North Central Washington

I was a Knapweed Millionaire: The Knapweed Chronicles

George Wooten
Twisp, Washington
November, 2001

I moved to the Methow Valley in Okanogan County, Washington in 1985, following my parents' move to the Chiliwist Valley nearby by about 15 years. The place I moved onto is a 20 acre parcel with gravity-fed irrigation canal (ditch) water. The house rests on the glacial till of a low bench on the valley floor, set up against a steep slope of slaty bedrock affectionately termed Knob Hill.

The old farmhouse sitting in the middle of the property is sometimes referred to as the old Funderberg place by unknowing individuals who don’t realize that it is really the old King place. The King family built the house sometime in the 1950s, starting with the building we now use as a tool shed. The original 40 acres already had a small house on it in 1912, which was as far back as our neighbor Ken White could trace the history of the place. Another neighbor remembers that in the thirties, all of the original two forty-acre parcels sold for back taxes of less than 50 dollars.

At the upper, east end of the property, where the irrigation ditch enters, is an abandoned apple orchard, which still has about thirty trees. This orchard is part of a contiguous strip of old resprouted or surviving original apple trees that runs along the base of the hill for about ten miles between Twisp and Winthrop.

At the lower, west end, the property borders a small depression which was formerly a seasonal pond, but which apparently went dry after a well was drilled in it by our neighbor, who had a newer ranchette and about thirty cattle on his ten acres when we moved in.

When we moved here, the north end of this spread had several dilapidated livestock sheds and a sprawling gash in the ground which was formerly used as a potato washing and bagging plant. The ruins, as this junkyard was called by the locals, has mostly collapsed over the years, and the lumber has been recycled into a number of other construction projects. Scattered piles of boards, junk, fence, and old tires lie in various states of composure.

Diffuse knapweed seed head in flower

In the late 1980s, most of the property that wasn’t sterile sand was dominated by a monoclone of diffuse knapweed (Centaurea diffusa), with lesser amounts of downy cheatgrass (Bromus tectorum) and quackgrass (Agropyron repens). To practice my botany skills, I once counted the number of species in the knapweed fields and came up with a total diversity of about a dozen species. In addition to the three species above there was also introduced crane’s bill (Erodium cicutarium), tumbleweed (Salsola kali) and several of the tumble mustards (Sisymbrium altissimum, S. loeselii, and Descurainia sophia). Native plants were scant - there was a single bush of native squaw currant (Ribes cereum), two plants of native chocolate tips (Lomatium dissectum), and a single plant of bitterbrush (Purshia tridentata).

A little knowledge...

Almost all of our knapweed is diffuse knapweed. We don't have spotted knapweed in this area (or didn't until it appeared on Forest Service property nearby), and it has been interesting to compare the differences in its ecology with other areas that have had spotted but not diffuse knapweed.

I was first introduced to diffuse knapweed by my father, who had lots of it on his cattle ranch in the Chiliwist Valley. He said that if someone were to figure out a way to get rid of it, they could make a million dollars. Interested, I brought a specimen back east to my botany professor, who banished it to a sterile container, and admonished that the way to get rid of it would be to study its biology for years and years and years, which I proceeded to do when I moved to the Methow Valley.

For starters, I learned that the plant isn't all bad, and in fact there are some species like downy cheat or quackgrass that are considerably more deleterious to the ecosystem, but which are nonetheless browsed in the spring by livestock and deer and so are left alone by most ranchers. But I also learned that mere discussion of any virtues for this despised plant could result in loud haranguing by whomever overheard such whimsy. For instance, diffuse knapweed forms the basis for a large honey industry in eastern Washington's dry interior. Its dependable blooms are practically the only flowers that bloom in mid-summer. Nay-sayers pooh-pooh this fact, and instead claim that knapweed honey is bad-tasting, but this isn’t true. Without knapweed, the honey industry of the interior Columbia Basin would be on hard times, as it appears to be headed.

The list of beneficial uses which I found for diffuse knapweed also include that (2) it produces a fine and even layer of litter which acts as an herbicidal mat to keep out other species. At least if you have knapweed you needn’t worry about new weeds invading. (3) The plants and their duff layer are highly flammable and quick burning, facilitating yearly burning along fence edges where it often dominates. (4) Monoclones of evenly tall plants direct the wind upward and protect the soil from desiccation. (5) Knapweed produces abundant crops, which are browsed by wildlife such as mule deer, finches and rodents, as well as livestock. Mule deer appear to be one of the main dispersal mechanisms, as the plant often travels along deer paths in mule deer migration corridors. In addition to dispersal, deer and cattle browsing encapsulates the seeds in fertilizer packets, which are handily planted by trampling hooves. Range cattle are less effective dispersers partly because they prefer wet areas where knapweed has a more difficult time establishing itself, and also because of their grazing habits, which tend to favor grasses which can outcompete knapweed. Deer mice are another main ecosystem agent affecting knapweed populations. Apparently, they have a fond preference for knapweed seeds. Thousands of rodent tunnels appear from the melting snow each spring, comprised almost entirely of knapweed seed caches. Our cats as well as a local great horned owl and occasional northern harriers spend many hours in the fields hunting among the network of tunnels.

The list of beneficial uses continues into the more obscure. (6) Noting that real estate agents have the cleanest undercarriages on their cars, a local mechanic reasoned that this was because they were always driving through knapweed fields to show houses. (7) Trespassing, and the need for fencing, is reduced in knapweed because the plants are unpleasant to walk through, and stray neighbor kids are deterred from tromping wherever they want. (8) When moving irrigation lines, there is only one plant that can push out plugged lines with the water still on, but without having to shut down the system – knapweed stems. (9) A visitor from California found some baled knapweed left by a local farmer. Noting the strength of the bales relative to grass hay, this person resolutely became convinced that knapweed could be the basis of a knapweed-bale house industry, because the bales were so much stronger than hay. Against our protests, he returned to California with a handful of seeds.

(10) From observing a herd of llamas which lived here for five years, we found that llamas cultivate knapweed as a bitter tonic. You won't believe this, but I swear it is true. Although llamas don't make much use of knapweed in their diet, they do cultivate and prune individual plants in a perennial garden. These plants are used following llama arguments. Understand, there are four types of llama oral expressions: Type 1 is a “puh” sound made with both lips and used as a note of concern, usually between two llamas deciding who was first in line. Type 2 is a sputter or cough, which is occasionally used when the two llamas disagree over the prior decision of who was first in line. This is the origin of the idea that llamas “spit”, however it is not used on humans (except accidentally) and it only involves a small amount of expelled vapor. Type 3 is an overtly aggressive ejaculation of whatever is in the llama’s mouth at the time. It is used during serious arguments between two or more animals. Type 4 is a defensive regurgitation used to fight off attackers. I have never seen this display.

But on a couple occasions, I saw llamas in type 3 aggressive behavior. Once, two of our llamas got into an argument (over food) and both spit a wad of green cud at each other. Within ten seconds they both trotted over to the knapweed patch, and piled knapweed leaves loosely onto their tongues. They remained with the green leaves sitting on their tongues, with their mouths open for about five minutes. Apparently only knapweed bitters have the necessary strength to cut through llama oral and gastric juices.

According to locals, knapweed really got out of control during the 1960s. There are many who insist that the Forest Service intentionally used it for seed stock following early wildfire restoration and this same anecdote is also sometimes applied to the Game Department. I think this unlikely, as such managers would quickly find out how noxious the plant is and recommend against it, halting any concerted efforts. While a few individuals may have tried some small seeding programs here and there, it is more likely that the original source of knapweed here is due to the use of contaminated grass seed originally imported from Eurasia.

Diffuse knapweed has some very effective biological mechanisms for dispersing itself. The whole plant seems to have been designed with wheeled vehicles in mind. For pious folk, this might suggest the idea of a wrathful creator, or at least a capricious one.

The alternative that evolution designed knapweed with the automobile in mind is somewhat hard to accept, however it might be noted that the home range of diffuse knapweed coincides well with the cradle of civilization in Eurasia. One might also note that these are also the lands of the Aryan invaders, so that the biology of the horse may have found a surrogate in automobiles. Regardless of the origin, the plant is actively adapting to automobiles and roads.

When mature, the dry, rigid skeletons of knapweed with seedheads intact readily snap off at the base when driven over, or if the car won't come to knapweed, knapweed goes to the car -- by travelling on the wind as a tumbleweed, whee! Upon finding a suitable vehicle, the knapweed stems will jam up under the undercarriage and exhaust system, where they can bounce along for miles while shaking out the seeds. The seed heads are just the height of a car door, and the stems are supple and springy, so that they fairly jump into the door upon one's getting out of a car in a field of knapweed. It takes several hands and feet to push all the stems back out, and then the door can’t be shut, so that inevitably one gives up and slams the door over the stems in disgust, ripping the seed heads off the plant and into the car while driving off. The seed heads are of the same gauge as most mud and snow treads, and flexible enough to fit in narrower passenger tire treads. The stiff, prickly bracts make sure that seed heads that lodge in tire treads stay there for at least a week, maximizing dispersal distance.

I am positive that if cameras existed during the time of the Roman Empire, there would be pictures of knapweed rolling along in the chariot wheels.

Controlled burning

The farm I moved to in the Methow Valley laid fallow for a number of years, following the mysterious disappearance of the manager after leaving for a farmers tractor march to Washington DC in the sixties. Knapweed management began with the owner who preceded me, who regularly burned the lands around the house and fields. Using this method, he was able to knock back the field population in some places by about 50%, but when I moved in, knapweed had reclaimed this ground while I concentrated my battle efforts in the immediate vicinity of the house.

The use of controlled burning on knapweed needs to be looked at in perspective. Until the advent of biological controls for knapweed, this was the most effective method on this property. On the other hand, no other method poses the risk of escaped fire, with such a large liability. Knapweed burns both hot and fast, and the margin between an out-of-control brush fire and a fire that won’t even stay lit can be due to just a change in humidity. In Okanogan County, fall and spring burning are common activities that seldom incur any permitting. But when fires get out of control, those who started them will be assessed the costs if a permit violation is found, which might be upward of $50,000 a day if air support is required.

My own experience with knapweed field burning occurred in 2000. I had been doing night burns for three nights, and was on the last night. I had black-lined three sides of the field, and with the ditch behind me, felt pretty secure to lengthen the line of fire to 100 feet. A slight down-slope breeze kept the fire less than a foot high and wide. Over an hour or so, the line slowly backed toward the ditch. With only about 100 feet left to the ditch, the line reached the toe of the slope. Then something happened which I hadn’t reckoned on. There was a zone of dead air at the base of the slope. When the knapweed reached that part of the field, the breeze died, the flames suddenly stood straight up, ten feet tall, and the flame front began moving by radiant conduction about a foot every ten seconds. At this rate, within about ten minutes the flames would be licking into the apple trees and cottonwoods along the ditch. With flames ten foot high, the fire could certainly jump the ditch, in which case the whole fire department would be called out. This was early spring, so the ditch wasn’t on to provide water yet.

My partner ran back to the house and called all the neighbors, while I jumped from the hot to the black side of the fire. My shovel was completely useless against the wall of flame. For about 10 minutes I raced back and forth along the line, watching the fire get closer to the trees. Suddenly, just as one of the neighbors arrived, the breeze returned, the fire laid down, and within about ten seconds the fire was essentially out. I had to discourage the neighbor from trying to re-light the blaze.

If you want to use this method, I suggest it might be appropriate for large, heavily infested properties abutting state or federal lands where government firefighters will assume most or all of the liability for doing a controlled burn, perhaps in conjunction with a burn on public land. One modification might be to use a drift fence to catch the tumbleweed knapweed plants, and then only burn the fence line. The goal is to prevent infection beyond the fence.

Strategy - defining and maintaining a defensible space

I decided to try and get rid of knapweed on this property. Herbicides were considered out of the question, partly because they are killing the planet, partly because they are killing my friends, partly because herbicide companies are getting rich off of this legalized murder, and largely because they smell bad and make me sneeze and cough while my face gets florid, my nose bleeds, my head pounds, I get dizzy and then go into a mental and physical depression. I formerly worked at a solvent-based coater at a paper company, where my lifetime exposure limit was already exceeded. Many herbicides use these same organic solvents. The commonplace use of herbicide by folks who have been duped into complacency is a personal sore spot which I cannot write without getting emotional and out of hand. But my friend and author Jeanne Hardy wrote a reasoned, calm, passionate dialogue in her long running newspaper column documenting her fall from health caused by exposure to wantonly abused bad chemicals, including herbicides.

My efforts to remove knapweed over the first three years of living in the Methow Valley began with a resolution to spend exactly 24 hours yearly clearing knapweed away from the house perimeter. The methods used were divided between removal of rosettes in the spring using a spade, pulling up plants at the bolt stage following rain events to make it easier, and mowing and cutting the tops off of plants that couldn’t be killed. The objective was to stop seed production. In my war against knapweed, I assigned a point system to encourage effectiveness, and keep my interest piqued. For each second-year plant pulled out by the roots or cut off below the crown, regardless of its height, I received ten points. This way I found out that cutting off the crowns in the spring was faster and more efficient than pulling, although the yard divots were somewhat unsightly.

But the yeoman's work of knapweed control remained hand-pulling, using leather gloves.

In removing knapweed, there are two types of plants to consider – rosettes and mature plants. First year rosettes seem to be about ten times more common than mature plants, yet their weak stems make them about equal difficult to pull up by the roots. I reasoned that my limited energy should be directed toward the seed plants. Since each mature plant can produce over a thousand seeds, seed plants should be worth lots more than first-year rosettes. For this reason, I granted myself only 1 point for each first-year rosette pulled out, with the idea being that these plants were like the pawns in a chess game, expendable decoys to take away energy better spent on mature plants. It wasn't worth the trouble to pursue them, but if they were killed in the course of getting at the big guys, then there was a small benefit. Where the first-year rosettes grew really thick, sometimes you could cut them below the soil ten at a time with a shovel. I used the shovel primarily in the driveway, where plants grew tended to grow thickly.

I learned a lot about knapweed biology from this point system. When an area is cleared of knapweed leaving only a single plant, regardless of whether that plant is a first- or second-year plant, it will grow to maturity in that season. I don't really think knapweed is a true biennial, only facultatively so. First year plants remain as rosettes for the first year unless the canopy is cleared, in which case they revert to seed plants. Thus, if you want to stop yearly seed production, you have to be 100% effective in removing all stems and rosettes from an area. But removing all the rosettes was impractical, because they are fragile, and frequently break off and resprout; furthermore, their numbers are often up to ten times that of the mature stems. The point system forced me to ignore the rosettes and accept that some bolting would be the cost. The consequences also meant that each area treated required at least a second pulling near seed time, to get the rosettes that managed to convert to mature plants.

The final pulling of the season amounted to a mop-up operation. The goal was to have zero percent seed production within the cleared perimeter. All resprouted plants within the perimeter were pulled out, preferably right after a precipitation event which would soften the soil and make pulling about ten times more effective, so I would accrue more points. If rain didn’t come, or if plants were not mature enough, there was a strong likelihood that ssome stems would break upon being pulled. Plants with broken stems invariably resprout to form multiple seed stalks and a strengthened root system. These become the most intractable of plants to remove. The short stems offer no place to grab on to, and even if a good grip is made, the thickened roots put up an effective fight. Typically, the stem slips through one’s gloves, and the seed heads roll happily away. These toughened plants then convert to perennials, where they will be found to reappear every year until pried out with steel tools.

Because of this changeling aspect of knapweed, a score of negative five (-5) points was assigned for each mature plant that broke off above the crown. The goal was to have a positive score after 24 hours. During the first two years of the knapweed war, 24 hours spent maximizing my score was just enough to maintain a 50-foot wide knapweed-free zone around the house. On the third year, the results finally began to show. The knapweed within fifty feet of the house was reduced to less than fifty percent of the original cover, where it was replaced by Kentucky bluegrass, my lawn.

Widening the perimeter

When a secure zone was thus established and maintained for a couple years, extra time became available to expand the battle line. By about the tenth year, the knapweed-free zone had expanded to a minimum of 100 feet in one direction, and close to 500 feet in two directions.

The outer fields do not receive any traffic, so the sandy soil is fragile and plants are easily pulled. Knapweed growth in the outer fields is balanced by two factors: deep soil which aids root growth, and rapid percolation which desiccates plants. The mortality of rosettes is much higher, and the crown density of mature plants is thus lowered (crown thinning is counteracted over time if the knapweed stand builds up a litter layer, in which case it can form impenetrable monoclones). When plants do reach maturity, they reach epic proportions, sometimes over four feet tall, and completely covered with fertile seed heads. On the other hand, pulling plants out of the loose, sandy soil is extremely easy, regardless of their size. The hard stems don’t break and no bending over is necessary.

On about year six, I over-extended my pulling line out to about 500 feet in one direction, however the following year, there were over twice as many mature plants in this area, and I had to retreat. I am not sure why this happened, but I believe this was due to soil disturbance that exposed more seeds. A year after the retreat I regained some of the lost ground.

Biological controls grasses

Two years after the battle where I was forced to retreat, the appearance of this field had undergone an amazing, even unbelievable transformation. The entire field, clear to the 500-foot line of the first battle, had become occupied by a native bunchgrass, sand dropseed (Sporobolus cryptandrus). I do not know whether the grass had been slowly expanding or whether I had merely failed to notice it, but since then, I have had no knapweed in that field. Even though it is a bunchgrass, sand dropseed forms monoclone populations able to exclude knapweed as well as other plants. I am currently experimenting in growing mixtures of sand dropseed and native needlegrasses (Stipa spp.) to see if a more diverse mix can be encouraged.

Another retreat occurred during a period of overzealous lawnmower use, not entirely of my own volition. Plants in the bluegrass around the yard never had a chance to elongate their stems under a regimen of frequent lawn mowing, hence they couldn't be pulled. Although the knapweed plants in the mowed parts of the yard weren't able to go to seed, the cut off stems became little daggers for bare feet, hardly the picture of a manicured lawn. Unlike smooth brome or quackgrass, bluegrass does not compete well with knapweed. I believe this is due to the fact that the roots and the stems of the plants do not overlap the same vertical spaces.

About five years ago, approximately 1997, I noticed competition occurring on the upper end of the property, only this time the knapweed was thinning out because of quackgrass (Agropyron repens) competition. Unfortunately, quackgrass is even harder to get rid of than knapweed. For this and many other reasons, it should be considered as undesirable in shrub-steppe ecosystems or agricultural settings. Nonetheless, there was nothing to be done of it.

quackgrass invading knapweed from the left (Click for larger image of quackgrass, green, invading knapweed) I began watching the battle between quackgrass and knapweed more closely, along a path which I cut through the knapweed to access the ditch headbox. This path cuts through the old apple orchard and receives no irrigation. Each year I used the same path as a short ski run in the winter, as soon as the knapweed is buried under the snow. During recent years, I noticed that the ski run was free to go in any direction because knapweed is disappearing from this part of the farm.

Quackgrass overcomes knapweed slowly, but surely. As its rhizomes advance through the soil, knapweed stands become islands in a sea of grass. The destabilization of the knapweed community brings in mice, tumblemustard and wild lettuce, the knapweed stand shrinks, and then one year it is gone. The quackgrass was advancing at about two feet a year, and could be expected to overtake all the knapweed on the farm, even in the most intractable areas, in about 10 or 15 more years.


Where there was originally a monoclone of knapweed above the house, I planted two fields of alfalfa-grass mixtures in the 1990s and irrigated them regularly. These fields were planted five years apart so that watering could be concentrated on the most recent, more productive field. The original field with finer soil has now naturally converted to bluegrass and orchard grass, while the second field is converting to smooth brome. Smooth brome maintains itself free of knapweed if provided with two waterings a season. From observation, it appears that smooth brome needs the equivalent of 15” a year precipitation if it is to maintain itself against knapweed. The average in Winthrop is 12 inches per year so the two waterings make up the deficit.


In the last five years I have spent the most time on the tenth mile of dirt driveway leading to the house. Knapweed thrives along roadsides where yearly grading pushes seeds along the drive and into the soft, prepared soil. Runoff from the compacted center percolates into the edges of the roadbanks where it helps to grow strong plants. At the edge of the compacted part of the road, plants that get driven over become hardened perennials, completely resistant to pulling. If someone wanted to design the perfect site for growing knapweed it would look like a dirt road with yearly road grading.

Control of knapweed along the road consisted of spading the compacted areas free of plants and later in the season pulling out the remaining large plants. By leaving the large plants for last, they shaded and suppressed the growth of smaller rosettes so that there was less pulling necessary when the time came to do it. Pulling occurred in two stages as before. Pulling the road often had to extend past the flowering season, which invariably meant that the seeds had already matured. As soon as flower petals are visible, plants should be considered viable. Pulled plants with viable seed must not be left lying where they came from. Typically, these pulled plants, along with their seeds, were taken to a boneyard or other area where knapweed is already common. Beyond some time in August, pulling is no longer effective. This is the point when the flower heads have completely opened, and the seeds scatter if the plants are touched.

Along roadsides, succession proceeds differently than in the fields or the yard. When knapweed density is lowered, roadside plants become dominated by three main plants: tumble mustard (Sysimbrium altissimum), wild lettuce (Lactuca serriola), and Russian tumbleweed (Salsola kali). The first two are a good sign that the area is in transition, and they often give way to grasses or wild roses or even shrubs. Tumbleweed, on the other hand, has taken over large areas along the roadsides that were formerly occupied by knapweed. Its biology is completely different – it has the rounded profile of a plant adapted to rolling across playas dominants and it relishes alkaline soils (and burning) as would be expected of a basin halophyte. Control of tumbleweed requires a different control system from knapweed. It is also resistant to herbicides, which the county sprays with reckless abandon along every road. It appears to be gaining ground, along with another long-time invasive that responds positively to herbicides – flannel mullein (Verbascum thapsus).

Biological controls – Sphenoptera jugoslavica beetles

In the late 1980s, with the help of Gary Piper of WSU, Okanogan county’s first population of the root-boring beetle, Sphenoptera jugoslavica, was established on my father’s Chiliwist ranch. Over the years, the site was used a number of times for collection and distribution of the beetles, using sweep nets to gather the adults. These were successfully established on the Methow property right about 1990, where they were released in the center of a mega-patch at the base of the hill. Within about five years, they began to appear as far away as 20 miles, and today they are ubiquitous along the sixty or so miles of valley floor, as well as most of the county.

Initially, their effect on knapweed appeared to be neutral. The big knapweed population they were released within is still going strong in 2001. Around 1997, the effects of Sphenoptera beetles began to show up, primarily in isolated plants at the periphery of big populations. Sphenoptera beetles apparently don't kill knapweed outright, but they convert the plants into short, stocky, multi-stemmed, perennial dwarves, more akin to bonsai plants than weeds. The root excavations produced by the beetles gradually enlarge the root over a number of years, forming it into a knobby hollowed-out home for the larvae. Although the number of seed heads is greatly reduced on the shortened plants, seed production per head does not appear to be greatly reduced.

This seed head predation by Sphenoptera contrasts with that of knapweed biological controls released several decades ago, the seed-head gall flies, Urophora affinis and U. quadrifasciata. Actually I should say “escaped” biological controls, since these two species reached the US by flying surreptitiously across the border from unsanctioned releases in Canada. Although they have become ubiquitous in Okanogan county, studies indicate that their presence has not reduced knapweed, and in fact they can actually benefit knapweed slightly, even when they consume up to 95% of the seeds in a plant. My guess is that they select for the 5% of fittest seeds.

In the year 2000, an unexpected ecological twist developed in Sphenoptera-infected knapweed. When I began my yearly pulling of knapweed in the driveway, I first finished off the compacted area with a shovel, then proceeded to pull plants out from a sandy area next to the road. As I gripped the first plant with my gloves, I was immediately attacked by a swarm of about a dozen large red ants. Leaving that plant alone, I went to the next plant and got the same result; in fact every plant in the area was being guarded by a small group of these vicious ants, which look similar if not identical to the same kind that normally produce large mounds in shrub-steppe habitats.

There is a trick to getting the ants off of these plants without getting bitten. If the plant is quickly shaken and then let go of, the ants all move out for the attack and keep right on going off to war somewhere else. In about five seconds or so, the knapweed plant is vacated of them and can be pulled with safety and ease, since the thick root offers a good handhold, and it does not go very deep in the sandy soil preferred in the Sphenoptera- ant ecosystem. Peeling open the rootstock reveals a catacomb, sometimes with the larvae still grazing away inside the tunnels. The inside walls are inevitably lined with many small, gray discs of some type of fungus. I suspect that the ants are using the plants as remote outposts for cultivation of this fungus, which may be later brought back to the main nest, or perhaps used in situ.

After this I no longer pull the stocky plants that grow in sandy soil; besides they are hardly a threat any more. Populations that formerly grew hundreds of 1- to 3-foot tall plants, support only a few dozen of the dwarf knapweeds. Mark my words, the day will come when knapweed is no longer a problem, and although few will miss it, what follows it could very well be worse. So I am hanging on to my best plants.

Biological controls Larinus minutus weevils (beetles)

In 1999, Steve Dupey, a local naturalist, showed up at the Twisp Farmer's market selling banks of mason bees and vials of knapweed-eating weevils, Larinus minutus.

Incidentally, Steve's primary market was the mason bees, which are being used by orchardists as surrogate pollinators to replace collapsing populations of honey bees. Honey bees have been decimated in the past few decades by a number of pests. In the Columbia Basin, I wonder whether honey bee losses may be related to steadily lowered crops of their main summer staple, knapweed. While one might counter that historically, wild flowers would have been the summer nectar source, rather than knapweed, consider that the role of summer nectar sources may just as well have been moot in this ecosystem as honey bees are also an introduced species. Perhaps the native wasps, yellowjackets, and solitary bumblebees do not require large supplies of summer nectar sources in the present-day range of knapweed, which is to say ponderosa-pine shrub-steppe.

Several of Steve's customers hired him as a contractor to disseminate Larinus widely. In the early summer of 2001, I found my first population of Larinus in downtown Twisp and was astonished at what I saw. Every plant had from a couple to several dozen of the weevils on the seed heads. Surveying the valley for more Larinus revealed that there were strong populations in widely dispersed areas such as the Chewuch River valley, around Carlton, Finley Canyon and Twisp, but that they thinned out to zero over about two miles. This works out to a tentative rate of spread of about a mile a year, considerably slower than Sphenoptera or Urophora. There were absolutely none on the ranch where I live, although I found them two miles south.

Remembering the words of my father years ago, I decided to become a knapweed-weevil millionaire, and began selling them through ads in the newspapers.

Releasing Larinus minutus in Republic, Washington (Click to see enlarged image) The price was 50 cents a beetle for the $25 dollar minimum, or $50 for about 500 of them. Collection of Larinus was incredibly easy at one site which averaged over a hundred per plant. One simply holds a plastic 5-gallon bucket under the plant and shakes them into the bucket. As long as the bucket is kept jostling, they cannot get airborne. When the desired number are in the bucket, they can be transferred to a smaller container.

Some of the collection areas with good Larinus populations also had Sphenoptera. Some areas also had attacking hordes of red ants which would get into the bucket and bite the crap out of the poor weevils, not to mention the collector, so a more complicated shake and sift procedure had to be performed to get the ants out. Ultimately, I opted to build a rearing pen, which was a 50 gallon barrel stuffed with fresh knapweed and covered with a screen. Whenever a knapweed order came in, I could reach in quickly and shake the weevils into a coffee can.

The weevils I had been harvesting were overwintered Larinus adults freshly emerged from the ground litter where they had been hibernating. I stopped selling the beetles when they became less prevalent during late summer. The female weevil must feed on the flowers of the knapweed for ovary development; therefore egg laying begins after the knapweed has started to bloom. Females lay eggs in clusters which hatch within three days. The larvae begin feeding on the flower parts and immature seeds within the seed head. Pupation takes place inside a cocoon made of seed and flower parts that is attached inside the developing seed head. When development of the larvae is complete in July and August, new adults emerge from the seedhead. These adults feed on the plant foliage before going into the ground litter to hibernate.

Both adult and larvae are destructive to knapweed plants. Larinus attacks both spotted and diffuse knapweed with a slight preference for the latter. Adults feed on young leaves in spring, and both leaves and flowers later on. Typically, a larva destroys all the seeds in a single seedhead.

Sites that are chosen for Larinus release should be considered for long term availability and should be left undisturbed by development or pesticide use for at least ten years. The release site should be dry with some bare ground. Frequently this will be the outer edges of the knapweed infestation.

Unexpected help from some unwanted allies

The knapweed complex has been targeted by the USDA as a high priority for control. Larinus minutus was introduced along with over a dozen other knapweed-eating species as part of that program. Coincidentally, an unrelated outbreak of entrenched insect pests also joined the war on knapweed in the Methow Valley in 2001.

In 2001, the Methow Valley experienced an outbreak of case bagworms and army cutworms which both caused great damage to knapweed, along with our gardens. No one is sure why these two generalist pests both exploded this year. During the spring, many gardens were ravaged by army cutworms which consumed every green leaf that sprouted. Since I grow catnip, I was at first relieved to find that the cutworms left the plants alone. But as the season progressed, the catnip also turned brown and dry, and the leaves became perforated. The plants were under attack by a second insect outbreak, in this case, caseworms (also called mudworms or bagworms, but which I call case bagworms to avoid confusion with other species of bagworms). These critters can be recognized as those familiar little curled balls the size of BB-pellets decorating the undersides of eaves, limbs and junk piles. In a decade of living on the same ranch, I had never seen case bagworms become so abundant or such a nuisance.

The cumulative effect was a complete surprise in the Methow Valley. Entire fields of knapweed disappeared overnight. On the ranch I live on, between the case bagworms and the army cutworms, every single rosette was consumed on one acre of ground. The main effect was total removal of the new and second-year rosettes, which typically make up the bulk of the year’s production. Areas which I can remember being barely able to drag my garden cart through the year before were now completely devoid of living vegetation, and were decorated only with dead stems from the previous year, or perhaps a few stems of cheatgrass and the occasional tumblemustard, also severely munched. Thus, competition from other plants could not explain what had happened because the large, bare patches were not yet invaded by new weeds. So profound was the loss of vegetative cover that several people swore it had to have been clandestinely herbicided either by me or my neighbor. To support their contention, the field area behind the house still had an apparently healthy knapweed population (click for closeup).

Knapweed completely eliminated on the ranch
Photo showing knapweed completely eliminated on one part of the ranch - dead stems from previous year are not followed by new rosettes
Knapweed collapsed faster than grass advances
Knapweed collapsed faster than grass advances
Knapweed collapsed faster than grass advances
Knapweed collapsed faster than grass advances

A healthy discussion of these bizarre phenomena ensued. One person described how on their property, knapweed had been decimated by bagworms eating the plants, while another person swore that their knapweed had all been killed by cutworms. Steve Dupey was quick to point out that on his property, the Larinus minutus were responsible and they had done the job so well that when the knapweed ran out they were forced to nibble on Russian knapweed, which wasn’t expected. Apparently it was a combination of either, both and neither, depending on vagaries of local habitats. One of the factors may depend on whether the knapweed plants were infected with Sphenoptera or not, in which case plants appear to have had an advantage, as in the mega-population behind my house, which wasn’t reduced. Perhaps the Sphenoptera-infected plants were tougher, or contained more of knapweed’s bitter chemical, cnicin. But in any case, throughout the valley, everyone who normally pulled plants every year finally began to straighten up and notice the effect.

Apparently most of eastern Washington has not seen the dramatic effects in the Methow, and it is knapweed-as-usual. But for those who have seen biological controls work in the Methow Valley, the question is not whether, but how will the monumental collapse of this dominant invader occur. Will it all end suddenly, leaving great bare patches; or will knapweed now behave peacefully as a minor ecosystem component; or will only certain populations with particular microhabitats succumb leaving the others to continue spreading? And what will become of the insects - will they die out when their host dies, or will they adapt to a new ecosystem?

The words of entomologist Gary Piper, who provided the original releases of Sphenoptera for us, come to mind: “Knapweed isn’t going to go away until there are a half-dozen or so biological controls all working on it.” And now that it is going away, the words of Tom Brannon, extension agent from Wenatchee, Washington, also need to be heeded: “Whenever man creates a disturbed bare area, nature finds something to fill it in with.”