Herbicide-free weed control
by Dr Neil Hipps, Horticulture Week, 14th February 2008

Is it possible to suppress weeds successfully without the use of herbicides?
A look at alternative methods used across Europe Controlling weeds on hard surfaces including roads, pavements and playgrounds is a major task that is completely dependent on the use of herbicides in the UK.

Dr Neil Hipps: Little if any resources are being given to alternative control methods. Yet in Denmark and Sweden, many municipalities do not permit the use of herbicides and alternatives such as flame, steam and hot foam are regularly used to control weeds.
Herbicide use is also restricted in Germany and the Netherlands. Research in the Netherlands has shown that the use of pesticides (including herbicides) on hard surfaces contributes to exceeding EU drinking water safety standards in surface water. This is due to direct run-off, little breakdown during the passage through the sewage systems and in the surface water itself.
Surprisingly, only Denmark and the Netherlands have current data on the pesticide and herbicide use specifically in urban areas. However, the Pesticides Safety Directorate is currently funding a survey of amenity horticultural usage in the UK, so data should be available later this year.
With trends towards greater environmental sustainability, including reduced use of agrochemicals, it is uncertain for how long the current UK position will remain acceptable or permissible.
To tackle this problem, East Malling Research has participated in international project Interreg IIIC, funded by the European Regional Development Fund, to consider the issues relating to reduced herbicide use on hard surfaces. Other UK participants included Warwick HRI, Mid Sussex District Council and Warwick County Council.
The project involved a mix of 18 research and municipalities (practitioner) organisations from seven northern European countries.
The purpose of the project was to consider how to minimise the use of herbicides for weed control in urban areas. A project website provides a comprehensive overview of the results from the project and details of its many partners.
A weed survey was carried out on pavements in several towns across northern Europe, which identified 107 different weed species. However, the most prevalent weeds were grasses – mainly annual meadow grass (Poa pratensis), mosses, procumbent pearlwort (Sagina procumbens), knot grass (Polygonum aviculare), Canadian fleabane (Conyza canadensis) and dandelions (Taraxacum officinale).
Persistent weeds, such as annual meadow grass, are well-adapted to urban situations as they tolerate soil compaction, are resistant to pedestrians, have protected growing points and produce large quantities of seed that can germinate from March until October, allowing several generations per year. Procumbent pearlwort spreads by producing both seeds and rooting stems. More specialist and rarer weeds were associated with specific situations - for example, rosebay willowherb (Epilobum angustifolium) was strongly associated with sunny spots.
The amounts of weeds present in the different towns was variable, but generally surface coverage was below five per cent. As might be expected, town centres tended to have the lowest amounts and the industrial estates the greatest. Factors that influenced the amount of weeds included adjacent vegetation, shading, footfall and the area of bare soil. Pavements with cobbles or that have been installed in sets have a much greater area of soil that can be invaded by weeds than those made of large slabs or asphalt. One noticeable difference between the UK and other countries was the preference for the use of asphalt for pavements in residential areas. While this is aesthetically unattractive, from the perspective of reducing weed invasion it is more effective.
The project also considered alternative control methods to the use of herbicides. Some of this work was carried out by Palle Kristoffersen in Denmark. He suggested that weed levels must be kept below two per cent of the total ground area for aesthetic acceptability. He compared the use of different thermal methods of control (flame, hot water, steam and hot air) with the use of herbicide (glyphosate) and found that all of these methods could be used successfully to control weeds to his prescribed two per cent level.
However, there were differences in the number of applications of the treatments. While herbicide application was required between two and three times, thermal treatments were needed between two and seven times to achieve the same result. The most efficient non-chemical method was hot water and the least hot air. This work is now being developed to determine the energy-efficiency of the different control methods.
Corne Kempenaar at Wageningen University in the Netherlands has investigated not only thermal control, but also mechanical methods, ie brushing and scraping. He has used a descriptive set of quality standards, which grades weediness into five different classes. While Class 1 is clean, Class 5 signifies a large amount of weed growth (a combination of both height and spread).
The advantage of this system is that it allows him to compare different levels of acceptability and to see what needs to be done to increase acceptability without necessarily eradicating all weeds. His results broadly agree with those from Denmark, meaning herbicide requires fewer applications than any of the other techniques. Thermal methods require four to five applications and brushing requires five to six applications to give reasonable control of weeds in the pavement joints. Dr Arnd Verschwele of the Institute for Weed Research, Germany, carried out a separate trial over two seasons, and also found that the hot-foam and hot-water techniques were more effective than brushing and flaming.
Only five brush applications per season were needed to maintain an almost weed-free pavement, but to improve from Class 3 to Class 1 needed seven applications. Brushing the weedier situation (Class 3) only five times maintained weediness at the same level. Care has to be taken with brushing as some of the more vigorous steel brushes are not appropriate for all surfaces and could cause damage. Good guards are needed in close proximity to the brushes to prevent small stones flicking out and injuring pedestrians or damaging parked vehicles. Obviously, safety is also a potential issue with flame weeding techniques.
Kempenaar has considered the cost implications of the alternative control methods, taking into account the amounts of labour involved. To achieve very little weed growth by brushing costs EUR0.20 to EUR0.40, by flaming EUR0.15 to EUR0.35, hot water EUR0.30 to EUR0.40 and herbicide EUR0.07 to EUR0.10 per square metre. A lower level of weed control causes a small reduction of costs.
A sustainable weed management system that minimises, but does not eliminate, herbicide use has been developed by the University of Wageningen. This provides guidelines and certification for Dutch contractors, which decrease the potential of polluting water with herbicides. The principle is to target herbicides accurately, only apply when necessary and consider weather conditions in advance and integrate herbicide application with a good brushing regime.
Several technologies are available to apply herbicide more accurately. Sprayers using optic sensors to detect weeds on the ground surface to precisely target weeds or individual plants are already available. Similar technology has also been developed for hot-water applications.
Alternatively, contact methods that don't involve the use of spraying also exist. One example, developed in Germany, uses a hand-pushed trolley that has an adjustable rotating drum. The drum brushes directly against the leaf surface, so herbicide is delivered only directly onto the plant. Low-volume spraying (controlled droplet application) that reduces run-off potential is well developed in the UK. It is known to be a very efficient method for glyphosate application.
A life-cycle analysis that considers many environmental parameters including pollution, climate change and ecotoxicity was used by the University of Wageningen to compare the effects of various weed-control methods, including their sustainable management systems. It showed that the reduced herbicide input methods can produce overall impacts no more deleterious than flaming and hot-water control, although the contributory factors are obviously different.
The project has quite clearly demonstrated that adequate control of weeds on hard surfaces can be achieved without use of herbicides. Much of the new non-herbicide control technology is being developed in Denmark and the Netherlands to satisfy local need, but could be used in the UK, where similar weeds grow. More research and development will be required to improve the efficiencies of the non-chemical technologies to comparable levels to herbicides.
In future, more thought could be given also to landscape design targeted at preventing the growth of weeds. This will require combining knowledge of weed biology with the development of appropriate construction techniques. Particular attention needs to be given to the materials used in joints between surface blocks, slabs and setts.