| Abstract |
- Purple loosestrife is a wetland perennial that came to the eastern seaboard of
North America in ship ballast and raw wool in the 1800's. Since coming to the
United States, it has spread across the country, forming dense monospecific stands
in wetlands. Mechanical and chemical control was expensive and ineffective, and a
biological control program was initiated in the 1980's. Four biological control
agents were released to control purple loosestrife: two leaf-feeding beetles
Galerucella calmariensis and G. pusilla (Coleoptera: Chrysomelidae), a rootmining
weevil Hylobius transversovittatus (Coleoptera: Curculionidae), and a
flower-feeding weevil Nanophyes marmoratus (Coleoptera: Curculionidae). The
leaf-feeding beetles are the most widespread.
Most biological control programs monitor the release and establishment of
control agents, but neglect to measure control agent spread, suppression of the
weed, and community succession. In this thesis I report on the suppression of
purple loosestrife, the spread of the beetles in relation to a spreading purple
loosestrife population, the potential to integrate biological control of purple
loosestrife with traditional management practices, and community succession
following purple loosestrife suppression in western Oregon.
In the first part of my study I monitored weed suppression and subsequent
community changes associated with. biological control at two sites in western Oregon. Two biological control agents, Galerucella calmariensis and G. pusilla
were released at Baskett Slough National Wildlife Refuge and Aquatic Gardens in
western Oregon in 1992 to control purple loosestrife. An intensive quantitative
monitoring program began in 1994. My objectives were to compare the outcome
of biological control at two sites, determine the stability and long-term pattern of
biological control, and to evaluate the ability of control agents to keep up with the
spread of purple loosestrife. I used estimates of Galerucella feeding damage,
purple loosestrife biomass, and purple loosestrife stem density to evaluate the
presence and density of beetles and suppression of purple loosestrife.
Beetle damage varied within and between years at both sites, but increased
faster at Baskett Slough than at Aquatic Gardens. Damage rates reached 100% at
least one year earlier at Baskett Slough. Initial purple loosestrife density was
higher at Aquatic Gardens (48 stems m2) than Baskett Slough (32 stems m2), but
final densities converged to less than 12 stems m2. At Baskett Slough, purple
loosestrife biomass varied over time. Galerucella beetle spread lagged 676 meters
behind the spread of purple loosestrife. As purple loosestrife declined, biomass of
forbs and grasses increased, and when purple loosestrife increased again,
abundance of forbs decreased.
To summarize the results of the first study, control developed one to two
years earlier at Aquatic Gardens than Baskett Slough over an eight-year study
period. Purple loosestrife biomass was suppressed to 10% of pre-control
abundance and seeds in the seedbank declined 55.2% over three years. Forbs and
did not vary over time, but reed canary grass was higher in 1998 than in the other
four years of the study.
In the second part of my study I evaluated how biological control of purple
loosestrife and the community responded to mowing, tilling, herbicide, and burning
to control reed canary grass using a randomized block experiment. Although there
have been several studies examining the effects of herbicides on biological control
agents, few studies have evaluated the potential to integrate biological control with
mowing, tilling, and burning. I evaluated the responses of our focal species and purple loosestrife - Galerucella interactions for two years after the application of
treatments by measuring the biomass of each species in the community and
Galerucella damage to purple loosestrife.
My experiment detected no effect of treatments on colonization of treatment
plots by purple loosestrife or biological control agents. Galerucella beetles
colonized all plots, and we detected no difference in insect damage levels among
treatments in either year. Tilling and herbicide treatments decreased reed canary
grass biomass the first year after treatment, increased purple loosestnfe biomass the
second year after treatment, and increased plant diversity both years. I detected no
differences in biomass or diversity measures in mow, burn or control treatments
either year.
To summarize the results of the second part of my study, tilling and
herbicide treatments reduced dominance of reed canary grass and increased purple
loosestrife biomass and plant diversity indices. The reduced dominance of reed
canary grass in tilling and herbicide treatments allowed purple loosestrife and other
plant species to successfully compete with reed canary grass and grow in treated
plots. The treatments did not interfere with the ability of biological control agents
to re-colonize and completely defoliate purple loosestrife in treated areas in this
small scale experiment.
The third part of my study used an observational study to evaluate plant
community changes accompanying biological control of purple loosestrife growing
along an elevation-moisture gradient on a lakeshore in western Oregon. Prior
literature documents the tendency of vegetation to form distinct bands along
lakeshores, and indicates little correspondence between the growing community
and the abundance and distribution of seeds in the seed-bank. Prior studies also
document the ability of insects to affect plant succession and zonation by feeding
selectively on preferred species. Weed management requires controlling specific
weeds as well as succession to control a weed and encourage desired vegetation. I
evaluated plant community changes following biological control of purple loosestrife along an elevation gradient over three years using estimates of biomass
of species along the gradient, buried seed density, and Galerucella feeding damage.
The dominant species in this study formed bands along the lake-shore, and
the buried seeds were more broadly distributed than the growing plants. This
agrees with prior literature showing that the actual range inhabited by growing
plants is narrower than the potential range indicated by the buried seeds. Reed
canary grass was found only at the high and dry end of the gradient. Purple
loosestrife was predominately found in the low and wet end of the gradient and was
uncommon in the reed canary grass zone, except in 1999, the only year that beetle
feeding did not cause 100% defoliation across the entire gradient. In 1999
defoliation levels increased linearly with elevation.
To summarize the third part of this study, zonation occurred in the dominant
species along the gradient. Purple loosestrife was uncommon in the reed canary
grass zone except in 1999, when Galerucella feeding damage was low, suggesting
that the beetles indirectly affect purple loosestrife - reed canary grass interactions.
Overall, biological control of purple loosestrife was successful at two sites in
western Oregon. The Galerucella beetles were able to control purple loosestrife
under a variety of conditions despite annual variation in beetle population. Further
studies are necessary to determine the direct effects of treatments to Galerucella
beetles, and to clarify the effects of Galerucella beetles on reed canary grass -
purple loosestrife interactions.
|
|---|