Download this weed on revegetation as a PDF

When land managers are faced with invasive forbs listed as noxious weeds, they have viable options for controlling them such as herbicide or biological control. While these tools can effectively kill weeds, they often do not result in a plant community that meets management objectives.  Remnant desired vegetation is often too scarce to fill voids previously occupied by an invasive forb, and the invasive forb may be replaced by an unwelcome secondary invader like the annual grass cheatgrass.  Integrating seeding of desired species with weed control may help to prevent secondary invasion and improve long-term restoration outcomes.

We conducted a study in which the goal was to restore desired grasses to rangeland dominated by spotted knapweed and cheatgrass through the integration of herbicides and revegetation. We expected that seeded species establishment would increase with control of spotted knapweed and cheatgrass.  We also expected that increased establishment of seeded species would result in lower abundance of spotted knapweed and cheatgrass over time. We tested eight herbicide treatments and six seeding treatments at a site near Missoula, MT.  Herbicide treatments were designed to provide control of spotted knapweed, cheatgrass, or both and applied August 2009.  Five perennial grasses were seeded in monoculture in November 2009.  Plots were sampled 1, 2, and 4 years following treatment application. 

graph showing effect of herbicide on bluebunch wheatgrass density 4 years after treatment

Effect of herbicide on bluebunch wheatgrass density 4 years after treatment.  See link to full article below for details.

In year 1 and 2, herbicide treatments performed as expected; spotted knapweed cover was reduced to less than 1% in all treatments that included aminopyralid, clopyralid, or picloram.  Cheatgrass cover was reduced to less than 0.1% by treatments that included imazapic.  In spite of good weed control, very few seedlings of seeded species were found.  However, by year 4 we found bluebunch wheatgrass and tall wheatgrass growing where it had been seeded.  Controlling weedy species with herbicides led to an increase in bluebunch wheatgrass establishment (see figure, right) but had no effect on establishment of tall wheatgrass.  We were surprised that bluebunch wheatgrass and tall wheatgrass performed so well 4 years after treatment in spite of finding very few seedlings 1 and 2 years after treatment. 

Establishment of seeded species resulted in lower weed abundance over time for spotted knapweed but not for cheatgrass.  Spotted knapweed was nearly absent when bluebunch or tall wheatgrass was seeded in combination with certain herbicide treatments.  Cheatgrass, on the other hand, did not seem to be affected by seeded grasses.

Although limited, this study still provides two valuable insights into revegetation of weed-infested rangeland. First, restoration of weed-infested rangeland is a slow process; it took 4 years before we saw the grasses we had seeded!  Seeded species need time to establish and subsequently compete with weedy species. Second, good weed control does not always equate to improved establishment of seeded species. Seedling establishment is influenced by many factors, only one of which is competition with weedy species. Once established, however, seeded desired species can limit re-invasion.  In our study, this was true for the large-statured exotic invasive forb spotted knapweed more so than the annual grass cheatgrass, whose abundance fluctuated greatly from year to year.   

Read more about this study here