Return to site

What the knotweed paper really says

Addressing inaccurate reporting of our research

In light of the recent press coverage about the publication of our knotweed paper, Optimising physiochemical control of invasive Japanese knotweed, this article is written to remove any ambiguity about the intent of our research and the main findings.

Let’s re-examine the headlines about the knotweed paper:

The Telegraph:
"Japanese knotweed cannot be cured, major study which tried 19 methods finds"

The BBC:

"Knotweed ‘cannot be eradicated’, Swansea Uni trial shows"

The Daily Mail:

"Japanese knotweed ‘cannot be killed off’: Scientists fail to destroy the invasive weed after 19 attempts"

Our first thoughts? These headlines are all a fundamentally inaccurate characterisation of our research.

All three headlines are also overwhelmingly negative and this obfuscates the valuable findings of the work  —  that of establishing a scientifically credible baseline of research for some of the main treatments for Japanese knotweed in the UK, that up to this point lacked evidenced-based, multi-year research.

A cure for knotweed?!

Firstly, The Telegraph.

Nowhere are we saying that Japanese knotweed is something to be cured or needs to be cured of something. You might ask, cured of what?!

Japanese knotweed is a plant, not a disease.

Knotweed is not a ‘deadly plant’ that is toxic to humans like Foxgloves; nor, is it inherently hazardous in the manner of Giant Hogweed, the sap of which can cause severe burns.

Japanese knotweed is in fact edible and a good source of resveratrol, although harvesting in the UK is not recommended, on account of potential heavy metal accumulation or prior treatment with herbicides, as we've written about here.


Secondly, and more seriously,  the BBC and The Daily Mail - does the paper say that knotweed cannot be eradicated in any circumstances?

No, it does not.

The paper says something subtly, yet importantly different; of the 19 methods of physiochemical control of tested, none resulted in complete eradication of Japanese knotweed (after 2 years of treatment).

A central point of the paper is that full mechanical excavation with off-site disposal (known as ‘dig and dump’), and also geomembrane encapsulation, were not examined on account of prohibitive costs and practicality.

The above two methods of physical remediation, favoured by many contractors for their rapid turnaround of knotweed infested sites, are widely used and generally regarded as effective measures of control. They are also by far the most expensive treatments for knotweed  —  often by an order of magnitude more  —  compared to control using herbicides.


Dig and dump in particular is often prohibitively expensive due to the large amounts of contaminated soil that must be disposed off in specially licensed landfill sites. For both homeowners and large landowners alike, dig and dump is often too costly to be considered as a viable option for remediation.

Notwithstanding costs, the large volumes of soil that need to be excavated also rule out dig and dump where it is too destructive, or where ground disturbance presents a safety hazard — say alongside a railway or reservoir. And lastly; many sites are inaccessible by heavy mechanical equipment needed to dig out soil or install geomembranes in the first place.

Some contractors claim they achieve total eradication but this is an extremely confused position — so much so that it needs a separate article.

Eradication is an unhelpful simplification of what is really a process better understood as achieving and subsequently maintaining effective control of Japanese knotweed, irrespective of the remediation methods used.

So dig and dump and geomembrane encapsulation are useful, but mostly on development sites where the downstream costs of delaying a commercial project justify the high initial cost of these methods of knotweed control.

The logic of the research focus of the paper is to examine the methods of knotweed control that are more widely applicable — chiefly methods centred on glyphosate-based herbicide application.

What was tested?

The focus of the paper is on herbicide based methods of knotweed control. Additionally, one practical physical treatment method (covering) and also some integrated physiochemical methods of knotweed control were tested , these were:

  • Aboveground covering with hand-pulling.
  • Cutting with herbicide treatment.
  • Rhizome disturbance/tillage using a mechanical excavator, followed by herbicide application. 

The last method, in particular is being confused for dig and dump which also uses mechanical excavation.

In total 19 different treatment protocols, across a range of methods, were tested concurrently. Experiments lasted three years; with one year for site preparation, and two years of treatments, followed by evaluation. From beginning of the research to final publication of the paper, the project lasted seven years in total.

Note how the Daily Mail completely mis-represents testing 19 different treatments as 19 successive treatments on the same area of knotweed.

What does the paper actually say?

There are seven main findings, none of which are accurately represented in any of the headlines. Broadly speaking they are:

  • Finding 1 - none of the 19 treatments we tested achieved total eradication.
  • Finding 2  -  of the all 19 treatments tested (across physical, chemical and integrated methods), glyphosate-based treatments are by far the most effective. The non-glyphosate based herbicides tested did not control knotweed more effectively.
  • Finding 3 - two foliar applications and one stem injection application of glyphosate-based herbicide (over successive years) were significantly more effective that all other treatments, though a single foliar application of glyphosate-based herbicide came in as a close 3rd place. All three treatments were significantly better than all of the other treatments, except one based on picloram (which is now banned in the EU). Note that stem injection used a lot more glyphosate-based herbicide than treatments using foliar spray application.
  • Finding 4 -  aboveground physical covering with hand-pulling was not an effective method of knotweed control. 
  • Finding 5 - cutting aboveground knotweed growth before applying glyphosate-based herbicide, either as a spray or by pouring it into the knotweed stem, reduced the effectiveness of control. This is because the leaves act as a pump, pushing the glyphosate into the rhizome  later in the growing season — cutting off the pump reduces the amount of herbicide getting into the rhizome. 
  • Finding 6  - applying glyphosate-based herbicides to knotweed more than twice a year is not necessary, and very early treatments may interfere with the effectiveness of later treatments in the growing season by reducing the ability of the knotweed plant to translocate herbicide from the leaf surface into the rhizome.
  • Finding 7 - don't apply huge doses of glyphosate-based herbicide to knotweed, as beyond a threshold dose, control does not improve. This is because translocation of glyphosate in the rhizome is limited — you will simply be wasting herbicide and potentially, damaging the environment.  

A key suggestion of the paper is also that very time consuming physical methods such as hand-pulling, hand-digging, strimming and cutting are not practical or effective methods of control, and should be discontinued in almost all circumstances. This is because exhausting the rhizome by removing aboveground knotweed growth could take decades. Also this increases the risk of spreading the plant around accidentally.

What next?

The practical outcomes of the knotweed paper are expected to be:

  • A reduction in the inappropriate use of non-glyphosate based herbicides to control knotweed.
  • Better seasonal targeting of glyphosate-based treatments, resulting in both better treatment outcomes and also lower amounts of glyphosate being applied over the treatment lifecycle. 
  • More sparing use of stem injection in recognition of its much higher use of glyphosate compared to foliar spray for knotweed control.
  • A reduction in inadvertent knotweed spread from physical cutting, strimming, and hand-pulling methods, which we advise against.

All of the above should be welcomed by contractors, academics, environmentalists and homeowners alike.

Firstly, our research helps to reduce the uncertainty and therefore the risks and costs of knotweed remediation. Secondly, it promotes effective knotweed control whilst reducing unnecessary herbicide use.

The knotweed paper caps a huge research effort that places the UK and South Wales in particular at the forefront of serious research into Japanese knotweed control - something we, and all of the co-authors of the paper are extremely proud of.