Smith, L. 2004. Impact of biological control agents on diffuse knapweed
in central Montana, pp. 589-593. In:Cullen, J.M., Briese, D.T., Kriticos,
D.J., Lonsdale, W.M., Morin, L. and Scott, J.K. (eds), Proceedings of the
XI International Symposium on Biological Control of Weeds. CSIRO
Entomology, Canberra, Australia.
Title:
Impact of biological control agents on Centaurea diffusa (diffuse knapweed) in central
Montana
Author:
Lincoln Smith, Research Entomologist, USDA-ARS Western Regional Research Center,
800 Buchanan Street, Albany, CA 94710. lsmith@pw.usda.gov
Summary
Centaurea diffusa (diffuse knapweed) is a major weed in semiarid regions in the northwestern
United States. Investigations on biological control began in the 1960s and have resulted in the
release of 13 species of insect biological control agents (most of which also attack spotted
knapweed: C. stoebe subsp. micranthos [often reported as C. maculosa]). In Central
Montana, three capitulum-feeding insects (Urophora affinis, U. quadrifasciata and Larinus
minutus) and three root-feeding insects (Sphenoptera jugoslavica, Agapeta zoegana and
Cyphocleonus achates) have become well established. Populations of diffuse knapweed have
rapidly declined at study sites at two locations in the presence of high densities of biological
control agents. Larinus minutus and the Urophora species infested up to 62% and 59% of
capitula, respectively. Cyphocleonus achates, A. zoegana and S. jugoslavica infested up to
64%, 57% and 31% of roots, respectively. By the summer of 2000, some study sites had no
mature plants that could be sampled. Impacts of these insect populations on seed production
and plant survivorship are discussed.
Running headline: Control of Centaurea diffusa in Montana
Keywords: biological control, Centaurea diffusa, diffuse knapweed, rangeland, success
1
Introduction.
Centaurea diffusa Lam. (Asteraceae), diffuse knapweed, is an important invasive weed in
semiarid regions of the northwestern continental United States and southwestern Canada
(Harris and Cranston 1979; Maddox 1979; Sheley et al. 1998; Roché and Roché 1999). The
plant presumably originated from Eurasia, and the first North American specimens were
discovered in 1907 in alfalfa fields in Washington State (Howell 1959). Since then, the plant
has spread exponentially (Fig. 1), and infested 1.4 million ha by 2000 (Duncan 20001). It is
an important weed in the states of Colorado, Idaho, Montana, Oregon, and Washington, and it
is designated as noxious in 13 states and four Canadian provinces (Rice 2000).
This plant, and its close relative Centaurea stoebe L. subsp. micranthos (Gugler) Hayek
(spotted knapweed; often reported in the literature as C. maculosa Lam. or C. biebersteinii
DC. [Ochsmann 2001]) have been the target of biological control for over 40 years (Piper and
Rosenthal 1995). Thirteen species of insect biological control agents have been introduced
(Table 1; Müller-Schärer and Schroeder 1993, Rees et al. 1996, Story and Piper 2001). These
species were multiplied and distributed by USDA-ARS, APHIS and Forest Service; USDIBLM; and by state departments of agriculture, university personnel, and county agents. All of
these species have established to one extent or another, and about half of them have become
abundant in at least some regions (Story and Piper 2001, E. M. Coombs personal
communication).
Diffuse knapweed populations recently appear to be declining at many sites in Colorado,
Montana, Oregon and Washington (Seastedt et al. 2003; personal observation; G. L. Piper, E.
M. Coombs and R. F. Lang, personal communication). However, because of limited
resources and general emphasis on releasing and distributing agents, rather than on
investigation, we lack quantitative documentation of the recent impact of these agents. I
happened to arrive in Montana just as the insect populations were beginning to impact
C. diffusa populations and here report the partial results of two years observations at two
locations that were previously heavily infested by this weed.
Methods
Studies were conducted at two locations in Fergus County, Montana where agents had
previously been released. Both are in habitat currently (or historically) dominated by
ponderosa pine (Pinus ponderosa P. & C. Lawson) at about 1,000 m elevation with annual
precipitation of about 35 cm. The Eickhoff site is a grass meadow used for cattle grazing, but
was enclosed by a fence in 1990 to protect it as a biological control release site. The Shannon
2
site was historically excavated for gravel. It is gradually reverting to rangeland and deciduous
forest and is grazed by cattle. Releases of Agapeta zoegana, Cyphocleonus achates, Larinus
minutus, L. obtusus, Pterolonche inspersa, Sphenoptera jugoslavica, Terellia virens,
Urophora affinis, and U. quadrifasciata were made between 1990 and 1997 at the Eickhoff
site (see also Table 1). At the Shannon site, the same species were released over the same
period, with the addition of Bangasternus fausti, and omission of Urophora spp., L. obtusus,
and T. virens. Releases generally comprised 50 to 300 insects except that about 10,000
Urophora spp. were released in 1990. By 1995, seed head weevils, primarily L. minutus,
were being collected at the Shannon site for redistribution.
Permanent transects were established at the two locations and permanent positions for
Daubenmire frames (20 cm x 50 cm) were marked along the transects at 5 m intervals at
Eickhoff and at 1 m intervals at Shannon, where tree clumps interfered with long continuous
transects. In the second year (1999), data were collected from additional nearby transects.
Number of mature plants, rosettes, and bolts (stems with flowers) were recorded in August
1998 and 1999. To estimate insect attack rates, seed heads were collected in Oct. 1998 and
roots in June 1999 from haphazardly chosen plants at uniformly spaced intervals adjacent to
the transects. Seed heads were held in a refrigerator and dissected during the winter to
determine insect infestation, and roots were dissected immediately in the field. Insect
identifications were based on morphology of immature stages except for L. minutus, which
was sometimes based on exit hole and characteristic flower head damage.
Results and Discussion
Seed heads were heavily infested at both sites: 99% at Shannon and 59% at Eickhoff
(Table 2). The most abundant seed head insects were L. minutus at Shannon and Urophora
spp. at Eickhoff (primarily U. affinis, which tends to displace U. quadrifasciata under
competition [Berube 1980]). A large proportion of roots were damaged at the two sites: 74%
at Shannon and 69% at Eickhoff. Some roots were infested by more than one insect and
sometimes by more than one species. Cyphocleonus achates was the most abundant root
insect at both sites. This weevil appeared to directly kill some plants at the time they began to
bolt because the mature larvae had girdled the vascular tissue from the inside.
The knapweed population drastically decreased during the course of this study (Table 3).
At Eickhoff, the knapweed population had already decreased substantially below historical
levels, and grasses had become a dominant component of the plant community. Grasses
composed a much smaller proportion of the canopy at the Shannon site, which is very
3
gravelly, yet the knapweed population decreased to levels similar to those at Eickhoff. Grass
species at both sites included Agropyron smithii (western wheatgrass), A. spicatum
(bluebunch wheatgrass), Poa pratensis (Kentucky bluegrass), Stipa comata (needle-andthread), S. viridula (green needlegrass) (McGregor et al. 1986). The Eickhoff site also had
Festuca idahoensis (Idaho fescue), Bouteloua gracilis (blue grama), and Deschampsia
cespitosa (tufted hairgrass). There were no nearby sites that were not infested with insects for
comparison, so these data do not prove that the insects caused this reduction. However,
C. stoebe populations being studied in the same region showed no decline during this period
(unpublished data).
Centaurea diffusa populations recently appear to be declining at many sites in Colorado,
Montana, Oregon and Washington, where insect biological control agents are abundant
(Seastedt et al. 2003; personal observation; G. L. Piper, E. M. Coombs and R. F. Lang,
personal communication). This decline has been attributed primarily to the impact caused by
high densities of the two Urophora flies, L. minutus and S. jugoslavica. In earlier studies in
British Columbia, the Urophora flies greatly reduced seed production, but generally not
enough to provide adequate control (Cloutier and Watson 1989; Myers et al. 1989). Although
S. jugoslavica is widespread in British Columbia, Oregon and Washington, it appears to be
unable to control the weed by itself (Powell and Myers 1988; Powell 1989). The impact of
L. minutus adults feeding on rosettes in the spring is suspected to be a deciding factor in the
decline of C. diffusa in some areas (G. L. Piper and R. F. Lang, personal communication),
although this has not been proven. Seastedt et al. (2003) found that in central Colorado within
4 years of the initial releases C. achates infested about as many C. diffusa plants as
S. jugoslavica, despite releasing seven times more of the latter species. Cyphocleonus achates
was originally expected to attack primarily C. stoebe (Stinson et al. 1994); however, it is
clearly an important agent for C. diffusa, at least in habitats that have a sufficiently warm
summer to allow the insect to emerge early enough for it to exploit its long oviposition period.
Centaurea diffusa, which is usually a monocarpic biennial, generally has a much smaller root
diameter than C. stoebe, which is a perennial (Story et al. 2001), so the weevil is likely to
cause much more impact on this plant than on C. stoebe.
Unfortunately, the apparently widespread success of biological control agents to control
C. diffusa lacks rigorous documentation of the declines in weed density and hard data showing
the direct impact of the agents on the weed population. Further study of these agents would
be useful to gain a better understanding of why some became abundant while others did not,
4
and which ones contributed to the decline of the weed population. Many of these agents also
attack C. stoebe, although this weed still appears to be uncontrolled in this region.
Acknowledgements
Many thanks to M. Mayer and S. Rosenthal (USDA-ARS), V. Roberts and C. Clark (USDIBLM), and others who established the release sites and released the agents. Permission to
conduct these studies on private land was generously provided by W. Eickhoff and J.
Shannon.
References
Berube, D.E. (1980) Interspecific competition between Urophora affinis and U.
quadrifasciata (Diptera:Tephritidae) for ovipositional sites on diffuse knapweed
(Centaurea diffusa: Compositae). Journal of Applied Entomology, 90, 299-306.
Cloutier, D. & Watson, A.K. (1989) Application of modelling to biological weed control.
Proceedings of the VII International Symposium on Biological Control of Weeds (ed
E.S. Delfosse), 6-11 March 1988, Rome, Italy, pp. 5-16. Istituto Sperimentale per la
Patologia Vegetale, Ministero dell'Agricoltura e delle Foreste, Rome, Italy.
Duncan, C.L. (2001) Knapweed management: Another decade of change. Proceedings of the
First International Knapweed Symposium of the Twenty-First Century (ed L. Smith),
March 15-16, 2001, Coeur d’Alene, Idaho, pp. 1-7. United States Department of
Agriculture, Agricultural Research Service, Albany, California.
Harris, P. & Cranston, R. (1979) An economic evaluation of control methods for diffuse and
spotted knapweed in western Canada. Canadian Journal of Plant Science, 59, 375-382.
Howell, J.T. (1959) Distributional data on weedy thistles in western North America. Leaflet
of Western Botany, 9,17-29.
Maddox, D.M. (1979) The knapweeds: Their economics and biological control in the western
states, U.S.A. Rangelands, 1, 139-143.
McGregor, R.L., Barkley, T.M., Brooks, R.E. & Schofield, E.K. (eds). (1986) Flora of the
Great Plains. Great Plains Flora Association. University Press of Kansas, Lawrence,
Kansas. 1392 p.
Müller-Schärer, H. & Schroeder, D. (1993) The biological control of Centaurea spp. in North
America: Do insects solve the problem? Pesticide Science, 37, 343-353.
Myers, J.H., Risley, C. & Eng, R. (1989) The ability of plants to compensate for insect attack:
why biological control of weeds with insects is so difficult. Proceedings of the VII
5
International Symposium on Biological Control of Weeds (ed E.S. Delfosse), 6-11
March 1988, Rome, Italy, pp. 67-73. Istituto Sperimentale per la Patologia Vegetale,
Ministero dell'Agricoltura e delle Foreste, Rome, Italy.
Ochsmann, J. (2001) On the taxonomy of spotted knapweed (Centaurea stoebe L.).
Proceedings of the First International Knapweed Symposium of the Twenty-First
Century (ed L. Smith), 15-16 March, 2001, Coeur d’Alene, Idaho, pp. 33-41. United
States Department of Agriculture, Agricultural Research Service, Albany, California.
Piper, G.L. & Rosenthal, S.S. (1995) Diffuse knapweed. Biological Control in the Western
United States: Accomplishments and Benefits of Regional Research Project W-84, 19641989 (eds J.R. Nechols, L.A. Andres, J.W. Beardsley, R.D. Goeden & C.G. Jackson),
pp. 237-241. University of California, Division of Agriculture and Natural Resources,
Oakland. Publication 3361.
Powell, R.D. (1989) The functional forms of density-dependent birth and death rates in
diffuse knapweed (Centaurea diffusa) explain why it has not been controlled by
Urophora affinis, U. quadrifasctata and Sphenoptera jugoslavica. Proceedings of the
VII International Symposium on Biological Control of Weeds (ed E.S. Delfosse), 6-11
March 1988, Rome, Italy, pp. 195-202. Istituto Sperimentale per la Patologia Vegetale,
Ministero dell'Agricoltura e delle Foreste, Rome, Italy.
Powell, R.D. & Myers J.H. (1988) The effect of Sphenoptera jugoslavica Obenb. (Col.,
Buprestidae) on its host plant Centaurea diffusa Lam. (Compositae). Journal of Applied
Entomology, 106, 25-45.
Rees, N.E., Quimby, P.C., Piper, G.L., Coombs, E.M., Turner, C.E., Spencer, N.R. &
Knutson, L.V. (eds). (1996) Biological Control of Weeds in the West. Western Society
of Weed Science., USDA-ARS, Montana State University, Bozeman, Montana. World
Color Printers, Bozeman, Montana.
Rice, P. (2000) Invaders Database. http://invader.dbs.umt.edu/
Roché, B.F., Jr.& Roché, C.T. (1999) Diffuse knapweed. Biology and Management of
Noxious Rangeland Weeds (eds R.L. Sheley & J.K. Petroff), pp. 217-230. Oregon State
University Press, Corvallis.
Seastedt, T.R., Gregory, N. & Buckner, D. (2003) Effect of biocontrol insects on diffuse
knapweed (Centaurea diffusa) in a Colorado grassland. Weed Science, 51, 237-245.
Sheley, R.L., Jacobs, J.S. & Carpinelli, M.F. (1998) Distribution, biology, and management of
diffuse (Centaurea diffusa) and spotted knapweed (Centaurea maculosa). Weed
Technology, 12, 353-362.
6
Stinson, C.S.A., Schroeder, D. & Marquardt, K. (1994) Investigations on Cyphocleonus
achates (Fahr.) (Col.: Curculionidae), a potential biological control agent of spotted
knapweed (Centaurea maculosa Lam.) and diffuse knapweed (C. diffusa Lam.)
(Compositae) in North America. Journal of Applied Entomology, 117, 35-50.
Story, J.M. & Piper, G.L. (2001) Status of biological control efforts against spotted and
diffuse knapweed. Proceedings of the First International Knapweed Symposium of the
Twenty-First Century (ed L. Smith), March 15-16, 2001, Coeur d’Alene, ID, pp. 11-17.
United States Department of Agriculture, Agricultural Research Service, Albany,
California.
Story, J.M., Smith, L. & Good, W.R. (2001) Relationship among growth attributes of spotted
knapweed (Centaurea maculosa) in western Montana. Weed Technology, 15, 750-761.
7
Figure 1. Rate of spread of Centaurea diffusa in the northwestern United States (data from
Rice 2000)
8
Table 1. Biological control agents released in North America to control C. diffusa and C.
stoebe
First introSpecies
Order: family
Releases at study sitesb
ductiona
Shannon
91,92,94
Eickhoff
Seed head insects
Bangasternus fausti
Col.: Curculionidae
1990
Chaetorellia acrolophi
Dip.: Tephritidae
1992
Larinus minutus
Col.: Curculionidae
1991
Larinus obtusus
Col.: Curculionidae
1993
Metzneria paucipunctella
Lep.: Gelechiidae
1980
Terellia virens
Dip.: Tephritidae
1992
96
Urophora affinis
Dip.: Tephritidae
1973
90
Urophora quadrifasciata
Dip.: Tephritidae
1981
90
Agapeta zoegana
Lep.: Tortricidae
1984
90,96
96,97
Cyphocleonus achates
Col.: Curculionidae
1988
95,97
96,97
Pelochrista medullana
Lep.: Tortricidae
1984
Pterolonche inspersa
Lep.: Pterolonchidae
1986
90,92,96,97
97
Sphenoptera jugoslavica
Col.: Buprestidae
1980
96,98
90
91,92
96
95
Root-feeding insects
a First introduction of the species to the United States (Rees et al. 1996).
b Year
of release omitting first 2 digits (e.g., 90 = 1990).
9
Table 2. Proportion of C. diffusa roots and seed heads infested by insects at two sites in
Fergus County, Montana
Infestation rate
Shannon
Eickhoff
Agapeta zoegana
20%
17%
Cyphocleonus achates
52%
29%
Sphenoptera jugoslavica
2%
12%
Root damage
74%
69%
86
99
Larinus minutus
62%
10%
Urophora spp.
37%
49%
No. seed heads sampled
145
299
Roots (June 1999)
No. plants sampled
Seed heads (Oct. 1998)
10
Table 3. Decrease of diffuse knapweed at two locations in Fergus County, Montana
Densitya
1998
1999
Reductionb
Eickhoff location
Bolted plants
8.3 + 3.6
2.3 + 0.8
72% *
Bolts
10.3 + 4.4
2.5 + 1.0
76%
Rosettes
10.7 + 3.4
3.5 + 2.2
67%
No. quadrats
6
10
Shannon
Bolted plants
24.3 + 3.6
2.4 + 0.8
90% **
Bolts
38.6 + 4.4
3.6 + 1.0
91% **
Rosettes
31.0 + 3.4
0.8 + 2.2
98% **
No. quadrats
12
60
a
Mean + SE per Daubenmire frame (0.1 m2) on permanent transects.
b
One-way ANOVA test of difference between years, * = P < 0.05, ** = P < 0.001.
11