WEED CONTROL USE OF HERBICIDES ON RAILWAY TRACKS IN SWEDEN Lennart Torstensson of the Swedish University of Agricultural Sciences in Uppsala outlines the methods used to clear weeds along railway tracks in Sweden Another important reason for removal of weeds from railway tracks is the working environment for track maintenance employees. The risk of accidents caused by stumbling or slipping on weeds is obvious and may have fatal consequences. A further reason for weed removal is the risk of train wheels skidding causing longer acceleration distances and more seriously extended braking distances which could result in the train being unable to stop at signals. If weeds on the railway embankment become dry during the summer there is always a risk of fire caused by sparks from the wheels.In addition the quality of wooden sleepers is rapidly destroyed when they are lying in polluted watersaturated ballast. Introduction In Sweden removal of weeds along railway lines has been carried out for more than 50 years by application of herbicides. The Swedish national railway network consists in total of about 12,000 km of which 25–30% is treated annually. A specially equipped spray train (Figure 1) is used by trained staff for the purpose. Theoretically there are several alternative methods of removing weeds from tracks i.e. chemical mechanical or thermal. Methods using hot water and steam have been tested but only delayed growth effects have been achieved. Weeds can also be controlled for short periods using open flames but this method is not possible on all types of track because of the fire-sensitive materials used in signal systems.There is also the risk of setting fire to dry vegetation along the tracks. Other approaches for weed management on embankments are use of impenetrable plastic layers along the sides of the embankments and strategies for establishment of weed competitive vegetation along the base of the establishments. The general opinion in Sweden is against use of pesticides and there have been suggestions that weeds on railway tracks could be removed manually. Manual removal would be possible on railway embankments where the ballast consists of gravel. However apart from the manpower required it is a dangerous operation.Weed removal has to be carried out on tracks without disrupting the traffic. Therefore the method used must have a high capacity. Because of this requirement it is difficult to find viable alternatives to chemicals. The Swedish National Rail Administration (BV) has found that weed removal by chemical means is the only practical and economically realistic method and overall is also the least risky method. Why remove weeds? One of the most important reasons for removal of weeds from railway embankments is to retain the quality of the track and to ensure safe railway traffic. The macadam or gravel which are components of the ballast should not become polluted by weeds or decaying plant residues. If this happens the pollutants will fill up the hollow spaces within the ballast resulting in the binding of water.During the winter this water will freeze to ice and swell thereby causing dislocation of the track. Pest ic ide Outlook – Fe b r u a r y 2001 This journal is © The Royal Society of Chemistry 2001 Environmental concerns To obtain good results from herbicide application on railway embankments requires a thorough knowledge of herbicide behaviour and persistence on and in the embankment. By good results is meant acceptable weed control but also the absence of side-effects such as injury to employees or damage to equipment or to the environment along the tracks. Chemical physical and toxicological properties of herbicides approved by the Swedish Chemicals Inspectorate (KemI) are well known but not always as applied to the environment of railway embankments which are characterized by very low contents of clay and organic matter and low biological activity.Herbicides are mainly absorbed to clay and organic matter. If these are low implying a low adsorption capacity the risk of transport out of the target area is obvious. A low content of organic matter normally means low microbial activity. Since the activity of microorganisms is the main factor controlling decomposition of herbicides in soil this results in slow rates of transformation and consequently long persistence times. This means that when a herbicide is to be chosen for use on a railway track it is not enough simply to be well informed about any effects on weeds appearing on the embankment.It is also necessary to know the binding mobility and decomposition of the substance in material from the embankment to be sure that there will be no risks to the surrounding environment. Today BV is investing in far-reaching experiments to identify effective herbicides and methods of application that reduce to a minimum the risk to the surrounding environ- 16 DOI 10.1039/b100802l Figure 1. The train used for full scale spraying operations. The train consists of an engine a staff carriage also containing equipment for recording all data concerning the spraying including GPS and finally a wagon with the spraying equipment and tanks for water supply. Spray booms are mounted on both ends of the wagon.This makes it possible to drive the wagon in either direction spraying only with the ramp at the back of the wagon. ment to BV employees involved in application or to other employees working along the track. A discussion of some of this research is given below. Herbicide testing Studies of herbicides that might be useful on railway tracks can include several steps such as l Gathering relevant information concerning weed effects appearance in the environment through mobility persistence and side-effects as well as methods for analysis of the compound and its metabolites. It is understood that information concerning human toxicity has been or will be evaluated by KemI. l Laboratory studies concerning adsorption to and desorption from material typical for railway embankments in Sweden as well as studies of decomposition in the sa m e material.The aim is collection of data showing kinetics of degradation and influence on rate of degradation of environmental factors. Laboratory studies may also include setting up suitable methods for analysis of the compound and its metabolites in soil and water. This has recently been done for diflufenican (Wennberg and Torstensson 1997) as well as for glyphosate and AMPA (Börjesson and Torstensson 2000). l Minor and full-scale field studies of the herbicide including – weed control effects. – studies of appearance of the compound and its metabolites in the embankment e.g. mobility and persistence. Minor field studies Minor field studies can be carried out on tracks heavily infested by weeds.Test plots are 25 m long with four replicates. To reduce the total length of the experimental WEED CONTROL distance the plots are divided along the middle of the track. The area of a plot is then (3 � 25=) 75 m2. Pesticides are applied using a 3 m boom with 6 VEEJET (11010) nozzles driven by compressed air and a spray volume counted after what is expected to be the volume in a full scale operation.The equipment is mounted on a trolley and speed regulation is done by an MTR 101 engine trolley or similar (Figure 2). Each experiment should include control plots not sprayed between each 5th test plot or at least 4 control plots per experiment. Minor field studies should be repeated at 3–10 geographically different places representing different climatic zones different weed populations and different types of embankments with variations in particle size pH etc.Evaluation of herbicide effects on the weed can be done by grading in a scale from 0 to 5 where 0 is as untreated control and 5 is all vegetation dead. The scale is relatively crude but handy for practical field work. By repeating the estimations at least 2–3 times during the growing period it gives a fairly good comparison oweed effects of different herbicides or dosages. In the practical evaluation half grades can be used. The same method can be used in both minor field studies and in full scale field studies. For studies of appearance of the compound or its metabolites in the embankment samples are taken from a randomly chosen area of 25 � 40 cm within the experimental plot.The uppermost layer (0–10 cm) is sampled by cutting a cubic sample of an area of 9 � 9 cm and 10 cm depth. After that the whole 10 cm layer within the sample area (25 � 40 cm) is removed. Then the procedure is repeated for each of the remaining layers to be sampled. Full-scale field studies Full scale field studies are carried out to verify if a herbicide tested and found useful in minor field studies can also be used on an operative scale. The train used for full scale spraying operations can be seen in Figure 1 with a closer view of the spray boom in Figure 3. The normal speed of operation is 25 km h–1 with the potential for higher speeds.Instead of spraying the herbicide may be wiped onto the weeds using equipment like that illustrated in Figure 4. The Figure 2. Equipment for spraying experimental plots with herbicide in minor field studies. 1 7 Pes ti cide Out look – Fe b r u a ry 2001 Weeds on railway embankments WEED CONTROL Figure 3. Detailed view of one of the spray booms in spray position. The boom can be moved more or less outside the wagon. The speed wind is used to press down the spray liquid against the ballast surface. The boom is equipped with individually fed triple nozzle bodies with diaphragm check valves for drip free shut off. Latin name Figure 4. Equipment specially designed for wiping herbicides on weeds on railway embankments. The wiping unit mixes the concentrated herbicide liquid to the desired concentration and then distributes it to the different ramp sections with its wiping fingers as can be seen in the figure.wiping technique is founded on contact application of herbicide directly on aerial parts of the weed. This type of application eliminates the risk of wind drift. It also reduces the amount of herbicide used since it is only applied on existing weeds. Wiping with this equipment is particularly useful where wind drift has to be avoided or where the application rate has to be as low as possible. However the speed of operation is low 5–8 km h–1. Evaluation of weed control effects as well as studies of the compound’s appearance in the embankment are monitored. Possible transport down to the watertable is studied through sampling in groundwater pipes of the kind described by Börjesson and Torstensson (2000).These are iron standpipes 1.5 m in length and 6 cm in diameter excavated down into the railway embankment as close to the tracks as Pest ic ide Outlook – Fe b r u a r y 2001 18 possible. The upper part of the pipes must not be an impediment for vehicles of any kind in traffic on the tracks. A layer of bentonite is put around the pipes to avoid mechanical transport of contaminated soil particles and surface water along the pipe walls. A layer of gravel filling is put around the bottom of the pipes to allow ground-water to percolate into them. At each sampling sits normally three groundwater pipes are placed in a row with about 10 m spacings.When choosing a herbicide and strategy for weed removal it is important to know what kind of weeds are to be removed. Inventories have been made showing that a large number of plant species both herbaceous and woody are found on the tracks. Examples of frequently appearing species are given in Table 1. Some weed species have caused particular problems on railway embankments. G. boreale and G. verum seem to have become resistent to diuron after it has been used for about 15 years. Glyphosate has no effect on the very common weed Equisetum arvense resulting in an increase of that species after some years of glyphosate application (Figure 5). Table 1. Examples of weed species found along the Swedish railway network.F = frequent VF = very frequent Frequency English name Achillea millefolium Artemisia vulgaris Campanula rotundifolia Equisetum arvense Galium boreale G. verum Gramineae (several spp.) Lapsana communis Linaria vulgaris Melilotus alba Plantago major Ranunculus repens Senecio vulgaris Taraxacum officinale Populus tremula Betula (several species) Picea abies Pinus silvestris Rubus idaeus Salix (several species) VF F F VF VF VF VF F F VF VF F F VF F VF F F F F Yarrow Common mugwort Harebell Common horsetail Northern bedstraw Ladies bedstraw Grass Nipplewort Toadflax Melilot Waybread Buttercup Groundsel Dandelion Aspen Birch Norway spruce Scots pine Raspberry Willow Herbicides Many different herbicides have been used on railway tracks.Active substances used before 1970 include amitrol bromacil diuron monuron and sodium chlorate. From 1974 only diuron formulated as Karmex 80 (800 g a.i. kg–1) was used. However after some years of application in certain areas damage to Scots pine was observed along tracks where diuron had been applied (Torstensson 1983 1985, WEED CONTROL identified in needles and branches of dead and damaged trees. Diuron has also been found to have a very long persistence several years in a railway embankment and in 1993 it was banned for use on Swedish railway tracks. A number of herbicides have been tested for use on Swedish railways.The tests have first identified the weed control effect. If that has been acceptable further tests have investigated mobility and persistence in the embankment. Some herbicides have passed these tests but with other considerations taken into account by BV or KemI few of them have been accepted for practical use. From 1986 glyphosate has been used for weeding of tracks. Since glyphosate has no effect on the very frequent weed common horsetail (E. arvense) that plant has increased in significance on Swedish tracks (Figure 5). Imazapyr provides good control of common horsetail and has been used since 1995. The only two herbicides registered by KemI for use on tracks to-day are glyphosate formulated as Roundup Bio (360 g a.e. l–1) and imazapyr Arsenal 250 (250 g a.i.l–1). The application rates are generally 5 l ha–1 of glyphosate or when there are problems with common horsetail a mixture of 3 l ha–1 glyphosate plus 2 l ha–1 imazapyr the first year followed by 2 l ha–1 imazapyr alone the second year. Figure 5. Common horsetail (Equisetum arvense) growing along railway tracks previously treated repeatedly with glyphosate. 1994). The explanation was that diuron had been transported downward within the embankment and at deeper levels come in contact with roots of Scots pine trees penetrating the embankments. Residues of diuron have been None glyphosate glyphosate 5 l ha–1 5 l ha–1 0 l ha–1 None imazapyr glyphosate + imazapyr 3 + 2 l ha–1 2 l ha–1 Effect on weeds The effects on weeds of the three herbicides used in practical application on railway embankments during the past 10–20 years have been tested in a number of minor field studies all over the country.The weed control effects of the herbicides vary of course depending on geographical site type of vegetation and yearly variations in climatic conditions. If one looks at the mean values for a number of experiments however the differences between the application strategies appear as shown in Table 2. Diuron gave an acceptable effect initially but during its last years of application the effects on certain weeds like G. apparine and G. verum seemed to disappear. Glyphosate has a good weed control effect in most cases on the condition 13 12 10 Table 2.Weed control effects of glyphosate* and imazapyr* on railway embankments after different spraying strategies. SD = standard deviation. N = number of field experiments N Weed control effect ±SD 3rd year 2nd year Treatment 1st year 3.1 ± 0.6 3.8 &pl0.5** 3.0 ± 0.7** 3.7 ± 0.6 4.6 ± 0.4 4.2 ± 0.8 0 l ha–1 15 11 10 * glyphosate formulated as Roundup Bio and imazapyr formulated as Arsenal. ** On condition that common horsetail is rare on the embankment. If it is frequent the effect will go down rapidly. 1 9 Pes ti cide Out look – Fe b r u a ry 2001 WEED CONTROL HERBICIDES ON RAILWAY TRACKS IN THE UK Maintenance of railway tracks in the UK is the responsibility of Railtrack PLC as owners of the 20,000 miles of track in the UK.Since the unchecked growth of vegetation can reduce the operational safety of the line and the life of the track most of the open track is sprayed annually with herbicide by contractors to facilitate maintenance. The aim is to control all weeds on the tracks and injurious or invasive weeds on the trackside. Over the last few years Railtrack has developed a best practice manual for vegetation management based on case studies and increasingly requires its contractors to give careful consideration to sensitive methods of controlling vegetation. In consultation with their contractors Serco Railtest Ltd. and the Environment Agency guidelines have been developed to ensure that control of vegetation is carried out on the basis of an evaluation of the risk to railway operations and the environment.Herbicides used Railtrack uses herbicides that work as selectively as possible. Three main herbicides are in current use l l l Roundup (glyphosate) – a foliar-acting contact herbicide Freeway (diuron) – a residual herbicide which remains in the soil killing seeds as they germinate Garlon 4 (triclopyr) – a selective herbicide which is specifically designed to kill broadleaf weeds and woody scrub growth These herbicides are normally applied using a fleet of specially designed Multi Purpose Vehicles (MPVs) fitted with spraying arms backed up by back-pack spraying in stations and sensitive locations. These MPVs are also used to apply anti-skid Sandite coating and water-jet conditioning of track. Environmental protection Railtrack works with the Environment Agency (EA) the Scottish Environment Protection Agency (SEPA) English Nature Scottish Natural Heritage Countryside Commission for Wales and the water companies to agree a list of sites for protection against potentially damaging effects of herbicide spraying.In these areas for example near water courses near abstraction sites for drinking water or where the ecology needs special protection the types of herbicides used by contractors are restricted. In 1999 728 such sites were agreed with these parties. Railtrack is also collaborating with the EA to carry out research into the leaching of herbicides. To minimise the effects on ecology all Railtrack zones have locally held databases. In all these databases contain over 3000 listed assets including sites of conservation interest and protected trees.Efforts are also made to protect and preserve wildlife wherever possible although in certain circumstances there is of course a need to control vermin. The following are two examples of ways in which Railtrack are seeking to protect the environment within its track maintenance programme. Protecting the sand lizard Railtrack is tailoring lineside works to help preserve the rail corridor as a linking habitat to Europe’s most northerly colony of the rare sand lizard (Lacerata agilis) on the Sefton coast on Merseyside. This is thanks to a joint initiative with English Nature the Life Fund Project Sefton Metropolitan Council and Jarvis Facilities Management on the Merseyrail Northern Line.Creating green corridors Within the Capital Greenlinks Lineside Regeneration Projects in South London Railtrack is improving the management and diversity of rail lines and linking public spaces with green corridors. At Wandsworth a seamless parkland landscape between the Common and the railway is being created with improvements to a mile and a half of rail corridor. Biodiversity Action Plans Railtrack has worked recently with the Wildlife Trusts and English Nature to develop a biodiversity action plan for their East Anglia Zone. The plan made available to contractors identifies the location of protected sites or species on the railway infrastructure on best practice for their management. For further information on Railtrack see http://www.railtrack.co.uk Pesti c ide Outlook – Fe b r u a r y 2001 20 that common horsetail is rare on the embankment.I f i t occurs a few years of treatment with glyphosate may render it dominant (Figure 5) but in such cases a mixture of glyphosate and imazapyr gives a good effect especially if followed the next year with a treatment of imazapyr alone. Normally there is then no need of any herbicide during the third year in southern and central Sweden. In the north of Sweden there may be no need for herbicide during the fourth year. Figure 7. Half-lives of diuron glyphosate and imazapyr in railway embankments. SD = standard deviation. Number of field experiments glyphosate 10 imazapyr 20 diuron 10. glyphosate 10 imazapyr 20 diuron 10. Mobility and decomposition Soil factors and climatic conditions influence mobility and decomposition of the herbicides.Diuron was very mobile in the railway embankment (Figure 6). All the herbicides studied move downward through the uppermost 20–30 cm of the embankment the part built up of macadam or gravel. This is probably a result of mechanical downward particle transport caused by the train traffic and transport in solution due to rain. The particles carry the applied herbicides adsorbed to them. With regard to glyphosate the main proportion of the herbicide is found in the uppermost 20 cm although amounts above the detection limit may be found down to a depth of 50–60 cm on certain sites. The main proportion of imazapyr is found in the upper 30 cm but amounts above the detection limit have been found down to a depth of 60–80 cm on several sites (Figure 6).Figure 6. Mobility of diuron glyphosate and imazapyr in railway embankments. Number of field experiments Diuron has a very long persistence in railway embankments (Figure 7). In spite of the fact that it has not been used since 1993 diuron and its metabolite demethylated diuron are still detectable in the embankments. Glyphosate and imazapyr have considerably shorter half-lives than diuron. Minor amounts of them however as well as of the glyphosate metabolite AMPA may be found a long time (1–2 years) after application. At sampling of groundwater pipes along tracks transport down to the groundwater has been detected for diuron and its demethylated metabolite.Glyphosate and imazapyr applied at the recommended rates have not been detected in WEED CONTROL amounts above 0.1 µg l–1 in water taken from the groundwater pipes. Conclusions l Weeds on railway tracks have to be removed to retain quality of track and to ensure safe railway traffic. l Acceptable weed control on railway embankments requires thorough knowledge of the weed control effect of the herbicide as well as its behaviour and persistence on and in the embankment. l To gather this knowledge there is a great need for testing of herbicides before they are introduced to the specific environment that a railway embankment provides. l From such studies it is an obvious fact that herbicides behave differently from what is known from their use on agricultural land usually with greater mobility and longer persistence.l After careful testing and evaluation of herbicides it is possible to select substances for weed control on railway tracks that fulfil high demands for a good weed control effect as well as a minimal influence on the surrounding environment. References Börjesson E.; Torstensson L. (2000). New methods for determination of glyphosate and (aminomethyl)phosphonic acid in water and soil. Journal of Chromatography A 886 207–216. Torstensson L. (1983). Investigations of mobility and decomposition of diuron in railway embankments. Swedish Environmental Protection Agency SNV PM 1764 43 pp. (in Swedish) Torstensson L. (1985). Further investigations of mobility and decomposition of diuron in railway embankments. Swedish Environmental Protection Agency SNV PM 2001 20 pp. (in Swedish) Torstensson L. (1994). Mobility and transformation of diuron in railway embankments. Proc. 5th Int. Workshop. Environmental Behaviour of Pesticides and Regulatory Aspects Brussels April 26–29 1994. Copin A. Houins G. Pussemier L. Salembier J.F. (Eds.) Developed from a symposium sponsored by the European Commission within the framework of COST Action 66 366–371. Wennberg E; Torstensson L. (1997). Gas-chromatographic method for determination of diflufenican in soil. International Journal of Environmental Analytical Chemistry 67 73–79. Dr Lennart Torstensson is Associate Professor of Microbiology at the Swedish University of Agricultural Sciences Uppsala Sweden. 2 1 Pes ti cide Out look – Fe b r u a ry 2001