Help tooltips: On  |  Off

Trials of two annual cover crops for controlling reed canary-grass Phalaris arundinacea invasion during restoration of wet sedge meadow at Minnesota Landscape Arboretum, Minnesota, USA

Published source details

Perry L.G. & Galatowitsch S.M. (2003) A test of two annual cover crops for controlling Phalaris arundinacea invasion in restored sedge meadow wetlands. Restoration Ecology, 11, 297-307


Rapid establishment by aggressive plants such as reed canary-grass Phalaris arundinacea often hinders establishment of the desired sedge meadow plant community in attempted restorations of prairie pothole wetlands in the mid-USA. Sowing a cover crop during the establishment period might suppress unwanted species such as P.arundinacea. In this study, two potential cover crops, barnyard grass Echinochloa crusgalli and nodding smartweed Polygonum lapathifolium, were evaluated for their potential to suppress P.arundinacea. The effects of E.crusgalli and P. lapathifolium on sedge meadow establishment was also assessed.


Study site: A trial of two annual cover crops for controlling Phalaris arundinacea during the establishment phase of sedge meadow restorations was conducted at the Minnesota Landscape Arboretum (44°51' N, 93° 36' W), Carver County, USA. A former wetland, drained and cultivated since the early 1900s, was split into four sub-basins and reflooded in 1994.

Site preparation The experiments were performed in one 0.2 ha sub-basin with a 5% slope. To eliminate the existing seed bank, the soil in the sub-basin was sterilized in situ to 20 cm depth using Basamid (a granular fumigant). Wooden frames (20 cm deep) to discourage rhizomes of experimental plants from growing beyond the experimental plots, were embedded in the soil in 15 × 1 m strips perpendicular to the sub-basin slope. Each strip contained 60, 0.25 m² plots, arranged in 30 × 2 grids. For the 1997 to 1998 experiment, three strips were installed on each of two sides of the sub-basin (n = 360). For the 1998 to 1999 experiment, two strips were installed on a third side (n = 120). Strips were separated by 1 m walkways covered with landscape fabric. Plots were sprayed with 50 mL of a filtered slurry of non-sterile sub-basin soil to reintroduce soil microbes. A constant water depth was maintained by watering in dry periods and draining in wet periods as required.

Experimental design: Two similar experiments were performed, each over two growing seasons: the first from early June 1997 to September 1998; the second late April 1998 to September 1999.

Each of the two cover crops (barnyardgrass Echinochloa crusgalli and nodding smartweed Polygonum lapathifolium) was sown with P.arundinacea at four sowing ratios: 100%, 87.5%, 50%, and 12.5% cover crop. P.arundinacea was also sown in monoculture (0% cover crop). The native wetland porcupine sedge, Carex hystericina was sown at a constant density (500 seeds/m²) to assess the effects on its establishment and that of P. arundinacea and the cover crops.

For the 1997 to 1998 experiment, plots were grouped into three blocks according to elevation above water level. Initially, average elevations were 37 cm (high), 24 cm (mid), and 10 cm (low), but the water level was raised 15 cm in April 1998 to improve C.hystericina germination. The average elevation above water level for the 1998 to 1999 experiment was 8 cm.

Species mixes and monocultures were each sown at three densities (1,000, 2,500 and 5,000 seeds/m²) covering the range of seed bank densities measured in natural sedge meadows. Each of the 30 treatments, randomly assigned to blocks (2 cover crops × 4 sowing ratios × 3 total sowing densities + 3 sowing densities for P.arundinacea and C. hystericina monocultures) was replicated four times within each of the three blocks in the 1997 to 1998 experiment and twice within each of the two blocks in the 1998 to 1999 experiment.

Seed sources and sowing: E.crusgalli and P.lapathifolium seeds were collected from wetlands within 50 km of the study site. C.hystericina and P.arundinacea ("common" variety) seeds were purchased locally. E.crusgalli and P.arundinacea. Seeds were appropriately stored and treated to break dormancy. Seeds were sown by mixing in 1 L of water, drizzled onto and spread over the soil by hand raking.

Site maintenance: Weeds were removed monthly from June to September each year. To minimize seed loss in run-off, plots were covered with white plastic sheets each winter (November to March).

Measurements: Aboveground biomass was harvested in the first week of September of each growing season from two replicates in each block (1997 to 1998 experiment), or one replicate in each block (1998 to 1999 experiment). To avoid edge effects only plants within the central 25 × 25 cm of each plot were harvested. Vegetation was clipped at the soil surface, sorted by species, dried and weighed. In September of the second growing season of each experiment, dead biomass on the soil surface of the central 25 x 25 cm of each plot also was collected, dried and weighed. Belowground biomass (within 25 × 25 × 15 cm deep blocks of soil) was harvested from one replicate in each block in September of the second growing season of the 1998 to 1999 experiment.

Percent light interception in each plot was measured in August of the first growing season of each experiment.


Echinochloa crusgalli, compared with no cover crop, reduced P.arundinacea biomass by 65% to 98% (dependent on sowing rate and elevation) after two growing seasons, but P.lapathifolium had no affect. Dense E.crusgalli growth in the first year and thick E.crusgalli thatch in the second, substantially reduced light availability for P.arundinacea. E.crusgalli also reduced porcupine sedge biomass by more than 1.8 kg/m² (99%) after two growing seasons. Sedge biomass was similar in plots sown with E.crusgalli to P.arundinacea monocultures.

Conclusions: An effective cover crop in prairie pothole wetland restorations must both prevent P.arundinacea invasion and allow typical sedge meadow perennials such as C.hystericina to establish. Neither E.crusgalli nor P.lapathifolium meet these criteria. Whilst E.crusgall was shown able to suppress P.arundinacea invasion for at least two growing seasons by establishing dense stands and creating thick persistent thatch layers in the first growing season, it also interfers with sedge C.hystericina establishment and is therefore unlikely to improve sedge meadow restoration success. P.lapathifolium did not influence P.arundinacea invasion. Until alternative methods are found, thorough site preparation to remove P.arundinacea propagule sources before restoration is considered essential.

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