It all started with picking up trash.

In the process, we noticed several Yoshino Cherry trees smothered by kudzu vines dropping from trees above and some from the kudzu on the railroad banks.

This was 2001 in Spartanburg, South Carolina, a city of about 40,000 located in the northwest part of the state an hour or so from Charlotte, North Carolina, and maybe 45 minutes east of the Appalachian mountains.

Thus began a Master Gardner community service project to save these cherry trees. Little did we know what lay ahead.

After several years of learning by trial and error how to save the trees through killing the kudzu, we invited the late Dr. Larry Nelson (Clemson University Department of Forestry and Natural Resources) to review our work in August of 2004.

In the process, he mentioned his idea about using plastic to kill kudzu foliage. We liked the idea and proceeded to try it out. When Dr. Nelson came back to appraise the results in September (the plastic did, indeed, kill kudzu foliage), he suggested that we form a "coalition" of organizations and individuals who are interested in Spartanburg's beautification.

From that meeting, Henry Pittman and Newt Hardie decided that we would give the coalition idea a try if the Spartanburg Men's Garden Club and the Spartanburg Mayor would support it.

The Board of Directors of the Men's Garden Club approved the initiative, even though it was still only in the concept stage.

The "coalition to control kudzu" came to life in October of 2004 when Mayor Bill Barnet and City Manager Mark Scott enthusiastically endorsed the idea.

In 2007 the Kudzu Coalition became an independent 501(c)3 nonprofit organization.

Newt Hardie with suggestions from Kate Gardner, writer and gardener in Bozeman, Montana
Revised June 2008
Lessons Learned from a Project
To Save a City's Decorative Area

Newt Hardie
April 2005

ABSTRACT: Kudzu can be controlled by a method other than using chemical herbicides. Specific removal of root nodes and root crowns, the point of connection of vine to root, is demonstrated to be effective. While labor intensive, the techniques described reduce the future growth of this invasive plant without the potential environmental damage from chemical use. Specific procedures for identifying, undercutting, and removal of the nodes and root crowns are given. This paper is a summary of the authorís two-year endeavor to save endangered cherry trees, to remove the kudzu eyesore, and to develop non-chemical methods for controlling this damaging plant.

Kudzu vines engulfed 14 Yoshino cherry trees and created an eyesore along a major entranceway to Spartanburg, SC. A pruning saw, hand trowel, and pulaski (firefighting axe and grubbing tool) were found to be the most useful tools for severing runners and dislodging root crowns, thereby permitting the kudzu removal or roll back. Dense "scaffolding" of kudzu draped shrubs, and vines must be brought down before removing root crowns. These experiences and labor-intensive techniques are still under development but provide a starting point for further study and work.


In the 1980ís, the Spartanburg Menís Garden Club and city personnel planted 250 Yoshino cherry trees (Prunus x yedoensis) along 1.2 miles of both sides of a major route into the city. By 2001 Kudzu (Pueraria lobata) vines had entangled the branches of twelve of these trees and threatened two others. The author, a member of the Spartanburg Menís Garden club and a Master Gardner, undertook to save these 14 trees.


The use of herbicides requires far less physical labor and time than does mechanical control, but the proximity of Yoshino cherry trees, the slope of the ground down to a creek, and the authorís general resolve to avoid herbicides for environmental reasons led to the decision to pursue only mechanical control.


A Norfolk and Southern Railroad overpass bisects the Yoshino tree line and crosses the highway. For many years kudzu had overrun both sides of the overpass. North and south of the railroad bed, the land slopes sharply downward to the Yoshino tree line. Behind the tree line, kudzu capped two small groves of hardwood trees, and the vines from these trees dropped down into four of the Yoshinos. The project required a series of different approaches in order to address the differences in the maturity of kudzu, the different sizes of root crowns, the terrain, and the assault from the overhead vines.


A curved pruning saw with a 13-inch blade and a 7-inch handle proved to be easier and more effective for this activity than a sling blade. The teeth of the pruning saw worked surprisingly well in picking up vines close to the ground even among grass. After the vine had been plucked from the ground on the teeth of the pruning saw, a flip of the wrist could toss it onto the wall or blanket of mature kudzu outside of the Yoshino drip lines. This approach was effective with young vines and further exposed nearby root nodes. Tossing vines back on top of the kudzu behind the desired perimeter cleared the area for further work and turned the vines back on themselves.

The pruning saw was then used to sever each vine close to its root crown. When the vines were lifted up and tossed away from the area being cleared, they occasionally split at a node. When this happened, the top part of the vine peeled away leaving the rooted node intact with approximately two thirds of the vine still in place.

Within two weeks the kudzu vines on the uphill side of the Yoshinos began to reach up into the low hanging Yoshino limbs. The uphill drip line perimeters were then expanded according to the slope in order to leave a three-foot gap between the kudzu tentacles and the lowest uphill limb.

The kudzu at the curb was immature and at the far drip line, away from the street, was mature. Vines were cleared from the curb in strips of approximately 10-foot width to these drip lines but were not cleared from the areas between these strips. Clearing the vines made possible the identification and removal of the root nodes and root crowns.


Roots form at each vine node that comes in contact with the soil or leaves. These plantlets with early root formation were about 8 to 12 inches apart. They can be described as herbaceous "root nodes" or immature root crowns and can be classified into two sizes:

  • Pinhead: Little or no enlargement of the vine at the node. Roots were roughly the diameter of a hair.
  • Pea: The node had begun to expand. Roots were the diameter of a straight pin.

The pinhead and pea size nodes were often pulled up from the ground by the teeth of the pruning saw during the process of picking up the vine. When larger, the pruning saw blade was slipped under the nodes to cut the roots for easy extraction. Also, both sizes could be easily removed with fingers since the roots offered little resistance.

When these nodes were pulled up by manually grasping the node, the roots often split from the underside of the node leaving a small piece of the node still attached to the root and shaped liked an upside down shoe. These did not grow back.


Root crowns are fibrous, woody knots at or near the soil surface where new vines originate. Root crowns enlarge downward as they grow into the ground rather than upward. Some were found as much as an inch below the surface and some at about the same distance above the surface. Root crowns required different removal techniques depending on their size.

Small, marble size root crowns are present in both mature and immature kudzu. This was the most frequently encountered size and represented perhaps 35% to 45% of total root nodes and crowns removed. Extraction was quick and easy. An estimated 600 were removed during the two years. One study showed an average time of thirty seconds per extraction.

The initial removal technique was simply to scratch around the crown with the tip of the pruning saw, sever the horizontal roots, slide the tip of the saw below the exposed crown, and under-cut the woody root crown knot. The flexibility of the pruning saw was useful since it could bend around the underside of the root crown. Midway through the first year, the teeth at the tip of the pruning saw were worn down to nubs and the saw was replaced. The convenience of using only one tool for digging and cutting was well worth the small cost of another saw. However, as reported below, the larger root crowns dictated different, multi-tool approaches.

Golf ball size root crowns were frequently found at the perimeter or "front" of mature foliage. Perhaps 15% to 20% of the total extractions were this size. The same approach was used for removal as for the marble size. Extraction took longer but was usually less than two minutes each. As larger crowns were encountered, particularly during the second year, a hand trowel replaced the pruning saw for digging around the larger crowns. This enabled deeper digging, more complete crown removal, and quicker extraction.

Lemon size root crowns accounted for about 8% to 12% of the total crowns removed. Extraction was noticeably more difficult due to many large root structures with the corresponding need to dig deeper. Roots often grew to 1 inch in diameter. Extraction typically required three to five minutes or more.

Baseball size root crowns were perhaps 3% to 7% of the overall total. These were difficult to extract even after the addition of the hand trowel. Removal often took five to ten minutes and longer.

Fortunately, softball size crowns were not found often. Perhaps a dozen of the approximately 1,400 nodes and crowns removed were of this magnitude. Extraction is even more difficult than the baseball size crowns. Stewart Winslow, Horticulturalist for Milliken & Co., calls crowns of this size "kudzu stumps".

During subsequent weeks second and third passes were required in each cleared area because some root crowns were missed in the first effort.

Extractions were easier when the ground was moist. After a rain, some root crowns were pulled completely out of the ground by hand even with carrot-like roots up to 2 feet long. Crown removal was scheduled for the day after rain to the extent possible during the second year. Since most of the extractions were done from hands and knees, one disadvantage was muddy jeans.

Midway into the second year a further improvement was made through the timing of root crown extractions. Root crowns of all sizes seemed to be easier to extract a week or two after the vines were severed rather than when tackled immediately. Thereafter, extraction was systematically planned one or two weeks after the vines were cut from the crown. It is known that repeated defoliation, such as occurs in grazing, weakens the kudzu.1 Perhaps severing the vine, which might be called the ultimate in defoliation, weakens the root structure. This question is addressed in a later study.

During the latter part of the second year, it was apparent that easier extraction methods were needed for the larger root crowns. After the end of the growing season a better tool was identified and obtained. The SE-EPPC Website mentioned a tool for mechanical control called a pulaski. Local garden supply houses were not familiar with such a tool. A picture and description of its invention was found on the Washington Trails Association Website The tool was named for its creator, Ed Pulaski, a firefighter. Tired of carrying both an axe and a grubbing hoe, Pulaski combined the firemanís axe into one head with a heavy duty grubbing blade.

Since the local hardware and garden supply stores were unfamiliar with ordering the pulaski, a picture and specifications were found in a Corona Clipper product catalog at The cost was $72.44 including tax when bought through a local garden supply store. Later, the Corona web site dropped pulaskis from their offering. An alternate web site was used to purchase subsequent pulaskis. For current info on availability, see "Tools".

Evaluation showed that the pulaski was effective in removing root crowns of golf ball size and larger. The broad grubbing blade was thicker and heavier than the usual grubbing hoe. After the blade was driven under the root crown, the strong axe handle permitted substantial leverage. This action often lifted the root crown and exposed the horizontal roots for easy cutting. The axe had a steep bevel for cutting roots in soil.

One advantage of the heavy-duty nature of the pulaski was demonstrated when the same leveraging action of forcing the root crown up was attempted with an ordinary gardenerís mattock. The mattock handle cracked on the very first attempt. The pulaski was acquired after the two-year project discussed here. Hence, the time saving benefits were not reflected in the results reported in this paper.


Approximately 6 feet of flat ground lie between the curb and the Yoshino tree line. Despite occasional mowing by the Spartanburg City grounds team, kudzu vines regularly extended to the curb in spring and continued to be an eyesore through the fall.

When the author explained this project and requested that the city mow the area regularly, Mr. Douglas Jones, Superintendent of the Spartanburg City Grounds and Maintenance team, readily agreed. Fortunately, Mr. Jones had participated in planting the original Yoshinos and was quite interested in the welfare of these trees as well as techniques for controlling the kudzu.

This was a very positive example of cooperation by civic officials. Thereafter this area between the tree line and the street was mowed regularly. The time for this mowing is not included in the average of one hour of work per week that was devoted to this project.


Inevitably, during the first year, vines from the mature kudzu beyond the perimeters around the trees reinvaded the cleared areas. Similarly, in the second year, after rolling back the kudzu all along the frontage, the vines regularly returned from beyond the "front". All of these vines were either cut at the perimeter or rolled back onto the top of the mature kudzu blanket as described above. These reinvading vines had no time for rooting within the cleared area and were observed to be less and less aggressive in terms of length of new growth after each cut back and roll back. Throughout this project, the kudzu seemed to always come back weaker after each setback.


An analysis of scaffolding proved helpful in reducing the amount of maintenance time. By its nature, Kudzu builds its own scaffolding as the vines crisscross, layer upon layer. The kudzu thrived best where briars, shrubs, and trees provided a support structure that enabled significant upward growth. The taller and sturdier the scaffolding, the more the foliage had thrived and the larger the root crowns had grown.

Experience during the second year showed that removal of scaffolding reduced the kudzuís aggressiveness as indicated by fewer and weaker returning vines. Moreover, removal of the scaffolding was usually necessary in clearing these areas.

With these observations in mind, by mid 2003 the author set out to collapse all scaffolding along the 483 foot "front" of mature kudzu. Scaffolding was usually easy to detect. Peaks and mounds in the kudzu blanket provided a sure indication that scaffolding was present. Briars and honeysuckle (Lonicera japonica) provided a less conspicuous support structure than trees and bushes.

Intensive thrusts beyond the perimeter were required to collapse the scaffolding. The primary culprits included other pest plants such as Chinese privet (Ligustrum sinense), Tree-of-Heaven (Ailanthus altissima) and found on page 373 in Rayner and Porcherís book "A Guide to the Wildflowers of South Carolina", dead tree limbs, dead trees, and most difficult of all, briars.

When the scaffolding was toppled, the foliage of the kudzu often collapsed to one half or less of the original height. Immediately there were fewer and less vigorous invading vines. This process of attacking and removing the scaffolding became standard practice and was carried out two weeks before pushing back the perimeter or "front".


The primary work of reducing the kudzu eyesore during the second year was to clear it from the areas between the Yoshinos and to push back the knee high mass of kudzu. Removing the tangle of matted vines was much more difficult than peeling back individual vines.

Gradually, as productivity improved, larger and larger areas were cleared within each weekís 30 minutes or so of work allotted to rolling back the kudzu. Initially, a 3-foot long frontage might be moved back 6 inches within 30 minutes. As the techniques improved, sections twice this long were rolled back twice this depth and more.

This enabled another process change. Rather than alternating between vine roll back and crown extraction many times during the course of an hour, it was more efficient to concentrate an entire 30 minutes on one task or the other. This change increased productivity and avoided the need to shift back and forth between the specialized tools.

Vines were cleared most easily by rolling them away from the protected area. Repeated roll back of separate vines formed a rounded mass similar to a loose bale of hay. Pushing this tangle back little by little exposed the root crowns from which the vines originated. The pruning saw then cut the exposed vines close to their parent crowns. Repeated cutting of vines freed up the mass of kudzu so that it could be rolled further away from the area being protected.

Late in the second year, this roll back process was modified again. Since only the vines at the exposed root crowns were being cut, those vines that originated in the main kudzu patch were merely being rolled back. Roll back did little to damage the main patch other than to shade the existing foliage due to the layering of rolled back vines on top of the patch.

A more effective process proved to be cutting a swath through the kudzu foliage parallel to the previously cleared area and a yard or so forward of and into the front. The mass of cut kudzu was then rolled toward the cleared area instead of away. By severing all vines at the swath, the root crowns in the remaining kudzu were dealt a setback.


Vines were hanging down into four of the Yoshinos from the canopy of kudzu atop other trees that stood approximately 20 feet behind the Yoshinos. These descending vines were the most threatening of all to the Yoshinos. The roots lay outside of the cleared and de-crowned frontage perimeter, and the vines enjoyed the ultimate scaffolding — trees.

Due to the threat of over-hanging vines to the Yoshinos, this was first chronological action of the entire project. It is reported late among the activities because, while the easiest to handle, it was only a temporary solution. The largest, semi-woody vines going up into these trees were cut with the pruning saw. Several were 1 inch or more in diameter. Only a few minutes was needed to cut these biggest vines. No attempt was made to cut the dozens of smaller vines that were climbing into these trees. A drastic effect on the kudzu canopy was clearly visible within three days after cutting the largest vines. This large, very obvious setback to the kudzu was very satisfying. The overhead canopy of kudzu and the descending vines returned twice each year after the first cutting. No permanent solution had been achieved at the end of the second-year growing season.


Seedlings were easily differentiated from root crown sprouts. Sprigs from seeds had no root crown. Instead, the seedlings had fewer, smaller, weak roots that permitted finger removal much like pulling up a weed from a garden. Very few seedlings — perhaps a dozen — were observed during this two-year project. Seeds were not the reason for sprouts reemerging in areas from which root crowns had been extracted.

No flowers appeared in the kudzu front that was attacked when this work began. Thus, seeds were not expected. In the kudzu patch beyond the cleared perimeters, flowers were present where undisturbed vines were cascading down from tall scaffolding. All kudzu along this street frontage was disturbed, including the vines hanging down from the trees above. Seeds were not expected or found in the areas cleared during these two years.


After ruling out sprout reemergence from seeds, further examination showed that when the pruning blade sliced through the middle or the bottom of the root crown rather than through the taproot, sprouts came back. Leaving just 10% of the root crown was sufficient to produce resprouting. This realization led to a switch to a hand trowel in order to dig deeper and to remove the larger crowns completely.

This observation may be germane to the question of how deep the bulldozer blade should cut into the earth when an area is being cleared of kudzu. Sprouts will definitely return where the dozer fails to cut below the depth of the largest root crowns.

It was difficult to determine whether sprigs reemerged from locations where the root crowns were completely removed. From the standpoint of the project objectives, this did not matter because regular follow up ensured that the few resprouts were removed. However, the question was worth study, "Do sprouts return from roots after the crown has been completely removed?" In order to address this issue, an additional study was necessary.


During the second year, a plot of mostly immature kudzu at a different location was used during to determine whether sprouts and vines would reappear after 100% removal of the root crown. Since no mowing was involved, construction flags could be used to identify the spot from which each root node and root crown was removed. Virtually all complete extractions were effective to the extent that new sprouts did not emerge during the same year. These observations may not be applicable to an area of mature kudzu and large root crowns or to subsequent year reemergence. The flags remain in place in order to identify any sprouts that might return in 2004. During the winter of 2003 a plot of mature kudzu was cleared of root crowns and similarly marked with construction flags to determine whether sprouts return or not.


A dozen or so large, horizontal vines up to 2 inches in diameter were observed to connect larger root crowns. More often than not, these connecting roots were dead. Even when not living, it was necessary to sever these horizontal roots with a pruning saw before the root crowns could be removed.

When roots were cut with the pruning saw, the cut was often distinctly not clean, leaving a frizzy, fibrous profusion at the cut. Since no sprouts were observed to return after complete root crown extractions were made, it was undetermined as to whether the clean cut damaged the kudzu more or less than the rough, fibrous cut.

Several obstacles were encountered including briars, ants, poison ivy, and a widespread mindset among friends that the only way to control kudzu is through chemical methods. The literature offered little help. As Dr. J. H. Miller observed in a private communication, " . . . very little to nothing has been written detailing manual removal techniques . . . " 2


During the first year, one hour per week during the growing season was sufficient for removing the kudzu vines, root nodes and root crowns from the expanded drip lines around the decorative trees, and for protecting them from the vines cascading down from overhanging trees.

At the end of the first year, kudzu vines still extended to the street on the ground between the Yoshino trees. This left an unsightly, scalloped view for the traveler who saw alternating areas of kudzu, no kudzu, kudzu, no kudzu.

The project expanded during the second year to include reducing the kudzu eyesore at the street, pushing back the kudzu front, improving techniques, and the reducing the maintenance time. With the help of city mowing of the curbside section, the cleared areas were maintained and enlarged. Virtually all techniques were improved. All vines and root crowns were removed between the curb and the line of decorative trees. The kudzu front was rolled back approximately 7 feet behind the Yoshino tree line along the entire highway frontage. The 483 feet of frontage was clear of kudzu for an average depth of 13 feet. Roughly 6,200 square feet, or one seventh of an acre, was free and clear of kudzu.


Decorative areas and highway frontages with 482 feet of frontage exposure to kudzu can be protected without application of herbicides by an expenditure of one hour per week during the growing season using manual and mechanical methods.

With perseverance one hour per week of time permits pushing the kudzu back 7 feet in the second year. Rolling vines back, extracting complete root crowns, regular mowing, collapsing scaffolding, and using the described techniques are effective in mechanically removing kudzu from around the Yoshino trees as well as in minimizing the maintenance needed to protect against future kudzu reentry.


1Miller, J. H.; Everest, J. W.; Ball, D. M.; Patterson, M. G., 1991. Kudzu in Alabama, History, Uses, & Control,

2Personal communication on 7 October 2003 from Dr. J. H. Miller, Research Ecologist, USDA Forest Service, Auburn, Alabama.

SUPPORT FROM OTHERS: The author very much appreciates the generous help and encouragement from Mr. Don Caldwell, CEO of Grace Management Co. the property owner; Mr. Douglas Jones, City of Spartanburg; Mr. Henry Pittman, Director of The Spartanburg Menís Garden Club; and Dr. J. H. Miller, US Department of Agriculture.

ABOUT THE AUTHOR: While the writer is not a professional horticulturist, Hardie is a certified Master Gardner, member of the Spartanburg Menís Garden Club and active participant in civic and church organizations in Spartanburg, S. C. He holds a Masters Degree in Industrial and Systems Engineering from Georgia Tech. Hardie is a former marathoner and is retired from Milliken & Company after 41 years in various management positions.

The author is interested in communicating with individuals and groups who have experience or interest in mechanical methods for kudzu control. The author may be contacted by email: or mail at 128 Bagwell Farm Road, Spartanburg, SC, 29302 or by telephone at (864) 582-0990 or through the web site