide application (August or September), it recolonized the year following seeding, with P.arundinacea biomass reaching 50% of that of the control, where less than 10% of the control a year earlier. It rapidly recolonized mid-May herbicide plots. One year after seeding, P.arundinacea biomass was still lower in herbicide plots than in the controls but was similar between herbicide timings. Burning did not influence P.arundinacea recolonization.

Colonization of planted and non-planted species was negatively correlated with presence of P.arundinacea. Colonization of species in the restoration seed mix was low and did not differ with burn treatments. Planted species rarely established in mid-May herbicide plots (93% failures) and failed completely in non-herbicide plots. Of the 26 species planted, 15 established, including the grasses Canada wild rye Elymus canadensis, swith-grass, blue-joint grass, and the forbs thoroughwort, common sneezeweed and swamp milkweed Asclepias incarnata.

Non-planted species had greater establishment success in herbicide plots whilst burning did not influence  establishment. Several unplanted wetland natives colonized, including several present on-site prior to treatment application (e.g. blue vervain Verbena hastata, stinging nettle, monkey flower Mimulus ringens) and also some persisting along ditches surrounding the sites (black-eyed susan Rudbeckia hirta, Solidago sp.), annual weeds also colonized.
a multiple season management strategy including later season herbicide applications over several growing seasons and subsequent selective Phalaris removal to limit reinvasion and facilitate native species establishment.

Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at:

Output references

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, terrestrial mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 18

Go to the CE Journal

Discover more on our blog

Our blog contains the latest news and updates from the Conservation Evidence team, the Conservation Evidence Journal, and our global partners in evidence-based conservation.

Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape Programme Red List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Bern wood Supporting Conservation Leaders National Biodiversity Network Sustainability Dashboard Frog Life The international journey of Conservation - Oryx British trust for ornithology Cool Farm Alliance UNEP AWFA Butterfly Conservation People trust for endangered species Vincet Wildlife Trust