Bunchgrass Ridge

Ecology and restoration of conifer-invaded meadows:
Research and adaptive management

Home > Key findings
Key findings and research highlights

Study area

Key findings
Cheryl Friesen
Cheryl Friesen
WNF Science Liaison
Hosted by
UW link
Privacy | Terms

Mountain meadows throughout the Pacific Northwest are undergoing rapid invasion by conifers. At Bunchgrass Ridge we are exploring the history, potential causes, and ecological consequences of these invasions, and testing the potential to restore naive meadows through tree removal and prescribed fire. Here we report the highlights and key findings of our work.

Click on the thumbnails below to see details of relevant research.

On this page:
1. Conifer invasion
2. Vegetation responses
3. Gopher disturbance
4. Restoration experiment
BG Conifer invasion link 1. Patterns of conifer invasion
  • Changes in climate and disturbance regimes. At Bunchgrass Ridge, conifer establishment during the 20th century was not correlated with variation in climate, but may have been triggered by a change in disturbance regime (grazing or fire).

  • Role of biotic interactions. Spatial and temporal patterns of recruitment of lodgepole pine and grand fir suggest that once establishment is initiated, strong positive interactions among trees can facilitate rapid conversion of meadow to forest—even when climatic conditions are not conducive to establishment.

  • Recommendation. Removing trees early in the invasion process can preempt these interactions and the ecological changes that accompany conversion of meadow to forest.

BG Vegetation responses link 2. Vegetation responses to tree encroachment
  • Replacement of meadow by forest species. Conifer establishment results in rapid replacement of meadow forbs and grasses by mesic forest herbs. Within 60-80 yr, understories may be dominated by forest species.

  • Tree species have differing effects. Lodgepole pine and grand fir differ markedly in their effects on the ground-layer community. Under grand fir, cover and diversity of forest herbs rise steeply with tree age (or size). Similar changes do not occur under pine. Instead its effects on the ground vegetation appear to be indirect, through its strong facilitation of grand fir.

  • Meadow species are largely absent from the soil seed bank. 75% of the species present in these meadows are absent from the soil seed bank. Seed banks are instead dominated by weedy, ruderal species. As a consequence, there is limited potential for seed banks to contribute to recovery of meadow species lost to conifer encroachment.

  • Seed dispersal and vegetative spread. Given their absence from the seed bank, species’ reestablishment during restoration hinges on natural seed dispersal or gradual vegetative spread from adjacent meadow openings.

  • Recommendation. Restoration will be most successful where it targets ecotonal (edge) environments or tree islands to maximize the potential for dispersal of meadow species.

BG Role of gophers link 3. Role of gophers in meadow community structure
  • Importance of gopher disturbance. Gopher disturbances (including mounds, tunnels, and castings) are common features of Cascade meadows and contribute in important ways to plant community structure.

    Mounds bury plants, reset succession, reduce dominance of competitively superior grasses and sedges, and increase variability of species composition at local and larger spatial scales.
    In the absence of gophers, meadows that currently support a diverse array of forbs and graminoids would become increasingly dominated by grasses.

BG Restoration experiment link 4. Potential for meadow restoration with tree removal and fire
  • Tree removal over snow. Felling and yarding over deep snow resulted in little disturbance to soils. Climate warming and reductions in snow pack could pose challenges to this approach in the future.

  • Operational tradeoffs of fuel-reduction methods. Broadcast and pile burning can be effective at reducing ground fuels, but involve tradeoffs. Creating slash piles can be labor-intensive, but piles can be burned in late fall or early winter when fire risk and containment costs are low. Broadcast burning does not require redistributing fuels, but weather conditions are more restrictive and costs of containment can be high.

  • Effects of burning on soils. In broadcast burned plots, fire consumed the majority of fine fuels within 2 hr. Burning resulted in some exposure of mineral soil, but most of the ground surface was covered by charred litter or duff. Effects on soil chemistry were limited to a transient (first-year) increase in available N.

    In pile-and-burn plots, slash piles burned to completion within 2 days (95-100% consumption). Burn scars covered <10% of the ground surface but only scar centers showed signs of severe burning. Effects on soil chemistry included small, but significant changes in soil carbon (loss) and pH (increase), and a dramatic increase in available N. N concentrations declined over time, but remained elevated in year 3.

  • Gophers reinvade meadows and dampen effects of burning. Soil disturbance by gophers increased significantly over time. Mixing of sub-surface and surface soils greatly dampened the effects of broadcast and pile burning.

  • Vegetation responses to tree removal. Short-term responses to tree removal suggest strong potential to reverse the effects of decades to well over a century of tree influence. Tree removal, with or without fire, benefited meadow species at the expense of forest herbs.

  • Necessity of fire? Fire does not appear critical to recovery of native meadow communities in the short term. For most measures of vegetation response, effects of burning were no different from those of tree removal alone.

  • Ruderal species. Ruderal species were uncommon in the post-treatment vegetation, despite their abundance in the soil seed bank. Exposure or heating of mineral soil may not have been sufficient to promote germination in burned plots, but heating beneath burn piles probably destroyed most viable seeds.

  • Re-invasion of conifers. Conifer establishment was sparse and no greater in burned than in unburned plots. Although of limited importance in the short-term, abundant seed sources and germination sites create the potential for future invasion.

  • Recovery from a broad range of initial states. Recovery of meadow species was clearly evident among the full range of initial vegetation states (young to old forest). However, restoration is far from complete and hinges on the ability of species to disperse from adjacent meadows and residual openings.