Canada: Manitoba: Tallgrass Prairie Restoration on Agricultural Fields Removed from Cultivation

Project Stage:
Implementation

Start Date:
1987-08-21

End Date:
2008-08-21

Primary Causes of Degradation

Degradation Description

Until 1986, approximately 225 ha of parkland were leased for agriculture.

Reference Ecosystem Description

Beaudry Provincial Park was created in 1975, and now comprises 950 ha of riparian forest and upland habitat. Lowland forest adjacent to the river is dominated by Populus deltoides (plains cottonwood), Acer negundo (Manitoba maple), Ulmus americana (American elm), and Tilia americana (basswood). Remnant tallgrass prairie and savannah in the park are characterized by Andropogon gerardii (big bluestem), Spartina pectinata (prairie cord grass), Dalea purpurea (purple prairie clover), Symphoricarpos occidentalis (western snowberry), and Quercus macrocarpa (bur oak).

Project Goals

To restore tallgrass prairie on agricultural fields removed from cultivation in Beaudry Provincial Park.

Monitoring

The project does not have a monitoring plan.

Description of Project Activities:
From 1987 to 2000, sections of three agricultural fields were sequentially removed from cultivation along a north-south gradient, tilled, summer fallowed, and seeded with prairie species. Seed was collected from local prairie sites within a 180 km radius of Beaudry Park using a manual harvester. This was supplemented by commercial seed (Prairie Habitats, Argyle, Manitoba) and seed grown at the park. Mixes had high proportions (85%) of grass seed by weight (Berg, 1992; Morgan, pers. com.). Reseeding of sites into standing vegetative cover has been infrequent, occurring twice at site 98 and once at site 99, where successful establishment of vegetation from a previous seeding was compromised by flooding (Friesen, pers. com.). These sites are also mowed repeatedly in the summer to control exotic species, including Cirsium arvense (Canada thistle) and Sonchus arvensis (perennial sow thistle). Commonly, patches within all restoration sites are mowed to control invasive exotics Melilotus alba (white sweet clover) and Melilotus officinalis (yellow sweet clover). Prescribed burning is conducted every 5 years and always in early spring.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
Three sites were randomly selected for analysis in each of the following classes: (i) those seeded between 1987 and 1990; (ii) those seeded between 1991 and 1994; and (iii) those seeded between 1995 and 1999. In addition, three unbroken remnant prairie sites were identified and analyzed as target reference points (i.e., reference sites). Species diversity differed substantially between restored and reference sites. Overall effective species richness was significantly higher in reference sites than restored sites, as were native species richness and effective species richness (ESR). In contrast, exotic species richness and ESR were significantly higher in restored sites. Species diversity was also affected by time since restoration. Overall species richness and ESR were highest in reference sites and were significantly lower in old and new restored sites. Interestingly, mid-aged restoration sites had significantly higher overall species richness than other restored sites. Differences in vegetation response became more apparent when species origin was examined. Native species richness and ESR differed significantly across all four age classes, being highest in reference sites, lower in mid-age and old restorations, and lowest in young restorations. In contrast, exotic richness and ESR predictably were highest in new and midage restorations and significantly lower in reference sites. Native evenness was lowest in reference sites and highest in old restorations, whereas exotic evenness was significantly higher in reference sites than in all restorations. Species composition differed substantially between reference and restored sites, and also among restored sites, and further separated sites into distinct clusters. Reference sites were characterized by native Bouteloua curtipendula (side-oats grama), Amorpha canescens (lead plant), S. pectinata, Galium boreale (northern bedstraw) and D. purpurea. The future restoration site was characterized by exotic ruderals Brassica kaber (wild mustard) and Malva rotundifolia(round-leaved mallow). Restored sites were characterized by natives Schizachyrium scoparium (little bluestem), Anemone canadensis (Canada anemone), and Potentilla arguta (tall cinquefoil), as well as exotics Thlapsi arvense (stinkweed), Phleum pratense (timothy), Amaranthus retroflexus (redroot pigweed), and Pastinaca sativa (wild parsnip). Young restoration sites were characterized by exotics Echinochloa crusgalli (barnyard grass), Chenopodium album (lamb_squarters), and Sonchus asper (annual sow thistle), and early successional natives Elymus canadensis (Canada wild rye), S. scoparium, and Ratibida columnifera (prairie coneflower). Old and mid-age restorations were characterized by exotics C. arvense and Alyssum alyssoides (alyssum), and prairie natives Agastache foeniculum (giant blue hyssop), and Rudbeckia hirta (black-eyed susan). Twenty-four species were significantly affected by time since restoration. Of these, 11 (45.8%) were seeded native species, four (16.7%) were non-seeded native species and nine (37.5%) were exotic species. Ten native species were primarily associated with reference sites, including prairie natives Artemesia ludoviciana (white sage), G. boreale, Solidago canadensis (Canada goldenrod), S. pectinata and woody species G. lepidota, Rosa spp. and S. occidentalis. Five exotic species were primarily associated with young restoration sites: C. arvense, M. alba, Polygonum convolvulus (black bindweed), S. arvensis and Taraxacum officinale (dandelion). Species that dominated older restorations included natives A. gerardii and Solidago rigida (stiff-leaved goldenrod), as well as exotics Bromus inermis (smooth brome) and Medicago sativa (alfalfa). Importantly, 10 seeded native species were not significantly affected by time and remained present in at least two age classes. These species seem to have been successfully established and included: Achillea millefolium (yarrow), A. canadensis, E. canadensis, Erigeron lonchophyllus (Fleabane), H. maximiliani, and Monarda fistulosa (wild bergamot) (data not shown). Two non-seeded native species, Aster ciliolatus (Lindley_s aster) and Convolvulus sepium (hedge bindweed), also had become successfully established. Similarity between restoration and reference sites generally increased with time. Overall similarity increased in the first 7 years since restoration initiation, but did not change thereafter. When species origin and growth form were considered, similarity of native seeded graminoids and native species that had been seeded in some restoration sites and self-recruited in others increased linearly with time-since-seeding. Similarity of native seeded forbs increased for the first 7 years, but decreased thereafter. Exotic species showed no significant changes in similarity. To determine the effects of restoration on soil properties, reference, restoration, and future restoration sites were compared. Although affected by restoration, changes in these variables were much smaller than those of vegetation. Soil phosphorus was significantly higher in restoration and future restoration sites than in reference sites. Soil sulphate was also significantly higher in the future restoration site than in restoration and reference sites, and soil nitrate also tended to be higher. In contrast, organic matter was significantly higher in the reference sites than in other sites. These differences in soil composition are likely a result of past agricultural activities.

Factors limiting recovery of the ecosystem:
The high diversity of mid-age (8 - 11 years) restorations relative to both new (3 - 7 years) and old (12 - 15 years) restoration sites suggests that desirable species had not become established in new sites, whereas old restorations appear to have lost species. Similar results have been observed in other tallgrass prairie restorations (e.g., Kindscher and Tieszen, 1998; Baer et al., 2002; Sluis, 2002). At Beaudry, increases in similarity between restoration and reference sites only occurred for the first 7 years. Likewise, 41% of seeded species were observed in mid-age restorations, whereas only 26% remained in old restorations. These changes suggest that simple "˜"˜sow-and-stop'' revegetation will not likely recreate high-diversity natural habitat and that these sites will require ongoing maintenance if they are to continue recovering.

Socio-Economic & Community Outcomes Achieved

Key Lessons Learned

Beaudry is one of the largest and longest standing tallgrass prairie restorations in Canada, and represents an important opportunity to examine the long-term impacts of prairie restoration on ex-arable land. Our results suggest that nearly 20 years of restoration have had desirable impacts on diversity and species composition at Beaudry, but this habitat will require ongoing restoration.

Long-Term Management

Analyses of the Beaudry Park project show that many of the older sites are losing species and thus, that this restoration should be understood as an ongoing process that requires a long-term commitment. Many of the species in decline tended to be forbs, which may reflect the use of grass-dominated seed mixes. However, the use of such mixes is common in tallgrass prairie restorations, as these are generally cheaper and more accessible (Sveinson, 2003). The resemblance of mid-age restorations and reference sites in our study suggests such mixes may successfully be used in prairie restoration. However, increasing dominance of grasses over time should be addressed by modification of management practices subsequent to establishment. Frequently, limited resources and convenience lead to spatially and temporally uniform management practices and over-reliance on a few routine maintenance activities, regardless of age, location, or soil status of restoration sites. This may facilitate the expansion of competitive native and exotic grasses and decrease the importance of less common species (Sveinson, 2003). Reduction of the scale of maintenance activities may aid in increasing restoration heterogeneity.

Colonization might be enhanced by small-scale disturbance, including patch mowing or spraying of herbicides on dominant C4 grasses, minimally disturbing soils, and interseeding of native forbs. Use of vegetative propagules, such as A. gerardii rhizomes collected from mature sites, could also facilitate establishment in new restorations while reducing grass dominance in older sites. Vegetative propagules have been successfully used in marsh restoration, yielding over 90% survival and extensive vegetative spread after four growing seasons (Fraser and Kindscher, 2001). These and other small-scale post-seeding activities will likely increase structural complexity and native forb diversity, and thus warrant further exploration (e.g., Sveinson, 2003).

Sources and Amounts of Funding

Funding for this project has been provided by the government of Manitoba.

Primary Contact

Organizational Contact