Effects of prescribed burning on herbaceous and woody vegetation in northern lowland meadows

Summary

The extent of meadows in the Lower Peace and Slave River drainages in the Northwest Territories, Canada, has been reduced by shrub and tree encroachment. Spring burning of sedge-grass meadows between 1992 and 1998 was undertaken to reduce shrub encroachment in an attempt to enhance bison Bison bison grazing habitat, although the impact of fire on preferred bison forage was unknown.

In the summer of 1998 a study in the Hook Lake area of the Slave River was undertaken to test the effect of burn frequency on herbaceous plant community composition and willow Salix spp. shrub survival and vigour.

Study area: The open meadows of the Slave River Lowlands (SRL) are characterized by flat land with many sloughs and old stream channels. A history of periodic flooding along the Slave River has resulted in accumulation of deep layers of alluvium overlying glacial tills. The meadows, surrounded by boreal forest, provide habitat to one of the few extant wild bison Bison bison populations in northern Canada.

Meadows in the SRL can be divided into two broad categories: Wet meadows dominated by sedges (awned sedge Carex atherodes, beaked sedge C.rostrata, water sedge C.aquatilis) and grasses (reed grass Calamagrostis spp., sprangle top Scholochloa festucacea, sweet-grass Glyceria spp.); dry meadows characterized by grasses (Northern reed grass Calamagrostis inexpansa), rushes Baltic rush (Juncus balticus) and forbs (Virginia strawberry Fragaria virginiana, Potentilla spp., American vetch Vicia americana and Canada goldenrod Solidago canadensis). Vegetation within the larger meadows in the Hook Lake area usually includes patches of both meadow types.

Study design: In the spring of 1992, 1993 and 1995 meadows were burned in the eastern SRL in the Hook Lake Bison Range. Burn treatments were carried out after snow melt but before leaf-break. The fires attempted to meet desired burning conditions, including moderately light wind, high fine fuel moisture code, low duff moisture code, moderate to intense heat, continuous flame front, and dormant vegetation or early bud flush. There were overall, lower intensity fires in 1993 compared with 1992.

In July 1998, six meadows were selected to evaluate the effects of spring burning on plant communities. The meadows were either unburned, burned once or three times between 1992 and 1998. Two meadows had been burned in 1992, 1993 and 1995; two burned only in 1995; and two had not burned since at least 1992 (prescribed fire was not used in this area prior to 1992).

The six meadows varied in size (from approximately 400 to 4,800 ha), shape and relative distance to one another. Five, 1,000 m² plots (20 × 50 m) were randomly established in July 1998 in dry meadow areas on each of the six meadows. If the randomly selected site did not meet the following criteria i.e. predominantly dry meadow vegetation, some shrub cover (20–40%), evidence of burning (for burned meadows), and little obvious disturbance by bison or humans (e.g. hoof prints, bison wallows, vehicle tracks), then the next closest site, which did conform, was chosen.

Ten, 1 m² quadrats were placed in each 1,000 m² plot, in which herbaceous plant species abundance was estimated using percentage cover categories. Soil pH was measured at depths of 10 to 15 cm. Organic horizon (OH) thickness was measured in the OH horizon of the soil profile in a small hole dug within each quadrat. Plant litter was removed from a 25 × 25 cm area in each quadrat, dried at 70°C for 72 hr and weighed.

Within each 1,000 m² plot all individual willow shrubs over 0.5 m tall were identified to species. The shrub canopy was measured (height and width at the widest point), and each shrub was assigned one of four 'vigor' classes: I = standing dead shrub with no new growth, II = new growth from the root crown, III = new growth to about 75% of the height of standing dead stems, and IV = robust leaf growth to full height of shrub.

Forty-eight herbaceous plant species were recorded in the quadrats and eight plant community groups identified (through cluster analysis). Forty-six percent of the 300 quadrats were classified in two groups: Calamagrostis spp.; and Calamagrostis spp., Fernald's hay sedge Carex aenea and Juncus balticus. Many of the communities shared dominant species, but the proportion of the dominant species distinguished them.

Litter biomass: The average litter biomass decreased from 50.8 g (±18.0 SD) on unburned meadows, to 40.0 g (±18.2 SD) on meadows burned once, and 30.5 g (±17.6 SD) on meadows burned three times. Average values of soil pH were 5.9 on unburned meadows, 5.8 on meadows burned once, and 6.0 on meadows burned three times.

Analysis indicates that Calamagrostis spp., an important bison forage taxon, had a low association with either the burn regime or litter biomass.Carex atherodes, also an important forage species, shows a low association to burn regime but has a relatively strong association with high litter biomass. Conversely, J.balticus and C.aenea, both poorer bison forage than Calamagrostis spp. and C.atherodes, appear to be strongly associated with meadows burned three times.

Two plant communities in particular appeared to be strongly associated with meadows burned three times: that characterized by Calamagrostis spp., C.atherodes, J. balticus) and that characterized by C.aenea and Calamagrostis spp. A third community (slender wheatgrass Agropyron trachycaulum, yarrow Achillea millefolium) also had a tendency towards occurring in these burnt meadows. There was a significant difference in shrub occurrence within the four vigor classes and this differed among the three burn regimes. Within meadows burned three times, 24% of the shrubs sampled belonged to class I (standing dead) compared with 12% in the single-burn and 0.3% on unburned meadows.

The relative proportion of willow in each of the vigour classes differed most between unburned meadows and burned meadows (once or three times). In the three times burned meadows there was more willow in class I (i.e. more standing dead) and fewer in class IV (i.e. fewer robust). On unburned meadows most willow appeared vigorous compared with fewer than a third on meadows burned three times.

Conclusions: In the Hook Lake area, the prescribed spring burns whilst reducing the vigour of the various shrub willow species, their survival remained high (76%) even on the most frequently burned meadows. Analysis suggest that multiple spring burns influenced plant community composition in dry meadow areas and that less palatable bison forage species (e.g. Carex aenea and Juncus balticus ) were correlated with a regime of three spring burns. These results suggest that frequent spring fires have only a small negative effect on willow cover (reducing shrub succession was a desired outcome) but may reduce the abundance of primary bison forage plants compared with less frequently burned meadows.


Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at: http://www.blackwell-synergy.com/journal.asp?ref=1061-2971