Diffuse Knapweed Control in Feral Babysbreath

Situation

Diffuse knapweed (Centaurea diffusa) and babysbreath (Gypsophila paniculata), both introduced from Eurasia, grow together on rangelands in northcentral Washington. Diffuse knapweed aggressively invades bitterbrush and sagebrush grass rangelands, reducing forage yields for livestock and wildlife. Babysbreath was originally planted as an ornamental flower in gardens and cemeteries, and had naturalized in eastern Washington by the 1950s. It invades sandy soils more aggressively than fine textured soils.

While diffuse knapweed continues to hold a negative value in rangeland management systems, feral babysbreath has been discovered by the dried floral industry. Wild babysbreath grows more compact, sturdier stems than its cultivated counterparts, and production costs are minimal. Valued at $1.50 per bunch, harvested feral babysbreath accounted for more than $5 million in the Northwest in 1989, primarily from Washington state.¹

Conflict arose when herbicide treatments for diffuse knapweed control rendered babysbreath unsalable in the dried floral industry. Picloram and dicamba are preferred herbicides for knapweed control because residual activity in the soil extends the duration of control. Eliminating annual reapplication costs, especially on rugged terrain, more than offsets higher chemical costs. The low rates effective on knapweed do not kill babysbreath, but prevent normal flowers, the marketable commodity. Control of diffuse knapweed remains important for slowing its spread and restoring rangeland productivity. However, the value of feral babysbreath on rangeland may exceed returns from grazing livestock. The object of this study was to find an economical method of controlling diffuse knapweed that does not injure babysbreath.

Diffuse knapweed is a taprooted biennial that reproduces by seed. Following fall or spring germination, diffuse knapweed normally remains a rosette during its first year. It flowers and seeds in July and August of its second year (hence biennial). In northcentral Washington, rosette leaves normally remain green over the winter (Fig. 1). Diffuse knapweed rosettes lie close to the ground, limiting their availability to cattle, but not necessarily to sheep and deer. Cattle graze diffuse knapweed most readily in the bolting stage, while the stem is still leafy and tender. Cattle, generally, unless short of feed, either ignore or avoid diffuse knapweed in other stages of development.Babysbreath is a long-lived perennial with thick roots that may penetrate 13 feet deep. Perennating buds overwinter on the root crown below the soil surface and emerge in April. Flowering usually peaks in July. Babysbreath is palatable to cattle from emergence until the stems begin to dry during flowering.

Diffuse knapweed overwinters as rosettes on top of the soil surface.

Dig around babysbreath crowns to locate the overwintering perennial buds beneath the soil surface.

Do not apply herbicides after new shoots of babysbreath emerge in April.

The herbicide 2,4-D amine was selected among rangeland-registered herbicides for its short persistence, lack of soil activity, and relatively low cost. Three rates (0.5, 1.0 and 2.0 lb/A acid equivalent) were applied before and shortly after babysbreath emerged to evaluate effectiveness in controlling diffuse knapweed and potential for damaging babysbreath. Babysbreath buds were just below the soil surface during the early application in late March (Fig. 2), and shoots were approximately 2 inches tall during the late application in April (Fig. 3).

To determine effects of repeated applications and longevity of control, researchers sprayed plots 1 year only (1989), 2 years (1989, 1990) or 3 years in succession (1989, 1990, 1991). All combinations of rates, timings and number of applications were tested in plots on an Ewall loamy fine sand located between Omak and Riverside on Pogue Flat, Okanogan County, Washington. Flowering and rosette diffuse knapweed plants were counted each July from 1989 through 1993. The cooperator recorded babysbreath quality at harvest in July. Perennial grass density and cover were recorded in October 1992 and 1993.

Effectiveness in knapweed control did not differ between early and late applications, except at the 0.5 lb/A rate. This rate did not consistently control diffuse knapweed when applied before babysbreath emerged. However, all postemergence applications damaged babysbreath by distorting or delaying flowers. Babysbreath flowers developed normally after preemergence applications at all rates. The level of control by 2 lb/A was not consistently better than 1 lb/A. Both rates reduced mature knapweed to 0 to 3 plants per 100 ft² in the year of application.

In 1990, mature diffuse knapweed was absent in plots that had been sprayed in each of 2 years (1989/1990) using 0.5, 1 or 2 lb/A, or a single year (1989) using 2 lb/A. Plots sprayed in 1989 and 1990 using 1 lb/A averaged 1.25 mature knapweed plants/100 ft² in 1990, compared with 4.5 plants/100 ft² in untreated plots. The number of rosettes in plots treated with 1 lb/A was about half the number in untreated plots.

In 1991, the number of rosettes and flowering plants in plots treated using 1 lb/A averaged 2.8 plants/100 ft², compared with 9.8 plants/100 ft² in untreated plots. No mature plants remained in plots sprayed for 3 consecutive years using 2 lb/A 2,4-D (see table).

One year following the last herbicide treatment (1992), the number of mature plants in plots that had been treated the previous year (1991) was significantly lower than in plots treated only in 1989 and 1990. Two years after the last herbicide application (1993), fewer rosettes and mature plants grew in all treated plots than in the check plots, indicating reduced knapweed seed reserves in the soil.

The number of knapweed rosettes fluctuated widely from year to year, depending on rainfall and temperature influences on seedling emergence and survival. The nonresidual herbicide 2,4-D controlled both rosettes and mature plants the year of treatment, resulting in fewer reproductive plants the year after treatment. Effects on rosettes beyond the year of treatment were limited to depletion of the seedbank, which declined after 2 to 3 years of seed prevention but could easily be replenished by a single year of normal seed production. Knapweed survival also appeared to be influenced by the amount of perennial grass. On unsprayed plots, knapweed density was low where perennial grass cover was high.