Controlled Traffic Farming: Methods applied to Grassland Silage Management

Transcription

1 Controlled Traffic Farming: Methods applied to Grassland Silage Management Research Partnership: Grasslands, Forage and Soil AHDB Dairy Grasslands, Forage and Soils Research Partnership Report prepared for AHDB Dairy Paul Hargreaves 1, Sven Peets 2, Tim Chamen 3, Paula Misiewicz 2, David White 2 and Richard Godwin 2 1 SRUC Dairy Research and Innovation Centre 2 Harper Adams University 3 CTF Europe Ltd June

2 Contents Farmer Recommendations..4 Executive Summary Introduction Background Objectives Hypotheses Outline method 9 2. Literature review - Controlled Traffic Farming on Grassland Introduction Grassland production agronomics Conventional mechanisation Soil conditions and forage yields Summary of the effect of compaction on forage yield Controlled Traffic Farming mechanisation Economics of Controlled Traffic Farming for forage production Introduction Economic evaluation of traffic systems on an imperfectly drained soil in Scotland Economic evaluation of controlled traffic for grass silage production for dairy farms in Sweden Conclusions Proposals for controlled traffic in mixed forage grass and arable production systems Objectives Methodology CTF Designs Multiple operating widths from 1.5 m to 9.0 m System with 4 or 8 m operating widths System with 5 m operating widths... Error! Bookmark not defined System with 9 m operating widths System with 12 m operating widths Summary of designs and discussion Assessment of the effect of controlled traffic on grassland dry matter (DM) yield for three silage cuts

3 4.1. Introduction Objectives Materials and methods Experimental site Field management Baseline measurements Grass harvesting Soil measurements and analysis Statistical analysis Results Baseline measurements Effect on the soil structure from vehicle passes Dry matter reduction related to vehicle passes Dry matter total off-take Difference between CTF and N in area covered Conclusions Grassland CTF Economic Analysis Introduction Yield Benefit Costs Cost benefit analysis Case studies Conclusions Conclusions References..67 Appendix I 72 Appendix II

4 Farmer recommendations Based on the findings of this study: Controlled Traffic Farming (CTF) management for the second and third silage cut increased DM yield by 13.5% (0.8 t ha -1 ) compared to the use of a conventional or normal (N) traffic system. This was in agreement with previous studies. The increased number of passes of a vehicle increased soil bulk density by 15 to 18% for 6+ passes compared to zero passes. Using existing equipment together with vehicle auto-guidance and employing simple changes to the movement across the field the CTF system reduced the area covered by vehicle wheelings by 57% compared to N traffic. Depending on the working width of the CTF system employed, along with vehicle guidance, calculated reductions of the trafficked areas for grassland management are from 40% to 13%. Assuming: i) An average UK value of a 13% increase in dry matter (DM) yield from the absence of wheel damage for 2 and 3 cut harvest systems and ii) A reduced vehicle traffic area from 80% for a N traffic system to 45% for a CTF system. The calculated yield is increased by 0.53 t ha -1 for a 2 cut and 0.73 t ha -1, for a 3 cut system. Similarly reducing the trafficked area to 15% increases the yield by 1.00 t ha - 1 and 1.36 t ha -1 for 2 and 3 cut systems, respectively. Assuming a dry matter value of 72 t -1 the above yield increases are currently valued at between 38 ha -1 and 98 ha -1. The cost of low accuracy and non-repeatable positioning manual steered systems is calculated at less than ha -1 for grassland areas in excess of 100 ha and ha -1 for fully integrated, high accuracy systems for grassland areas in excess of 200 ha. This cost reduces to ha -1 for areas greater than 1500 ha cut -1. 4

5 Executive Summary This study was instigated to evaluate the hypotheses that economically viable controlled traffic systems could be introduced for grass silage production that would reduce the area and extent of compaction and maintain increases in and reduced variation in dry matter yield compared to a field with no restriction on traffic movement. Experimental work on a silage field in Dumfries, southwest Scotland (SRUC), with an 7 ha field with a relatively uniform soil type was split into two smaller fields, where one field was managed with Normal (N) traffic and the second field had controlled traffic (CTF) practices, based on a 9 m working width. Both the resulting soil conditions and grass yields where monitored at selected locations in the two fields and total dry matter (DM) yield from each field were compared. In parallel with this, Harper Adams University and CTF Europe undertook desk studies to: Review the previous literature on CTF in grassland Devise possible grassland CTF mechanisation systems by matching commercially available equipment Conduct an economic analysis by comparing the costs and benefits of grassland management with CTF. From which the following conclusions can be made: 1. Previous studies in UK grassland have shown that random traffic of a single cut of silage results in a trafficked area of 65% of the field and that wheel damage can reduce yields by between 5% and 20%. A mean potential grass yield benefit of c.13% was identified. 2. The effect of the traffic in the field studies have shown that there was a 13.5% increase for the second and third silage cuts for the CTF compared with the N. As the CTF system was only employed from the first silage cut then any increase in DM yield would not be seen until the second or third silage cut (Table S1). Table S1. Total Dry Matter (DM) off-take and differences (t ha -1 ) from the two field systems, normal (N) and controlled traffic farming (CTF). Silage cut Field system Difference s.e.d. P value N CTF 1 st silage cut nd silage cut rd silage cut nd +3 rd silage cut Total silage

6 The employment of the CTF system ensured a relatively rapid increase in DM yield compared to the N system by the end of three silage cuts. 3. The soil structure was affected by the number of passes from the vehicles during the management of the N and CTF fields, with the increased number of passes increasing the soil compaction as measured by soil bulk density, penetrometer (soil resistance) and a visual evaluation of soil structure (VESS). After the three silage cuts and fertiliser applications over one season the soil bulk density had increased by 14.7% (P<0.001) for the N field and 18.2% (P<0.001) for the CTF field between the zero passes and 6+ vehicle passes. 4. The use of existing equipment at a working width of 9 m, together with vehicle autoguidance for the three silage cuts and fertiliser applications, reduced the trafficked area by 57% from 87.4% for the N field to 30.4% for the CTF. The trafficked area covered by the management of the CTF field would have been reduced to (19.4%) if the second slurry application could have been conducted with the same wheelings as the first. Unfortunately, the spread width had to be reduced as the slurry was too thick to spread at 9 m and an extra wheeling was introduced. 5. Commercial equipment is available to configure CTF mechanisation systems for grassland based upon mower widths of 3, 4, 5, 9 and 12m along with vehicle guidance, can reduce the range of trafficked areas from 40% to 13%. This range increases to between 48% and 18% when considering arable crops in the rotation. 6. Based upon a 13% increase in yield from the absence of wheel damage and the average dry matter yields for 2 and 3 cut harvest systems in the UK, reducing the trafficked area from an assumed 80% for random traffic to 45% for CTF increased the yield by 0.53 t ha -1 and 0.73 t ha -1 for 2 and 3 cut systems respectively. Similarly, reducing the trafficked area to 15% increased the yield by 1.00 t ha -1 and 1.36 t ha -1 for 2 and 3 cut systems, respectively. 7. Assuming a dry matter value of 72 t -1 ; the above yield increases are currently valued at between 38 ha -1 and 98 ha Based upon the assumptions that: it is only the cost of the guidance system that is needed to implement CTF systems and that four guidance systems were required to equip the harvester and the accompanying tractors. The cost of low accuracy and nonrepeatable positioning manual steered systems is less than ha -1 for areas in excess of 100 ha and ha -1 for fully integrated, high accuracy systems for areas in excess of 200 ha reducing to ha -1 for areas greater than 1500 ha cut -1. 6

7 9. The break-even area for implementing CTF depends upon the level of investment, the trafficked area and the number of cuts each year. This ranges from 28 ha for low accuracy, manual steered systems with a 35% trafficked area with 3 cuts year -1 to 250 ha for the fully integrated, high accuracy real time kinematic navigation systems, reducing to 175 ha with a trafficked area of 15%. 10. The data from the economic analysis also provide a basis for the negotiation between the benefit to a farmer and the costs to a contractor when considering the implementation of grassland CTF systems. There is sufficient evidence to support the hypotheses that reduced traffic coverage of a silage field with a planned system of controlled traffic will reduce compaction and therefore increase yield and reduce variation compared to a field with no restriction on traffic. 7

8 1. Introduction 1.1 Background In the UK, grassland fieldwork is generally conducted in an ad hoc manner with no conscious attempt to re-use equipment wheel ways or pathways in the way they are used in arable fields, such as tramlining or controlled traffic farming (CTF) 1. Two passes from tractor compaction in autumn field conditions has been shown to reduce yield of the following years 1 st cut by between 16 and 25% (earlier DairyCo funded compaction studies). In arable agriculture a number of studies have shown that in a single year approximately 90% of a field can be covered by tractor/harvester/trailer wheels at least once with a number of areas within the field repeatedly trafficked. More recently and specifically, studies at Harper Adams University in 2012 showed approximately 64% coverage during a single grass harvesting operation for both forage chopper and baling operations. A current study with winter wheat at HAU has shown a significant (19%) yield improvement by controlling field traffic to reduce the extent of soil compaction from field operations. Both silage yield and operational cost reduction benefits from controlled traffic have been shown with grassland in Denmark where grassland CTF systems are well developed. 1.2 Objectives 1. To devise and demonstrate an effective Controlled or Reduced Traffic Farming system for grassland. 2. To determine the effect of repeated traffic of a traditional management system on the within-field variation and overall yield in grassland silage yield through a season. 3. To identify patterns of controlled traffic that will result in a net yield benefit. 4. To include other vehicles and trailer sizes and axle widths. 5. Determine the potential economic and environmental benefits of CTF (or Reduced TF) systems. 1.3 Hypotheses A planned system of controlled traffic will reduce the area and extent of compaction within a silage field and therefore increase yield and reduce variation compared to a field with no restriction on traffic. 1 CTF is the confinement of all field traffic to the least possible area of permanent traffic lanes. 8

9 1.4 Outline method In order to achieve the above the following work was planned and executed by the research partners of SRUC, Harper Adams University (HAU) and CTF Europe by: 1. Undertaking a review of the relevant UK and international literature. 2. Developing proposals for controlled traffic in forage grass production systems by reviewing currently available commercial equipment. 3. Conducting field studies, in Dumfries, on the impact of CTF and reduced traffic systems by splitting a 7 ha field, with a uniform soil type (Crichton/Midpark freely/moderately freely drained) and management history into two smaller fields with identical layouts of 3.5 ha. This original field was re-seeded during late summer 2014 to minimise the carryover of historic compaction damage. The two 3.5 ha fields were sampled on a grid system of 40 sampling points for soil structure (soil bulk density, soil resistance (penetrometer) and a visual evaluation of soil structure (VESS)), along with the ph, phosphorous (P) and potassium (K) to ensure conformity across the fields and no differences between fields that could effect the yields. i. The first 3.5 ha field was farmed with normal (N) traffic management, with vehicles applying slurry, cutting and harvesting with the current ad hoc management of vehicle and trailers movements. ii. The second 3.5 ha field was managed with controlled traffic (CTF) and had a pattern of set wheeling s (tramlines) established by auto-steer/tramline markers for the vehicles to travel along the length of the area. The tractors and trailers exited the section of the field when fully loaded along pre-set wheeling s and the along the headlands. Three silage cuts were taken (May, July and August) from each section of the field through the year, with associated applications of fertiliser. Prior to each silage cut 40 small (2 x 2 m) areas were cut for DM; these were 8 cuts each from vehicle passes of zero, 2, 4, 6 and 6+. The total DM off-take from each field was also measured by weighing the fresh matter in each trailer and sub-sampling for DM. 4. Conducting an economic analysis of the viability of CTF in grassland by comparing the yield benefits and the costs of machine guidance systems in enabling CTF practices to be adopted. 9

10 2 Literature review - Controlled Traffic Farming on Grassland 2.1 Introduction Agricultural production systems are increasingly characterised by extensive in-field trafficking of heavy machines. For example, the extent of random traffic of forage harvester and round baler traffic across a field in a study by Harper Adams University has shown to be 63.8% and 63.4% of field area respectively as shown in Figure 2.1 (Kroulik et al., 2014). The first wheeling causes most of the compaction (Davies et al., 1993) with subsequent incremental effects as the number of passes increases (Raghavan et al., 1977). Traffic on the soil causes compaction and has the effect of increased bulk density, increased shear strength, reduced porosity (Douglas et al., 1998) and reduced air and water permeability (Davies et al., 1993) and (Chyba et al., 2014). Figure 2.1. Traffic patterns during a grass harvest using a forage chopper (left) and round baler (right). Trafficked area 63.8% and 63.4% of field area respectively. The bulk density of soil depends on the amount of air space in the soil; vehicle traffic closes the larger air-filled spaces, making the increase in bulk density an indication of compaction (Davies et al., 1993). The extent of compaction depends on soil texture, structure and moisture content. Clay loam soils have a wide range of moisture contents in the friable range, with silt loams and sandy clay loams having smaller friable ranges (Baver et al., 1972). At moisture contents below the friable range, soils are described as hard or cemented 10

11 and are relatively strong, with field traffic causing relatively limited soil structural damage. However, when soils are trafficked at moisture contents in the friable range and above, in the plastic range, they deform more easily (Davies et al., 1993). Above critical moisture content compaction declines because pores are water filled, but severe soil damage may occur because the strength of the soil decreases. Generally moist soil is more susceptible to compaction than dry soil (Alakukku, 1999), hence, in years with normal weather patterns soil/crop damage will most probably occur as a result of operations during the spring or autumn months. Soil compaction causes yield reduction and results in costly remedial actions. The potential to improve forage grass yields by controlling the traffic can be as high as 35% according to Chamen (2011). Controlled Traffic Farming (CTF) is a traffic management system where all the field traffic is confined to permanent wheel ways across the field (Chamen, 2011). To enable CTF the track gauges of the machines have to be matched. The working widths of the implements also have to be matched so that they are equal to, or are an integer multiple of, the base module width (Figure 2.2). The base module is the narrowest working width. Figure 2.2. Controlled Traffic Farming. Source: CTF Europe Global Navigation Satellite Systems (GNSS) such as Navstar Global Positioning System (GPS) with Real Time Kinematic (RTK) correction based auto-guidance systems are commonly used to adhere to the permanent wheel ways. The trafficked area depends on the base module width: for example a 6 8 m system results in approximately 25% of the field 11

12 having a trafficked area, this reduces to approximately 17% for wider 12 m systems (Chamen, 2011). Chapter 3 develops this for a range of grassland options. 2.2 Grassland production agronomics Grass is grown as permanent grassland (pasture) or in a crop rotation (grass ley). Grass fields can often be of a small size and irregular shape. The grass ley is often established for a period of 3 to 4 years dependent on the yield and quality of grass and is then ploughed in as the yield and quality decline. Establishing a grass ley costs from 260 per ha for a 3 to 5 year ley to 338 per ha for 1 to 2 year ley (Nix, 2015). If extra years of high quality grass could be achieved, then there is a benefit by spreading the cost over those years, with a saving of 10 to 170 per ha year -1 depending on ley type. Forage yields in the UK are in the range of t ha -1 for grass leys and t ha -1 for permanent pastures (Nix, 2015) with the dry matter (DM) yield on average 7.6 t ha -1 (Nix, 2015). Forage is preserved as clamped grass silage, wrapped grass silage or hay. There are usually three cuts of grass per season in the UK for clamp silage: May, July and September. In addition to yield, there are a number of key quality parameters such as crude protein, which are important in forage production as they determine the animal performance. 2.3 Conventional mechanisation The conventional mechanisation system for the establishment of a grass ley uses the mouldboard plough as the primary tillage operation, as a good seedbed is essential to accommodate the small grass and clover seeds which need to be in close contact with the soil. For secondary tillage harrow or disc and rolls are used. Sowing to a depth of no more than 10 mm can be undertaken by either drilling or broadcasting. During the life of a grass ley, patches of re-seeding may be conducted. Additionally, pressing stones into the topsoil using a heavy roller may also be required in spring. A grass ley will require nutrition and the applied fertiliser can be either organic (from solid manure or slurry) or an inorganic mineral form. The spreading widths of fertiliser spreaders can vary from 8 to 48 m. Sprayers have working widths from 12 to 45 m, but the most common is m (Bell, 2008). Manure spreaders have working widths of 3 14 m and capacities range from 3 to 30 tonnes of farmyard manure depending on the model. Slurry spreaders have a tank capacity from 3,500 to 35,000 litres with gross weights up to 45 tonnes. Slurry can be spread onto the land with a deflector plate behind the spout in a band up to 20 m. Because of environmental considerations slurry is injected into the land with the help of tines that cut slots in the soil (Warner and Godwin, 1988). Light tines with narrow spacing that make shallow slots are preferred for grassland (Bell, 2008). Typical working widths of slurry spreaders are as follows: meadow injectors m, meadow spreaders 9 30 m, and arable injectors m. The tractor/tanker track gauges are typically mm with tyre widths of mm ( Some slurry 12

13 tanker models may have 4 parallel wheels (e.g. 650/65R42) for an optimized load distribution in order to reduce ground compaction in meadows (Joskin TETRAX2). Grass may be cut three times per season in the UK with a mower which can be front mounted, rear side mounted or trailed (Bell, 2008) with working widths from 3 to 12 m. The widest currently available mower is 14 m, manufactured for North America. In Europe, however, 9 or 12 m widths are typical. Mower-conditioners combine cutting with the first treatment (splitting to increase wilting speed) of the new swath. This is considered in more detail in Chapter 3. Freshly cut grass will have moisture contents of approximately 75%. For safe storage in a hay barn this must be reduced to approx. 20%. Cut grass is left on the field to wilt. Wilting grass is conditioned, turned over, collected to a swath; the swath may then be spread out with tedders and rakes. Working widths of tedders and rakes are up to 15 m and working speeds are up to 15 km/h (Bell, 2008). The working widths of these machines are adjustable. Harvested grass is conserved as silage or hay. Hay is collected into bales on the field. Work rates of balers are in excess of 40 ha in a day for big square bales (Bell, 2008), which is typically suited for agricultural contractors, with big square hay bales weighing kg m -3 and for green silage between kg m -3. The bales are collected from the field with bale collectors, tractor front-end loaders, material handlers or self-loading trailers; these trailers vary in size. Silage is made as clamp silage, bag silage or tube silage (e.g. Ag-Bag). Forage harvesters are used for chopping the cut grass into suitable size lengths for clamp silage and propelling the chopped grass into a trailer drawn behind or by a tractor running either behind or alongside the forage harvester. The tractor/trailer transports the chopped forage from the field to storage. As forage harvesters are large and expensive machines they are usually offered as a service by a contractor. Tractor and trailers leave the field usually the shortest way, i.e. the traffic is random. As an alternative to the forage harvester a forage wagon can be fitted with a pick up roll and chopper. Forage wagons capacities vary from m 3. A 30 m 3 forage wagon has a capacity for approximately 20 t of grass (Bell, 2008). A 47.5 m 3 wagon is rated at 26 t with track gauges mm and wheel/tyre widths of mm. The total number of machine passes on the field can be 11 or more with conventional mechanisation systems (slurry spreading, 3x fertilizer application, 3x mowing, 3x tedding/raking, 3x lifting, 3x trailer) if these are not controlled a large area of the field can suffer from compaction damage. Alleviation of soil compaction can be achieved with the following machinery: aerators, sward lifters, and subsoilers. The power requirement of sward lifters and subsoilers is relatively 13

14 high, for example a minimum of kw is recommended for a twin legged subsoiler (Godwin and Spoor, 2015) with working widths in the range of m. 2.4 Soil conditions and forage yields Summary of the effect of compaction on forage yield The effect of soil compaction on forage crops has been recognised from the results of a number of scientific studies that have quantified the yield decrease due to soil compaction. A summary of these is given in Table 2.1. This table has been adapted from Alvemar (2014) by significantly increasing the number of studies reviewed and including the effects of field traffic on soil conditions. The following conclusions can be drawn from the collection of literature in Table 2.1: Dry bulk density increases. Air-filled porosity decreases. Cone penetrometer resistance increases. Water filled porosity increases. The following conclusions can be drawn concerning forage yield from the literature cited in Table 2.1: The yield decrease due to soil compaction is in the range 5 74%. The long-term yield decrease for UK conditions is in the range of 5 20% with a mean of 13%. The largest yield decrease takes place during the first cut. Although the yield may increase significantly for the second cut with the conventional or normal traffic systems, the total yield for the year is not significantly more than for the zero traffic system. The yield of dry matter declines with accumulated amount of traffic (Douglas & Crawford, 1989a & 1991) (Hakansson et al., 1990), which contradicts Frost (1988ab) who found there was no cumulative effect on grass yield decrease following traffic in the same wheel tracks at each silage harvest. Frequent or delayed wheeling (in excess of six days from further field operations) has much greater adverse effect than infrequent or un-delayed wheeling (Frame, 1985 and Frame & Merrilees, 1996). 14

15 Table 2.1. Grass yield decreases due to soil compaction in the trafficked area. Study Yield Forage crop Soil type Location Effect on soil conditions decrease Are et al. (2015) 21 40% Ryegrass & Lucerne Sandy loam Estonia No difference between water-filled porosity. Air-filled porosity % greater on uncompacted soil. Douglas & Crawford 32% Ryegrass Clay loam Scotland See Figure 2.3 (1989a & 1991) Douglas & Crawford (1989b & 1993) Douglas & Crawford (1989c) & Douglas et al. (1992) Douglas et al. (1995) 42, 28 and 16% 1 st cut in successive years 13% over four years 13% over four years Ryegrass Clay loam Scotland Macro-pore volume close to the soil surface 11.9% for zero traffic, 4.4% for severe traffic. Dry bulk density increased for severe traffic by Mg m -3 at mm depth. Ryegrass Clay loam Scotland Soil bulk density, strength and water content indicated poorer soil structure for compaction treatment. Ryegrass Clay loam Scotland Bulk density greater for compaction treatment. The volume and number of macropores in topsoil significantly smaller for compaction treatment. Elonen (1986) 8 68% Clay loam Finland Frame (1985) & 25* 33%* and Scotland Frame (1983) 17** 31%** Frame & Merrilees (1996) Red clover* and red clover/perennial ryegrass** Sandy loam on loamy sand Wheelings increased bulk density (1.24 Mg m -3 no wheeling, Mg m -3 wheeling) and decreased drainable porosity (9.1% no wheeling, % wheeling). 14%, 6% and Diploid and Sandy loam Scotland 9% annual and tetraploid on loamy mean 8% ryegrass sand % v/v for wheeling. Dry bulk density Mg m 3 for no wheeling, Mg m 3 for wheeling. Airfilled porosity % v/v for no wheeling, 15

16 Frost (1985) & Frost (1988a) Frost (1988b) 5 20% per year; 12 78% 1 st cut 8 20% per year; 27 32% Ryegrass Clay loam sandy clay loam Ryegrass Clay loam sandy clay Northern Ireland Northern Ireland Slightly higher cone penetrometer resistance values for compacted soil at depth mm. 1 st cut loam Håkansson et al. 9% Grass/clover Various Sweden Yield losses result of damage to the sward (1990) rather than of soil compaction. Hansen (1996) 27% Grass/clover Sandy loam Norway Air-filled porosity reduced from 12 to 7% because of compaction. Hargreaves et al. 13.1%* and Ryegrass England* & (2012) 15.0%** 1 st cut Scotland** Hargreaves et al. (2014) Jorajuria et al. (1997) 15.2%** 1 st cut; 14.3% Total cut* Ryegrass 74% Ryegrass & white clover Sandy loam* & silty clay loam** Sandy loam* & silty loam** Silty loam on top of silty clay loam England* & Scotland** Argentina Jørgensen et al % Ryegrass & Denmark (2009) clover Negi et al. (1981) 44* 58%** Silage corn Clay** & Canada See Figure 2.4 Sandy loam* Rasmussen & Møller 21 54% Ryegrass & Sandy loam Denmark Increase of bulk density by 0.08 Mg m -3 at mm; 0.17** Mg m -3 at mm. Increase of penetrometer resistance by * MPa at mm for compaction; * MPa at mm for no compaction; * MPa at mm for compaction. Water-filled porosity significantly greater (largest difference 23%) for compacted**; No significant changes in bulk density and penetrometer resistance. Cone index: at mm depth decreased up to 40%, at mm depth increased by 20 90%. Bulk density: at mm increase 2 10%, at mm 38 45% increase, mm 18 25% increase. 16

17 (1981) Grass/clover & Silty loam Reintam et al. (2013) 10 50% Ryegrass & Lucerne Sandy loam Estonia Air-filled porosity below 5% for fertilised ryegrass on compacted soil. Volden et al. (2002) 16** 47%** Timothy-grass Sandy soil* and peat soil** Norway On sandy soils no significant difference. On peat soils air-filled porosity 13.6 % for no traffic, and % for trafficked soil 17

18 Table 2.2 shows the results of a more detailed study by Frost (1988ab) who conducted a four-year experiment to investigate the effect of traffic treatments on forage yield in Scotland. The traffic treatments for each harvest were: tractor and slurry tanker traffic in February followed by tractor and trailer traffic 7 days after the first and second cut tractor and slurry tanker traffic in March tractor and slurry tanker traffic in February following compaction in November 1984 twice with a loaded rough terrain forklift truck in previous year s wheel tracks tractor and slurry tanker traffic in March and tractor and trailer traffic after the first and second cut. Table 2.2. Grass yields in % of yield in zero traffic (Frost, 1988ab) Year/Cut Treatment 1983 Zero 1 pass 3 pass 1 st 100 a 87 b 69 c 2 nd 100 a 94 a 83 a 3 rd 100 a 112 a 100 a Total 100 a 95 ab 80 c 1984 Zero 1 pass 2 pass 1 st 100 a 70 b 48 c 2 nd 100 a 105 a 101 a 3 rd 100 a 88 b 93 ab 4 th 100 a 97 a 97 a Total 100 a 88 b 80 c 1985 Zero 1 pass 1 st 100 a 73 b 2 nd 100 a 108 b 3 rd 100 a 97 b Total 100 a 92 b 1986 Zero 1 pass 1 st 100 a 68 b 2 nd 100 a 82 b 3 rd 100 a 94 a Total 100 a 81 b Row values with the same letters are not significantly (P < 0.05) different The following conclusions can be drawn concerning forage yield from Table 2.2: Over the four-year period 12.1% (4.33 t ha -1 ) more dry matter was produced in the zero traffic treatment than in the 1 pass compaction treatment. Traffic in the spring (February March) causes the most adverse effect on yield (Frost, 1988ab). The largest yield decrease takes place during the first cut. If there is no further traffic and the soil is in a good condition and resistant to compaction, then yield can recover in subsequent cuts (Frost, 1988ab). If compaction takes place then there will be consistent decrease of yield. Hargreaves et al. (2012) and Douglas et al & 1995 also reported a yield increase for the second cut.

19 There is no cumulative effect on grass yield decrease following traffic in the same wheel tracks for each silage cut (Frost, 1988ab). In addition, Table 2.3 shows the results of a more detailed study by Douglas et al. (1992 & 1995) who conducted an eight-year experiment to investigate the effect of zero and reduced ground pressure traffic systems on forage yield in a comparison with a conventional system in Scotland. The soil and crop management used was the following: Application of fertiliser in the last week of March, ground rolling in May, first cut in the first week of June followed by an application of fertiliser, second cut in early August followed by the final fertiliser application, and third cut in early October. Tyre inflation pressures were in the range of kpa for conventional, and kpa for reduced ground pressure system. The following conclusions can be drawn concerning forage yield from Table 2.3: Over the eight-year period 14.7% (13.7 t ha -1 ) and 14.0% (13.1 t ha -1 ) more dry matter was produced in the zero traffic and reduced ground pressure system, respectively, than in the conventional system. The largest yield decrease was generally in the 1 st cut for conventional traffic which was attributed to wet soil conditions (the amount of rainfall following the spring application of fertiliser). The yield trend at second cut was variable. Those years characterise by low summer rainfall, the yield at second cut was higher for conventional traffic system. Traffic with wet soil conditions during cutting decreased the yield of the next cut with the conventional traffic system. 19

20 Table 2.3. Grass yields in % of yield in zero traffic (Douglas et al., 1992 & 1995). Year Cut Treatment Zero Reduced Conventional pressure st 100 a 89 b 80 b 2 nd 100 a 165 b 177 b 3 rd 100 a 77 b 59 c Total 100 a 104 a 98 a st 100 a 99 a 78 b 2 nd 100 a 89 b 75 c 3 rd 100 a 78 b 72 b Total 100 a 91 b 76 c st 100 a 93 a 65 b 2 nd 100 a 131 b 134 b 3 rd 100 a 100 a 86 b Total 100 a 104 a 87 b st 100 a 88 b 67 c 2 nd 100 a 114 a 109 a 3 rd 100 a 100 a 97 a Total 100 a 99 a 87 b st 100 a 103 b 89 c *** 2 nd 100 a 100 a 101 a 3 rd 100 a 119 b 113 c ** Total 100 a 105 b 98 c st 100 a 107 a 91 a 2 nd 100 a 107 b 98 c ** 3 rd 100 a 90 b 72 c ** Total 100 a 101 a 87 b st 100 a 95 b 77 c 2 nd 100 a 114 a 99 a 3 rd 100 a 92 a 90 a Total 100 a 100 a 87 b ** st 100 a 95 b 77 c ** 2 nd 100 a 88 b 80 c ** 3 rd 100 a 97 a 95 a Total 100 a 93 b 81 c *** Row values with the same letters are not significantly (P<0.05) different. For rows with ** P<0.01, and for rows with *** P< Douglas & Crawford (1989a & 1991) investigated the correlation between the amount of traffic and dry bulk density and dry matter yield (Figure 2.3). The amount of traffic (N s) was defined as the product of number of wheel passes (N) and maximum (front wheel) tyre/soil contact stress (s). In relation to rainfall, they found that a wet June disadvantages the swards on the most dense soil, but with a dry June there was no significant difference in yield between zero traffic and compaction treatment. Both Frame (1985) and Hakansson et al. (1990) state that wheel traffic during silage making should be minimised by undertaking as few operations as possible. While later Frame and Merrilees (1996) suggested that in practice silage operations in the field should be conducted 20

21 Dry matter yield, t/ha with the lightest equipment and fewest traffic activities. Frost (1988a) and Hakansson et al. (1990) suggest that yield reduction from a field would be reduced if the same wheel tracks could be utilised, an early suggestion for controlled traffic practices. Figure 2.3. The effect of the amount of traffic applied over two years on soil bulk density in upper 120 mm ( ) and dry matter yield (- - -) (Douglas & Crawford, 1989a). The effect of soil bulk density on maize silage yield in eastern Canada is reported by Negi et al. (1981) where a small increase in bulk density causes a noticeable yield decrease (Figure 2.4) Soil dry density, t/m 3 Figure 2.4. Relationship between maize silage yield and soil bulk density (Negi et al., 1981). 21

22 The effect of wheel loads and tyre pressures was reported by Jørgensen (2009) who found that increasing wheel loads have a greater negative effect on grass yield than increasing tyre pressure. Jorajuria et al. (1997) has found that the same level of compaction within the mm depth could be achieved with several passes of a light tractor (2.3 Mg) as readily as with fewer passes of a heavy tractor (4.2 Mg), where both tractors had similar mean ground pressure. 2.5 Controlled Traffic Farming mechanisation Although the core concept of CTF is relatively simple as given in Figure 2.2, namely confining the field traffic to permanent, common wheel traffic lanes, it requires the matching of both wheel (or track) gauges and implement working widths, which in practice may be more difficult. As a result there is no universal layout because of a wide range of diverse machinery to consider, the main issue is with fitting the harvester into the CTF system and ideally matching the wheel (or track) gauges of all of the equipment. This can be a problem in the UK and Europe as there are road traffic regulations that govern the overall width of tractors. To overcome this problem in arable farming, compromise solutions are used as shown in Figure 2.5 (Galambošová & Godwin, 2012; Chamen, 2011). OutTrack one wheel (or track) gauge is used for all machinery except the combine harvester, which runs on a wider gauge due to its inherent design. Examples are that at 8 m module width, 25% of the field is trafficked, while at 12 m, 17% is trafficked. HalfTrack two wheel (or track) gauges are used with one being half the other. The implement width is the integer multiple of one or other of the gauges. TwinTrack a system that uses two wheel (or track) gauges with vehicles on the narrower gauge straddling adjacent passes of vehicles on the wider gauge. The implement width is the sum of the gauges, or a direct multiple of this addition. For example, with a 5 m combine cutter bar width (track gauges of 2.0 m and 3.0 m), 25% is trafficked while with a 10 m cutter bar, 17% is trafficked. AdTrack uses two track gauges, the narrower using one track of the wider. Implements can be any common width or direct multiple of the base module. Typical trafficked areas of the field range from 20% - 25%. Although CTF systems are becoming accepted for cereal production in the UK, currently with some 50,000 ha in production (Godwin, 2012) there is no robust data about the CTF for forage crops. However, CTF is being practiced for forage production in Scandinavia as reported by Kjeldal (2013) and detailed in Table 2.4 and Appendix I (CTF in Forage Grass). It can be concluded that the most popular module width is 12 m (trafficked area from 13 to 26%) and mower widths from 6 to 12 m. 22

23 (a) OutTrack (b) HalfTrack (c) TwinTrack (d) AdTrack Figure 2.5. CTF systems used in arable farming (Source: Tim Chamen, CTF Europe, 2012). 23

24 Table 2.4. Four case studies of farms practicing CTF for forage production in Denmark (CTF Europe Ltd website). Case Contractor Mosegaarden in central Jutland in Denmark (Pedersen, 2012) Farmer Jörgen Sønderby in Knepper Hedegard in Denmark A farm in Grøndal in Denmark Farmer Carl Einar Sorensen in Kjellerup in Denmark Description Using 12 m system for forage production with the following mechanisation: mower 12 m (JF Stoll GTX 13005), rake and tedder 12 m, slurry injection 12 m (Samson TGX tanker and Vredo injector), forage harvester (CLAAS Jaguar), trailers pulled by forage harvester, GPS RTK autoguidance (Source: - Accessed 14/10/2015). Using a 6 m mower in 12 m system for forage production with the following mechanisation: mower 6 m, rake 12 m, slurry spreader 12 m, forage wagon 12 m, trafficked area 31% of field area providing a 10% yield increase of the organic grass leys (Alvemar, 2014) (Source: sonderby - Accessed 14/10/2015). Using an 8 m mower in 12 m system for forage production with the following mechanisation: mower 8 m, rake 12 m, slurry injection 12 m and 24 m, forage wagon 12 m, the mower is running every third time in a 12 m track, i.e. there are two extra tracks for every 24 m, GPS and RTK auto-guidance in combination with visual markers on the field (Source: - Accessed 14/10/2015) Producing forage with a 12 m system, grass is harvested with a forage wagon, no further details available (Source: - Accessed 14/10/2015). In the UK Crathorne Farm in Yorkshire, an AHDB Dairy demonstration farm, has experimented with CTF for clamp silage forage production (Dugdale, 2015 and James, 2016) where two fields were selected for comparison. The first was treated with random traffic and the second field as controlled traffic. The results of one season are summarised in Table 2.5. Although the trafficked area was significantly reduced, there was no difference in crop yield. However, fuel consumption was reduced. Table 2.5. Comparison of random traffic farming and CTF in 2015 season at Crathorne Farm in Yorkshire (Dugdale, Personal communication). Random Traffic Farming Controlled Traffic Farming Mower (m) 9 9 Rake (m) 12 9 Spreader (m) Harvester (m) 12 9 Trailer random offset one track line Total distance (km/ha) Trafficked area 57.4% 23.5% Work rate (ha h -1 ) Fuel consumption (l ha -1 ) Fuel consumption (l t -1 )

25 2.6 Economics of Controlled Traffic Farming for forage production Introduction Although it has been demonstrated in a number of scientific studies (see Section 2.4) that conventional random traffic system results in soil compaction and a reduction in forage yield, only two economic studies of CTF for forage production were located. These concerned work by Stewart et al. (1998) in Scotland and a more recent study by Alvemar (2014) in Sweden Economic evaluation of traffic systems on an imperfectly drained soil in Scotland Stewart et al. (1998) reported on an economic evaluation of perennial ryegrass grown for silage in contrasting wheel-traffic systems on clay loam soil in a Scottish climate based on eight years of experimental data. The wheel-traffic systems were zero traffic, reduced ground pressure and conventional traffic. For conventional system standard tyres were used, for reduced ground pressure larger tyres at lower inflation pressure were used, and for zero traffic machinery operated on permanently positioned wheel-tracks with 2.8 m centres. During the eight-year period a perennial ryegrass crop was harvested three times each season and dry matter yield, herbage digestibility and crude protein content were measured. The economic analysis was based on the gross margin (GM) concept which was defined as GM = (Price of crop x Yield) Variable cost. Revenue was calculated as a function of digestible dry matter yield and an estimated price of equivalent dry matter nutrients. The variable costs consisted of fertiliser costs and nominal annual re-seeding and spray costs. GMs were calculated for each traffic system including equivalent crop losses in the permanent wheel tracks of a 12 m gantry. The zero (gantry system) and reduced ground pressure traffic systems produced GMs of 500 ha -1 and 503 ha -1 respectively, which were 19% greater than the GM in the conventional traffic system ( 421 ha -1 ). The estimated current value is 714 ha -1, 718 ha -1 and 601 ha -1 respectively taking into account the Consumer Prices Index published by the UK Office for National Statistics Economic evaluation of controlled traffic for grass silage production for dairy farms in Sweden Alvemar (2014) conducted an economic study on profitability of CTF for grass silage production in dairy farms in Sweden. He investigated the investment costs of CTF, decisive factors for the profitability of CTF investment, and profit margins of Random Traffic Farming and CTF. The economic models were developed and tested on a hypothetical dairy farm. The Hypothetical farm consisted of 300 ha of clay soils with 300 dairy cows in Sweden, of which ha was grassland, with single farm payments of 143 ha -1 (exchange rate 1 GBP = 12.1 SEK March 2016). He assumed that the investment in tractor guidance had been previously been made by the study farm but not that for the cost of the addition of Real Time 25

26 Kinematic correction signal to improve the control resolution to a few centimetres and provide a stable year on year control system. He assumed a grass ley management system over a three to four year period with three cuts per season. The crop rotation included winter wheat, oats, spring barely, and oil seed rape. The following factors were considered in an economical optimisation model: mechanisation, fixed and variable costs, dry matter yield, and silage quality in terms of crude protein content. The following mechanisation system was considered for the Random Traffic Farming: mower 6 m, rake 12 m, slurry spreader 24 m, fertiliser spreader 24 m, roller 8 m, self-propelled forage harvester, tipper, and 170 kw tractor. The trafficked area was 74% of the field area. The proposed Controlled Traffic Farming had a 12 m module width and the following mechanisation system: mower 12 m, rake 12 m, slurry spreader 24 m, fertiliser spreader 24 m, roller 12 m, self-propelled forage harvester with trailed tipper (changed at headland) and 220 kw tractor. The trafficked area was 20% of the field area. Both fixed (capital) and variable (maintenance, fuel consumption, and labour) costs were considered, a summary of these is given in Table 2.6. Table 2.6. Machinery costs Alvemar (2014) in GBP Cost Random Traffic Controlled Traffic GBP GBP/ha GBP GBP/ha Fixed 86, , Total variable 82, , Total annual 168, , In terms of dry matter yield, he assumed 12% overall yield increase for CTF vs RTF (8960 kg DM ha -1 and 8000 kg DM ha -1 respectively), a higher silage content in the feed ration provided a lower feed cost. In terms of silage quality, he assumed a 10% increase in clover content providing a 25% increase of crude protein content. A higher nutritional value of the silage decreased the required share of grain and protein concentrate in the feed ration, thus a lower feed ration cost was achieved. The economic analysis compared three alternatives: 1. RTF and standard feed ration against CTF assuming 12% increased silage yield. 2. RTF and standard feed ration against CTF assuming 12% increased silage yield and 10% increase of silage clover content resulting in an increase of crude protein by 25%. 3. RTF and standard feed ration against CTF assuming higher quality silage. The main results of the analysis are given in Table 2.7, these show that the yield increase is the most important factor affecting the profitability when converting to CTF. An increase in clover content without an overall yield increase does not provide sufficient increase in 26

27 revenue to switch to CTF. The potential profit due to CTF was 40 to 49 ha -1 (Alternative 1 and Alternative 2 respectively). The profit margins were 0.5 % for RTF, and 1.25% to 1.55% for CTF (Alternative 1 and Alternative 2 respectively). Table 2.7. Results of economic analysis Alvemar (2014) in GBP Grass area (ha) Winter wheat area (ha) Grass ley number of years Variable machinery cost (GBP/ha) Total farm revenue (GBP) Increase in profit (GBP) Alternative 1 Alternative 2 Alternative 3 RTF CTF RTF CTF RTF CTF ,693 16,654 4,693 19,347 4,693 5,047 11,961 14, Conclusions The following conclusions can be drawn from the literature review: 1. There is unequivocal evidence in the literature that traffic has a deleterious effect on soil structure and forage yield. The long-term forage yield decrease in the UK is in the range of 2 24%, with a mean of 13%. 2. In the longer term 12.1 to 14.7% more dry matter can be produced in the zero traffic system than in the conventional traffic system in the UK. 3. Trafficking in wet soil conditions has the most damaging effect on the forage yield. The largest yield decrease takes place during the first cut, which is attributed to wetter conditions in the spring. 4. The yield trend at second cut is variable on years characterised by low summer rainfall, the yield at second cut for the conventional traffic system can improve the compared to the CTF. 5. Conventionally random traffic mechanisation results in a trafficked area of 64% for a single cut in England, for multiple operations in a given year this will be higher. The economic analysis in Sweden estimated 74% of the field area was trafficked. 6. Controlled Traffic Farming is a new technique in forage production in the UK. The experience of Danish farms suggests 12 m module is the most common for CTF 27

28 mechanisation in forage production. The trafficked area is 20 30% depending on the mower width in the CTF system. 7. Economic evaluation suggests CTF is profitable for forage production. The key factor for profitability is the increase in yield and a contributing factor is an increase in quality of forage (crude protein content) (potential profit 40 ha -1 to 49 ha -1 ). There is an indication of an increase of gross margin of 19% ( 117 ha -1 ) in case of zero traffic system compared to a conventional or normal traffic system. 8. Whilst the main aim of this work was to focus on CTF, cited references also show a benefit in forage yield from the use of reduced inflation pressure tyres. 28

29 3. Proposals for controlled traffic in mixed forage grass and arable production systems 3.1 Objectives The aim of this part of the study was to identify a range of commercially available equipment that could form the basis of controlled traffic systems, principally for forage grass production on farms of different sizes. In addition, consideration was given to equipment that might be suitable for arable operations on these different farms. Where necessary minor modifications to machines or alterations in working practice, were deemed acceptable. 3.2 Methodology As there are few, if any, fully integrated controlled traffic systems for grassland farms in the UK, it was necessary to start from first principles. In the case of forage production, the greatest limiting factor was found to be the available widths of grass cutting equipment. These machines tend to be offset rear mounted or trailed, or a combination of front mounted and rear mounted or trailed. The widest offset rear mounted or trailed machines were between 3 and 5 m but there is then a jump in width to about 8 m associated with the combined front and rear machines. Necessarily, the machines listed are very specific because every manufacturer offers slightly different axle gauge settings (distance between wheels on one axle), tyre equipment and operating widths for implements. The inclusion of one manufacturer over another does not imply endorsement, merely the equipment listed happened to be the most compatible with the particular controlled traffic system being proposed. One shortcoming of many manufacturers literature is the absence of sufficient information to easily determine track gauges. This was overcome in most instances by interrogating data sheets to find overall widths of machines and checking that this dimension related to the tyres rather than machine metalwork. Trafficked areas were calculated using axle gauges and tyre section widths taken from tyre manufacturers literature. The calculated figure will be a small underestimate because the error of tracking has not been included. Assuming auto-steer based on an RTK correction signal, this error will be a constant of around ± 20 mm. Thus, for the narrowest equipment considered here (3 m) the error will be of the order of 1%. The inclusion of arable operations assumed that the grass would be ploughed out with a larger tractor introduced for this and subsequent operations; if this was considered necessary. The tracking associated with ploughing was not included but to reduce compaction, particularly that in the furrow, the recommendation would be to use the autosteer system to achieve on-land ploughing rather than in the furrow. Chemical application equipment was not considered as this would run within the tracks of other equipment. This report is complemented by an Excel workbook which provides more detail about the equipment proposed for each system ( Documents/Dairyco%20project.xlsx). 29

30 3.3 CTF Designs Multiple operating widths from 1.5 m to 9.0 m The basis of these systems are tractors with a 1.5 m track gauge. These can use either a nominal 1.5 m rear-mounted offset mower (e.g. Lely Splendimo 165 Classic leaving a swath width of up to 1.4 m) or a similar 3 m mower and a three swath (9 m) tedder and swather (narrower versions of these machines could be used providing they operate at 1.5 m, 3 m or 6 m). Potentially this system would work with little or no modification of the equipment being used, but it does rely on the track gauge of the tractor used for mowing being equal to or half the width of the mower. To maintain the tracking with this system while mowing, all work must be carried out in lands working either outwards from an initial cut or inwards from two cuts, both in virgin grass. The second swath must be precisely positioned for everything to match up when the land is finished, as indicated in Figure 3.1. Figure 3.1. Illustration of a tractor on 1.5 m track gauge operating with a 3 m mower and 9 m tedder and swather. A second tractor on 1.8 m track gauge is used with a loader wagon and to pull a slurry tanker having a 6 m trailing shoe. 30

31 This is similar to normal practice except that with an RTK auto-steer system positioning of the second cut across the field should be precise, easy and more efficient than manual selection of the cut position which can leave part widths. A minor issue is the fact that the tedder s working width is greater than the width at which it will be operating and it will therefore be offset by around 0.48 m from the centre of the three rows but the manufacturer did not consider that this would be an issue. The reality is that the tractor operating the tedder will be running on cut grass rather than the small gap between swaths. The equipment used in this hypothetical scenario is listed in Table 3.1. The larger of the two tractors is used to operate the loader wagon and the slurry tanker, both of which come within each other s footprint but slightly outside that of the smaller tractor. With this set up and tyre sizes, the trafficked area is approximately 36%, assuming all vehicles are auto-steered to within ± 20 mm. The Claas Volto 64 operated at 6 m with the Elios, could take the place of the Volto 1100 without affecting the trafficked area. Table 3.1. Machinery assumed for a CTF operation illustrated in Fig. 1. Trafficked area with grass production is 36%, rising to nearly 40% if arable cropping is introduced. Machine Working width (m) Operating width (m) Tyres on principal axle Track gauge, (m) 76 kw Claas Elios 240 tractor 420/70 R kw Claas Arion 650 tractor 20.8 R Claas Disco 3150 offset n/a n/a mounted mower Claas Volto 1100 tedder n/a n/a Claas Liner 2900 swather n/a n/a Claas Cargos 8300 loader / wagon Abbey 900 slurry tanker & R trailing shoe 3 m or 6 m cultivator 3/6 3.0 or 6.0 n/a n/a 3 m or 6 m drill 3/6 3.0 or 6.0 n/a n/a John Deere 9660 combine /75 R Bailey TB12 grain trailer /65 R The horizontal line in the table divides the forage from the arable operations The loader wagon could be replaced by a trailed forage harvester and separate tractor and trailer combination, as shown in Figure 3.2. The benefit of this combination (assuming that the tractor and trailer are running in the prescribed wheel tracks) is that the larger tractor (Arion 650) would not be required because the forager could easily be powered by the smaller tractor. The disadvantage is that at least one of the forager wheels will be running on a non-trafficked bed but it is a relatively light machine ( kg per wheel) and may therefore not inflict too much damage. Figure 3.3 shows that with careful planning, one of the wheels of the forager can be made to run within the wheel track of the tractor by hydraulic adjustment of the drawbar. The tractor and trailer could run on the adjacent tracks at 3 m, 4.5 m or 6 m from the tractor driving the forager because loading can be achieved at up to at least this distance (Jim Hunter, Lely personal communication). 31

32 There is an option to hitch a trailer directly to the forager, but this would be unacceptable because it means that the trailer wheels would be running on the bed and would also put significant extra load on the forager wheel or wheels which are also running on the bed. The trafficked area for this system is around 34% reflecting the small advantage of not having to use the Arion tractor to power the loader wagon. Figure 3.2. Example of trailed forager and side loading to a trailer Figure 3.3. The Lely Storm shown here has an hydraulically adjustable drawbar that can alter the position of the forager in relation to the tractor as shown on the insert (Photo: Jim Hunter, Lely). Equally, a baler and in-line bale wrapper could be employed in place of a forager. These implements tend to have similar track gauges to tractors so would not increase the trafficked area. An offset wrapper would not be appropriate because this would place the tractor in the wrong place. Picking up the bales which are dropped in the centre of the traffic lane could almost certainly be loaded from the adjacent traffic lane with this system that has traffic lanes 1.5 m apart. 32

33 Another alternative to the loader wagon is a self-propelled harvester delivering to a towed trailer, as shown in Figure 3.4. However, because of the larger axle gauge and tyres fitted to the harvester, the trafficked area for this system would be around 47%. As will be seen from the video shown in this clip ( there is also some constraint on output because if controlled traffic is to be maintained, trailers must be swopped over on the headlands. These systems could be integrated with a mixed arable/livestock farm having 9 m cultivators and drills without any effect on trafficked area because the two tractors have already been accounted for. However, if a 6 m wide cultivator and/or drill were used, the trafficked area would rise to 39% because the larger tractor was now being used every 6 m rather than every 9 m. If a combine harvester, operating at 9 m, was introduced, because of the wider tracking of this machine, the area covered would increase to close to 50%. If a 6 m cut was assumed, the trafficked area, as in the case of the 6 m cultivator, would rise to around 58%. If the straw is to be baled (which is likely on a livestock farm) rather than chopped and spread, the swath may need to be moved to align it with the pick-up system as the swath is laid over a wheel track (Figure 3.1) rather than directly between a pair of tracks; which is where it needs to be. Figure 3.4. Example of a self-propelled harvester loading into a towed trailer. As with the forage system, the bales may need to be moved if they are to be picked up by a chaser or chased to a headland with an adapted front end loader. 33

34 3.3.2 System with 4 or 8 m operating widths In this system a semi-mounted 4 m mower produced by Massey Ferguson (DM 408), (which would have just one supporting wheel on the non-trafficked bed) was considered. Details of the equipment used are shown in Table 3.2. The forage only system resulted in a trafficked area of just under 28% but this increased to nearly 36% when an arable system was introduced with a combine having a 4.57 m cutting platform used at 4 m. Table 3.2. Equipment for a controlled traffic system based on a 4 m semi-mounted mower. Trafficked area with grass production is 27.7%, rising to around 36% if arable cropping is introduced. Machine Working width (m) Operating width (m) Tyres on principal axle Track gauge (m) MF kw tractor 460/85 R Case Puma 140 kw tractor 580/70 R MF 408 semi-mounted mower MF TD 868 DN tedder MF RK 802 TRC swather Lely Tigo S loader wagon / Conor 1850 slurry tanker 8 550/60 R with trailing shoe Väderstad Carrier L cultivator Väderstad Spirit 400 S drill New Holland CX7.90 combine /75 R The horizontal line in the table divides the forage from the arable operations Although an operating width of 8 m could be considered for the combine, the nearest actual cutting width of these machines is 9.14 m. Running this at 8 m might not be acceptable in the long term, but despite its wider tyres and greater track gauge, it only increases the trafficked area to 40%, or to 37% if used with rubber tracks. A trailed forage harvester was also investigated for this system but found to be incompatible if a tedder and swather are used (Figure 3.5). The swath needs to be moved about 1.8 m nearer to the forager, which according to one manufacturer, would need some implement modification that might render it unstable. An alternative would be to let the cut swath wilt where it is. However, a swather would be needed to reduce its width from around 3.3 m to the 1.8 m needed for the pick-up of the forager and to bring it closer to the tractor. A baler and in-line wrapper could also be used but unlike with the 1.5 track gauge system discussed in section 3.3.1, picking up the bales would be problematic; unless they could be dropped or moved around 1.8 m to the side (see Figure 3.5). 34

35 Figure 3.5. Illustration of the problem associated with trying to use a trailed forager and tedding and swathing for a 4 m system. The swath needs to be moved 1.8 m to the left for the tractor to be in the right place System with 5 m operating widths In this category a 5 m system operating with tractors on standard gauges is proposed. Equipment at 5 m width is more limited in availability but a full range of machines was eventually found and are listed in Table 3.3, along with the tractors to operate them. 35

36 The machines required for forage production listed in the first part of the table lead to a trafficked area of about 22%, but this rises to 32% when the arable operations are introduced, largely due to the greater gauge of the combine harvester. If a combine with double the width of cutting platform is used, this has little effect on the trafficked area because of its larger tyres and greater gauge between them. The situation is very similar if rubber tracks are introduced on the combine. Table 3.3. Equipment used to create a 5 m controlled traffic system based on an OutTrac (refer to Chapter 2 Fig. 2) design. Trafficked area with grass production is 22%, rising to 32% if arable cropping is introduced. Machine Working width (m) Operating width (m) Tyres on principal axle Track gauge, (m) 97 kw MF 5613 tractor 460/85 R kw T7.210 FordNH tractor 650/65 R Lely 550 P Splendimo mower MF TD 524 DN tedder MF RK 662 TRC swather Claas Cargos 8300 loader wagon Swath / Conor 1850 tanker with trailing /60 R shoe New Holland CX720 combine /75 R kw Ford New Holland T /65 R tractor Simba/Great Plains X-Press VX Kongskilde Vibroseeder VS 500 H The horizontal line in the table divides the forage from the arable operations A shortcoming of the mower in this system is the fact that it weighs 2.86 t and this load is carried by two 15.0/ ply tyres which run on the so called non-trafficked bed (Figure 3.6). The potential damage from these might be reduced if they could be replaced by wider radial ply tyres, but this would require some modification to the machine and would impact on the present 2.6 m transport width. A TwinTrac system was also assessed (effectively combine tracking at 3 m track gauge), but was only an advantage if the arable system was Introduced, as the trafficked area then fell from approximately 31% to around 27%, compared with the equipment in Table 3.3. As with the 4 m system, use of a trailed forage harvester leads to even greater difficulties with picking up of the swath and delivery of the material to trailers. In this case it might make more sense to wilt the swath where it is and then bring it inboard and narrow it with a swather. 36

37 Figure 3.6. The 5 m Lely Splendimo P showing the wheels supporting this 2.8 t mower when in work System with 9 m operating widths The equipment for this system is governed by the operating width of triple gang mowers. These consist of front mounted units complemented by two rear-mounted offset mowers, as illustrated in Figure 3.7. Figure 3.7. Triple gang mower of the type envisaged for wider CTF systems. The Kuhn GMD 9530 has a working width of between 9.13 and 9.53 m, and would be run at 9 m. Tedding of the swaths to assist drying could be with a Kuhn GF TGII or a Claas Volto for example, both operated at 9 m. Alternatively, some mowers in this width range offer simultaneous conditioning and spreading of the material over the entire working width (e.g. Claas Disco 9200 or 9400), saving an additional operation. With these larger cutting widths, the aim will be to select tractors with a track gauge as wide as possible to match both grass and grain harvesters. Subsequently, if the material has been spread, a 9 m swather would be used to present a single swath to the forage harvester. In this case, a self-propelled machine was assumed, 37

38 but as before, delivering to a rear-hitch trailer because there is no technology at the moment to deliver 9 m to the side. The main choice of harvesters lies between John Deere and Claas; and in this instance, the Claas machine is preferable because it has a narrower track gauge to more closely match that of a middle range tractor (250 hp, 186 kw) at around 2.0 m. A 3000 litre Conor slurry tanker is assumed with a 9 m trailing shoe. The operations associated with this system are shown in Figure 3.8 which amount to a trafficked area of just over 18%. Mower Tedder Swather Forage Harvester Trailer Slurry tanker Figure 3.8. Series of machines and operations involved in a 9 m controlled traffic forage grass operation, including slurry spreading. Trafficked area for this system and machines listed in Table 3.3 is around 18%. 38

39 Assuming a mixed farm with some arable, a combine harvester with a 9.14 m (30 ft) cutting platform was chosen and operated at 9 m. In this instance, a John Deere 9000 series combine has one of the narrowest track gauges (2.61 m) but the New Holland CX 820 comes a close second at 2.62 m. As cultivators and drills of this width are less common, some modifications might be needed, but Agrisem make a 9 m stubble cultivator and drills of this width are now being offered, e.g. Dale Eco-Drill and the Amazone Primera or Condor. As the combine has a larger track gauge than all the other equipment, the trafficked area for the system including arable operations rises to 22%. The machinery line-up for this CTF system is listed in Table 3.4. Table 3.4. Machinery assumed for the 9 m CTF system illustrated in Fig. 8. Trafficked area with grass production only is around 18%, rising to 22% with the introduction of arable cropping. Machine Working width (m) Operating width (m) Tyres on principal axle 1 Track gauge (m) Claas Axion tractor n/a n/a 710/70 R Kuhn GMD 9530 triple gang mower Claas Volto tedder Kuhn GA 9531 swather Claas Jaguar 980/ /70 R harvester Richard Western silage n/a x trailer PR Conor slurry tanker /60 R trailing shoe Agrisem Vibromulch stubble cultivator Dale Eco-Drill John Deere 9000 series /75 R combine Grain chaser /40 R Tyres on the axle that creates the widest footprint for the particular machine The horizontal line in the table divides the forage from the arable operations System with 12 m operating widths As with the 9 m system, grass cutting operations are based on triple gang mowers. The particular machines identified are listed in Table 3.5 and provide a trafficked area of just over 13% with a grass only system and just over 18% when arable operations are introduced. This system is probably one of the easiest to put together and some growers may be using equipment of this size already. It will be noted that slurry application in this instance is with a dribble bar rather than a trailing shoe and is simply because a 12 m trailing shoe did not seem to be available. However, one could probably be found or made up if the dribble bar were an issue. 39

40 Table 3.5. Machinery assumed for 12 m CTF system. Trafficked area with grass production was just over 13% rising to just over 18% with the introduction of arable cropping. Machine Working width (m) Operating width (m) Tyres on principal axle 1 Track gauge (m) John Deere 7210R tractor n/a 710/70 R JF GXT triple front and trailed mower Claas Volto 1300T tedder Claas Liner swather Claas Jaguar 980/970 n/a /70 R harvester RW SF20 triple axle trailer n/a x ply 2.0 SlurryKat 16,000 L tanker /65 R together with dribble bar John Deere 8320 R tractor n/a 710/70 R Horsch Joker 12T disc cultivator Dale 12 m Eco-Drill Case Axial Flow VF710/70 R combine harvester RW SF20 triple axle trailer n/a x ply 2.0 The horizontal line in the table divides the forage from the arable operations 3.4 Summary of designs and discussion Table 3.6 provides an overview of the different designs proposed, together with the trafficked areas associated with them. Of the narrower systems, the one using 4 and 8 m widths stands out as being the most effective. However, as some farms might find a 4 m cutting platform rather narrow for the arable part of the operation there is the option of using a 9 m platform at 8 m without too much penalty in terms of trafficked area. For grass only, the 3 m system does reduce tracking significantly compared with traditional practice (36% compared with 80%), but the economics of the guidance system needed to achieve this would need to be considered as to whether there would be any net return purely on the grass yield (see Chapter 5). To reduce the trafficked area for the larger 9.15 m (8 m operating width) combine, an alternative is to have one on rubber tracks. Most of these reduce the track gauge to around 2.8 m and as can be seen, gives a small improvement on the wheeled equivalent. This machine may be beyond the scope of a farm operating equipment of this size, but could be brought in by a contractor. A combine equipped with rubber tracks was also considered for the 12 m option, but there was no advantage in terms of trafficked area because a wheeled option was no wider, and the internal width was already tracked by other vehicles. 40

41 Wider CTF systems do of course track a smaller area and in Denmark both 12 m and greater widths are used regularly; being embraced by both farmers and contractors (Appendix I). At these widths, most operators use self-propelled harvesters and most trail loader wagons behind them, suggesting that the need to drop these off on the headland is not considered a serious impediment to the adoption of these systems. An interesting introduction at this year s Scotgrass 2016 was a m triple mower from Claas, opening the way to a 10 m CTF system that could be fully integrated with an arable operation of the same width. Tracked area would lie somewhere between the 9 m and 12 m systems. Table 3.6. Summary of the different CTF systems proposed and their associated tracked areas Controlled traffic base widths Trafficked area, % Mowing width (m) Other widths (m) Forage system Grass only Grass & arable 1.5 or & 9.0 Loader wagon (6 m combine) 49.7 (9 m combine) Trailed forager (6 m combine) & 9.0 Self-propelled (6 m combine) 44.1 (9 m combine) Loader wagon (4 m combine) 40.5 (8 m combine) 37.0 (8 m combine) Loader wagon Self-propelled Self-propelled See Excel spreadsheet for other options and trafficked areas 2 Combine on rubber tracks Although every attempt has been made to provide accurate data, most have been acquired from internet-based brochures and technical literature that are open to misinterpretation or lacking in detail. However, what this study does show is that forage production systems using controlled traffic are certainly feasible and without the need for a great deal or any modification to the machines. Some of the systems proposed here could almost certainly be improved if all feasible setting aspects are explored on all the machines involved. As with arable systems, farmers need to embrace the idea and then plan future machinery acquisition accordingly. In very few instances is it feasible to make a complete change to CTF overnight, it is a process of planning and gradual transition until all compatible equipment becomes available as part of the normal machinery replacement policy. Such a process would be greatly helped if manufacturers literature provided the few extra data needed for farmers to plan their systems. 41

42 Due to the lack of web-based data from some manufacturers, there has been some natural selection in that their machines have not been considered because not enough information was available. 42

43 4. Assessment of the Effect of Controlled Traffic on Grassland Dry Matter (DM) Yield for Three Silage Cuts Introduction In the UK grassland fieldwork is generally conducted in an ad hoc manner with no conscious attempt to re-use equipment wheel ways or pathways in the way they are used in arable fields, such as tramlining or controlled traffic farming (CTF). Two passes from tractor compaction in autumn field conditions has been shown to reduce yield of the following years 1 st cut by between 16 and 25% (DairyCo Report, 2015). Research has shown that in a single year 90% of a field can be covered by tractor/harvester/trailer wheels at least once (University of Nebraska, 1999) with a number of areas within the field repeatedly trafficked. More recently, pilot studies at Harper Adams University (HAU) in 2012, showed approximately 65% coverage during a single grass harvesting operation for both forage chopper and baling operations. A current study with winter wheat at HAU has shown a significant (19%) yield improvement by controlling field traffic to reduce soil compaction from field operations (Smith et al, 2014). Both silage yield and operational cost reduction benefits from controlled traffic have been shown with grassland in Denmark (Kjeldal, 2013) Objectives This study aimed to: To devise and demonstrate an effective Controlled or Reduced Traffic Farming system for grassland. To determine the effect of repeated traffic of a traditional management system on the within-field variation in grassland silage yield through a season. Determine the potential economic and environmental benefits of CTF (or Reduced TF) systems Materials and Methods Experimental site An 8 ha field at the SRUC, Crichton Royal Farm, Dumfries, Southwest Scotland (N55:02:45, W03:35:56) that consisted of two sandy loam soil series (Crichton - freely/moderately drained and Midpark freely drained) was split into two halves of 3.5ha. The management history of the grass sward was the same up to the point the field was split on the 16 th April The field had been re-seeded with perennial ryegrass (Lolium perenne) during late summer of 2014, after tillage to minimise carry-over of historic compaction damage. The layout of each field area was the same rectangular shape (218m x 160m), including an entrance; these were located at the bottom right hand corner of the normal traffic field and the top right hand corner of the controlled traffic field (Figure 4.1). There was a discard of 10m between the two field areas to allow movement of machinery. 43

44 Figure 4.1. Normal (N) and Controlled Traffic Farming (CTF) field layout. One of two traffic management treatments were imposed on each area: Tradition or Normal traffic (N) Controlled traffic (CTF) Field management The field was managed as a three cut silage system in May, July and August (Table 4.1). Silage was harvested via a three beam chopper with a central beam of 3m with 3m beams either side (Figure 4.2) and outside wheel width of 4.15 m, the overall working width was of 9 m. Following the mowing, spreading and raking operations were done with a double windrower rake to bring the three cut rows into one central row. The cut grass was harvested the following day with a forage harvester of wheel width to outer edge of 3.1 m in to two tractor trailer combinations (tractor wheels outer width 3.2 m) running alongside the harvester (approximately 2 m) for the N and along the next set of parallel chopper wheelings for the CTF (Figure 4.3). Dairy cow slurry was applied at a rate of 30 m 3 ha -1, at twice through the season (May and July (Table 4.1). Slurry was applied via a slurry tanker and splash plate (wheel width to the outer edge 2.9 m) the second application in July could only be applied with a narrower application band as a result of the thicker consistency of the slurry. The CTF management pattern had to follow the one of the wheelings and add a further wheeling to each side of the 44

45 original pattern. Inorganic fertiliser, as urea, was applied to both the N and CTF fields at a rate of 60 kg N ha -1 after the first cut and followed the N or CTF systems (outside wheel width 3.0 m). Table 4.1. Dates of operations on the fields during Date Operation 20 May 1 st silage cut 22 May Rake 22 May Harvester and tractor and trailers 26 May Inorganic fertiliser application 1 June Slurry application 1 July 2 nd silage cut 2 July Rake 2 July Harvester and tractor and trailers 3 July Slurry application 20 Aug 3 rd silage cut 21 Aug Rake 21 Aug Baler and tractor and trailers All the activities on the CTF field were based on a 9m operating width; this was based on the width of the cutter. The wheel width was based on 3m and where ensured through the use of both a GPS system with a 15cm pass to pass accuracy and sight posts set up at the ends of each A-B line to allow for a manual check on the accuracy. The operations on the N field were done first to reduce any bias of the machinery operators from any of the controlled movements. Each field was cut around the headlands first three times and then the A-B lines completed using the cut headlands for turning (Figure 4.4). 45

46 Figure 4.2. Working width and wheel width as set down by the chopper a) Normal (N) b) Controlled Traffic Farming (CTF) Figure 4.3. Area covered from a) N with new area covered by tractor and trailer and b) CTF with tractor and trailer following chopper wheelings. 46

47 Figure 4.4. Controlled traffic pattern across the field Baseline measurements Forty initial measurements points were set up (20 th April 2015) on a grid system across both the treatment fields once the fields had been marked out and before the first silage cut. Measurements of penetrometer resistance, soil bulk density (0-10 cm) and a soil visual evaluation of soil structure (VESS) were completed at the sampling point to confirm the uniformity of soil structure before the start of the first silage cut. At the same time, soil samples were taken from each sampling point to assess the phosphorous (P) and potassium (K) levels along with the ph across the field, to determine any variation that could contribute to yield prior to management during the season. These initial locations were fixed using GPS Grass harvesting RTK GPS tracking of all vehicle movement across both treatment fields (N and CTF) during grassland management i.e. inorganic fertiliser, slurry application and harvest were recorded. The use of the Real Time Kinematic (RTK) systems give adequate year-to-year accuracy and enable operators to return to the same location in the field +/- 25 mm repeatedly without the need to re-locate from a fixed bench mark in the field. The intensity and spatial coverage of traffic across each treatment area for each field operation were identified from the RTK traffic maps. Forty 2 x 2 m 2 plots were superimposed across each treatment field and these were related to the number of times these locations within the field were trafficked. Within the N and CTF field, eight of the 2 x 2 m 2 plots were identified which had received zero number of vehicle passes following the first management operation (1 st silage cut). The number of individual 47

48 passes at a particular point were determined from the GPS data collected and eight plots were established at four further levels of traffic intensity depending on the number of passes to produce a distribution curve. Overall, eight plots were established at locations identified as having received: 0 passes 2 passes 4 passes 6 passes 6+ passes This process was repeated for each field (N and CTF) ahead of each subsequent silage cut (2 nd and 3 rd silage cuts) to reflect an accumulating number of passes over the season. Grass dry matter yield was determined for each of these sampling points and harvested with a Haldrup harvester (Haldrup Ltd, Germany) with a balance incorporated and cutting blades set to give an aftermath of 60mm. The harvester directly cut and provided a fresh weight of grass on each plot. A sub-sample of the fresh sample was taken for dry matter analysis. Dry matter analysis was completed by oven drying at 105 o C for 48 hours. These forty areas were cut immediately prior (the day before) each of the silage cuts. Total yield off-takes from both fields were recorded at each harvest using a static weigh bridge at the farm building. As each full trailer came off the field it was weighed on the weigh bridge and then taken to the silage pit. A weight of the empty trailer was used to calculate the total weight of fresh grass in each trailer. Ten grab samples for grass DM content were taken from each trailer after it has been weighed as it was emptied into the silage clamp Soil measurements and analysis Soil sampling was completed with a 30 mm diameter soil auger down to a depth of 10 cm. Three soil cores were taken from each sampling point to ensure sufficient soil for analysis. The soil samples were placed in a plastic bag and sealed and stored at 4 o C until analysis. The fresh soil samples were analysed for P (modified Morgans (McIntosh, 1969)), K (modified Morgans (McIntosh, 1969)) and ph (in H 2 O (Robertson et al, 1999)). Soil penetrometer measurements of soil resistance at all the sampling points across the two field areas down to a maximum of 25 cm were made with a cone penetrometer with a 20mm base, size 3 cone and built-in data-logger (Eijkelkamp Soil & Water, Netherlands). The soil visual assessments (VESS) were undertaken with the method outlined by Ball et al, (2007) on soil blocks of approximately 25 cm depth. Soil moisture across the two field areas were assessed from soil cores taken with an 30 mm diameter auger down to a depth of 10 cm before each management operation was undertaken i.e. silage cutting or fertilizer application. A final set of soil penetrometer and VESS measurements were taken from the 40 areas cut on each field area after the 3 rd silage cut in November 2015 as the soil moisture increased and conditions became suitable for the VESS assessment. 48

49 Statistical analysis Data were analysed using Genstat version 16 (VSN International, Hemel Hempstead). The main treatments of tractor passes and N compared to CTF were analysed on a randomised basis using Genstat ANOVA. Any significance was investigated with a student t-test at a level of significance of P< Results Baseline measurements The baseline physical measurements taken at the 40 sampling points in the grid across the fields, N and CTF, showed that for the soil structural component the soil bulk density, soil resistance and visual evaluation of soil structure (VESS) there were no significant differences between the two fields (Table 4.2). The chemical assessment of the soil at the same sampling points of the grid across the two fields did produce a significant difference (P<0.001) between the N and CTF for soil extractable phosphorous (All results shown in Appendix II, Table 1). Table 4.2. Mean baseline measurements, taken before the treatment wheelings, for soil bulk density 0 to 10cm (g cm 3 ), soil resistance (penetrometer; 0-20cm) (kpa), Visual Evaluation of Soil Structure (VESS) (Score 1 to 5), soil extractable phosphorous (P) (mg l -1 ) and potassium (K) (mg l -1 ) and soil ph. Measurement Field s.e.d. P value Normal CTF Soil bulk density CV% Soil resistance CV% Visual Evaluation of Soil Structure (VESS) CV% Soil extractable phosphorous <0.001 CV% Soil extractable potassium CV% Soil ph CV% However, these statistics were done on the raw data of the analysis from the individual sampling points. If these points were reported on the SAC scale or on the Defra index they gave moderate to moderate + for the SAC scale or 2 for the Defra scale and would indicate 49

50 that the addition of extra P was unlikely to provide any additional growth response to grassland (PDA, 2011). The potassium levels measured across both fields would also be moderate or Defra index 2 and would also have the recommendation that further additions of K would be unlikely to provide any additional grassland yield (PDA, 2011) Effect on the soil from vehicle passes The increased number of vehicle passes increased the soil compaction in both the N and CTF fields. This was seen from the increased measurements for the soil bulk density with an increase of 14.7% (P<0.001) for the N field and 18.2% (P<0.001) for the CTF field between the zero passes and 6+ vehicle passes (Table 4.3). These increases were also seen from the penetrometer measurements with a 48% (P<0.002) increase for the N field and a 70% (P<0.001) increase for the CTF. There were significant increases (P<0.001) in the VESS score between the zero and 6+ passes for both the N and CTF fields. Table 4.3. Effect on soil physical measurements of soil bulk density 0 to 10cm (g cm 3 ), soil resistance (penetrometer 0 to 20cm) (kpa) and Visual Evaluation of Soil Structure (VESS) (Score 1 to 5) from the number of vehicle passes. No of passes Soil Bulk Density (g cm 3 ) Soil Resistance (kpa) VESS (Score 1 to 5) N CTF N CTF N CTF The soil bulk density measurements indicated that the greatest single increase in bulk density compared to the overall change was the difference between the zero and 2 passes with a 54% increase for the N field and a 45% increase in the CTF field of the total bulk density. The CTF field did show a greater increase of the total increase in soil bulk density between the 6 and 6+ passes (23.8%) compared to the N field (9.1%) Dry matter reduction related to vehicle passes The first silage cut for both the N and CTF fields did not benefit from the controlled traffic movement as these were established with the first silage cut. Therefore the yield from both fields should have been similar (P=0.20) as they had been subjected to the same management; this was the case from the 40 grid points sampled for yield across the two fields, with the CTF only being 0.24 t ha -1 greater than the N field (Table 4.4). The 40 small cut samples taken across the two fields prior to the second and third silage cuts were taken from areas of the field that had been subjected to a varying number of passes 50

51 from the machinery (0 to 6+ passes). The increasing number of passes decreased the DM yield for both the silage cuts (Figure 4.5). Table 4.4. First silage cut Dry Matter (DM) yield (t ha -1 ), before the controlled traffic system was employed. 1 st Silage Cut N CTF s.e.d. P value DM Yield (t ha -1 ) There was a 19.5% reduction in yield for the 6+ passes compared to the 0 passes for both the N (0.66 t ha -1 ; P<0.01) and a reduction of 29.9% for the CTF (1.04 t ha -1 ; P<0.01) with the second silage cut and a reduction of 36.7% for the N (0.83 t ha -1 ; P<0.01) and a reduction of 38.2% with the CTF (0.87 t ha -1 ; P<0.001) for the third silage cut (All results shown in Appendix II, Table 2). Figure 4.5. Second and third silage cut Dry Matter (DM) yields (t ha -1 ) after the two traffic management systems were imposed (Normal traffic (N) and Controlled Traffic Farming (CTF)). The 4 passes (0.48 t ha -1 reduction (P<0.05)) and 6 (0.61 t ha -1 reduction (P<0.05)) passes showed a significant reduction in DM yield from the 0 passes for the CTF for the second silage cut and the 6 passes a significant 0.60 t ha -1 reduction (P<0.05)) for the third silage cut compared to the 0 passes Dry matter total off-take As expected there was no significant difference between the two fields for the first silage cut, as the systems had not been set up prior to this cut and all operations had been the same 51

52 across both fields. The difference in yield between the two fields was only 0.15 t ha -1 (P=0.27) (Table 4.5). There were increasing differences in yield between the two fields as there were more traffic movements across the N field compared to the CTF field as management operations continued for the 2 nd and 3 rd silage cuts. The 2nd silage cut gave a 0.30 t ha -1 increase in DM yield for the CTF compared to the N field (P=0.72) and a significantly increased DM yield (0.5 t ha -1 (P<0.01)) for the 3rd silage cut with the CTF field compared to the N. Table 4.5. Total Dry Matter (DM) off-take and differences (t ha -1 ) from the two field systems, normal (N) and controlled traffic farming (CTF). Silage cut Field system Difference s.e.d. P value N CTF 1 st silage cut nd silage cut rd silage cut nd +3 rd silage cut Total silage Although the DM yield increased by 8.5% for the CTF compared to the N field (0.96 t ha -1 ) over the three silage cuts the effect of the CTF system would only have been for the second and third silage cuts. When the combined yield of the second and third cuts are compared this gave a 13.5% (0.80 t ha -1 ) increase for the CTF field compared to the N field a significant increase after only 2 silage cuts indicating a relatively rapid increase in yield Difference between CTF and N in area covered The use of the GPS in-cab tractor systems and the details of the working widths and the tyre widths of the machinery involved in the management operations across the two fields allowed the calculation of the area covered by the wheelings and the overall distances covered (Table 4.6). The area covered in the CTF field (30.4%) was reduced by 57% compared to the coverage of the N field (87.4%) over the whole of the three silage cuts. This was as a result of the wheelings for the N field not following set areas and covering areas of the field outside the original area covered during the first silage cut. This was especially true of the movement of the tractors and trailers moving the cut silage to the clamp. In the N field, the full trailers would be driven directly towards the exit gate and on returning would drive across the field to fall in line behind the tractor and trailer being loaded, ready to take over. As the tractor and trailers that collected the cut silage from the N field also did not follow the next set of wheelings, as with the CTF system, but ran between the wheelings, this ensured a greater area of the field was covered (22.2%). The CTF system set up the wheelings with the first operation of cutting with the chopper and accounted for the same area cover as with the rake and harvester as these followed the same wheelings (19.4%). The tractor and trailers collecting the cut silage and transporting it 52

53 to the clamp also followed the same wheelings, even if they had to travel further to reach the exit gate. Unfortunately, the slurry applied between the second and third silage cut was much thicker than that spread between the first and second silage cuts and would not spread as far, therefore a reduced operating width was used to ensure coverage across the field and the wheelings had to be off-set with one new wheeling being produced for every A-B line this resulted in further area of the field being covered in the CTF field (11.0%). Table 4.6. Area of the field covered (%) and total distance covered (km) for the field operations on the normal (N) and controlled traffic (CTF). Operation Area covered (%) Distance covered (km) N CTF N CTF 1 st silage cut Cutting, Rake & Harvester Tractors & Trailers Slurry application Tractor and Tanker nd silage cut Cutting, Rake & Harvester Tractors & Trailers Slurry application Tractor and Tanker rd silage cut Cutting, Rake & Harvester Tractors & Trailers Total The reduction in the 2 nd and 3 rd silage operations in N field area covered were as a result of some of the field having already been covered during the previous cutting and slurry applications. The operations in the CTF field covered a further distance as the equipment were confined to the set wheelings and would use these, especially the ones that were round the inside perimeter of the field, to move into and out of the field gate Conclusions Even though there was uniformity across both fields prior to the first silage cut by the end of the third silage cut there was a significant difference between the two field for DM yield. Although the CTF field produced 0.96 t ha -1 more DM yield than the N field over the three silage cuts only the 2 nd and 3 rd cuts were subjected to true CTF. There was a benefit of 13.5% DM yield for the CTF system over current N practice coinciding with the literature 53

54 (Douglas et al., 1995) and in the economic model (Chapter 5). This was a relatively rapid increase in yield after only two silage cuts with the CTF system. The number of passes for 6+ significantly reduced the DM yield for both the N (by 20%; P<0.05) and CTF (by 30%; P<0.01) for the second silage cut. This reduction in DM yield was also significant for both fields for the 6+ passes for the third silage cut N (by 37%; P<0.01 and CTF (by 38%; P<0.01) compared to the 0 pass areas. The use of existing equipment together with vehicle auto-guidance reduced the CTF traffic area to 30.4% compared with 87.4% across the N field. The area covered for the CTF field would have been reduced further to 19.4% if the second slurry application could have had a larger spread width. A less thick consistency would have allowed a greater working width and the same pattern of wheelings would have been maintained. The only penalty associated with CTF was the greater distance (26%) traveled by the field machines as they were fixed into the set pattern whereas the machines in the N field were allowed to travel directly to and from the gate. 54

55 5. Grassland Controlled Traffic Farming (CTF) Economic Analysis 5.1 Introduction This analysis follows a similar format to that used by Godwin et al. (2003) when assessing the potential for precision farming in cereal production for the Home Grown Cereals Authority (Report No. 267). It considers the direct economic advantages from any improvements in forage yield alongside the additional costs of so doing. It does not include other less tangible benefits such as the savings in re-establishment costs and fuel savings referred to in Chapter 2. A range of trafficked areas (%) was used in these calculations, as these will depend upon the operating track gauges and wheel widths of the available equipment as discussed in Chapter 3, thus enabling a farmer and/or a contactor to estimate the benefits for their system. It also enables an individual farmer and/or contractor to estimate future benefits with the longerterm investment in replacement equipment. The cost of CTF is based upon four different levels of capital investment, as the level of expenditure will depend upon the type and scale of the operation. The additional costs are focused on the costs of the guidance systems only, as it is assumed that initially a farmer or contractor would use existing equipment (as demonstrated in Section 4) and that improvements to equipment matching would be part of the normal longer-term replacement policy. In practice multiple guidance systems would be required to support the harvester and other associated field equipment and management operations. Therefore in the analysis the costs of each individual guidance system are estimated, from which the costs of multiple guidance systems are considered. 5.2 Yield Benefit Detailed analysis of the most robust UK data reviewed earlier, namely Douglas et al. (1992 and 1995) and Frost (1988a and 1988b) indicated that the removal of wheel traffic increased the average grass yield by 14.7% and 12.1% respectively. Hence for the purpose of this study a mean potential yield increase of 13% was selected. From which the actual yield benefit for a range of reduced trafficked areas from a maximum of 45% to a minimum of 15% could be calculated and is shown in Table 5.1. These data were calculated by assuming: 1. The average forage dry matter yield for 2 and 3 cut systems per year in the UK is 12 t ha -1 and 16.5 t ha -1 (mean of the range t ha -1 ) respectively (Hargreaves - SRUC, 2016). 2. The fields are currently managed with random traffic patterns covering 80% of the field. Figures 2.1 and 5.1 illustrate the trafficked area from a single cut of silage. 55

56 The figure of 80% was chosen given the recorded data of the trafficked area for a single harvest operation of 65% from Kroulík et al. (2014) and that during a growing season where there could be in excess of 11 multiple operations (Chapter 2). It is higher than the 74% chosen by Alvemar (2014) and for the purpose of the analysis more conservative than assuming a total (100%) trafficked area. These assumptions result in calculated yield increases from 0.53t ha -1 to 0.99t ha -1 and 0.73t ha -1 to 1.36t ha -1 for 2 and 3 cut systems respectively. Figure 5.1 Illustration of the impact of vehicle traffic during and one week after forage harvest. (Courtesy: Chamen, CTF Europe). The above yield increases were estimated by adding the trafficked and non - trafficked yields given by the following equations in proportion to the trafficked and non-trafficked areas. (Where the trafficked areas and yield benefit are expressed as proportions rather than %). 1. Trafficked yield: Yt (t ha -1 ) = Average forage dry matter yield for the UK (t ha -1 ) ((Trafficked Area + (1 Trafficked Area)(1 + Average potential yield benefit)) 2. Non-trafficked yield: Yo (t ha -1 ) = Trafficked yield (t ha -1 ) x (1 + Average potential yield benefit) Hence for a 2 cut system with an average forage dry matter yield of 12t ha -1 and a potential yield benefit of 13% the trafficked yield is: and the non-trafficked yield is: Yt = 12 (t ha -1 ) = t ha -1 (0.8 + (1-0.8)( )) Yo = t ha -1 x ( ) = t ha -1 Which gives a combined yield Yc (t/ha) for a 45% trafficked area as shown below: Yc = ((0.45 x 11.70) +((1-0.45) x 13.22)) = t ha -1 56

57 Table 5.1. Estimated yield and yield increase (t ha -1 ) as affected by reductions in the trafficked area (%) for 2 and 3 cut systems. Trafficked Area, % 2 Cut 12.0 t ha -1 3 Cut 16.5 t ha -1 Yield, Yield increase, Yield, Yield increase, t ha -1 t ha -1 t ha -1 t ha Figure 5.2 shows these in economic terms assuming that the crop has a value of 72 t of dry matter (Hargreaves - SRUC, 2016) giving improvements of 38 ha to 71 ha -1 and 53 ha -1 to 98 ha for the 2 and 3 cut systems respectively. This equates to 1.10 ha -1 and 1.50 ha -1 for every 1% reduction in trafficked area for the 2 and 3 cut systems respectively. These economic benefits are in agreement with Alvemar (2014) in Sweden at 40 ha -1 to 49 ha -1 and Stewart et al. (1998) in Scotland at 79 ha -1 ( 113 ha -1 when adjusted for retail price inflation). Figure 5.2. The effect of reducing the trafficked area (%) on the potential economic benefit. Red - 2 Cuts (12 t ha -1 ); Blue - 3 Cuts (16.5 t ha -1 ) 5.3 Costs Typical costs of alternative machine guidance systems are given in Table 5.2. They include the capital expenditure for the receivers and subscription fees, at four levels of investment. These ranged from elementary systems using a light bar and manual steering with a low level of accuracy, to assisted steering with both low and higher accuracy and fully integrated 57

58 steering systems with high accuracy levels. To provide both improved accuracy and repeatable positioning real time kinematic (RTK) navigation systems are required, therefore adding the cost of a network subscription fee. This was chosen as an alternative to the capital costs of a base station. It should be noted that the lower accuracy systems are not ideal and will result in significantly higher trafficked areas. Although not considered in this analysis the benefits from the reductions in overlap between adjacent machine passes would also be reduced. Table 5.2. Current typical commercial equipment guidance costs/system (less VAT). Investment Equipment Capital Expenditure Level 1 Level 2 Level 3 Level 4 Low accuracy (+/ mm) Manual steering Low accuracy (+/ mm) Assisted steering High accuracy (+/-20mm) Assisted Steering High accuracy (+/-20mm) Fully integrated steering RTK Annual Subscription Fee 1,500 - No 5,000 - No 10, Yes 15, Yes Repeatable Positioning The individual cost components and the total annual costs are given in Table 5.3; these are based upon the following assumptions: 1. Interest rates 4.5% (base rate + 4%) (Nix, 2015) 2. Depreciation 15% per annum for 5 years with a residual value of 25% (Nix (2000 and 2015)) 3. Maintenance costs 5% 4. Training - 100/year The annual cost/ha for the four system levels are given in Figure 5.3 for a range of CTF harvested areas up to 1500 ha. The figure of 1500 ha is the assumed upper limit for a given harvester and associated guidance equipment during a single cut. This is based upon harvesting rates of 75 ha day -1 (Farmers Weekly, 2016 and Cottey, 2016) and an assumed minimum 20 workdays/cutting period for a contractor-based system. This is based upon Nix (2000) who quotes workable time coefficients of 70% to 75% for the May to September period. The data also enable grassland farmers to estimate the system cost for smaller harvest areas. 58

59 Table 5.3. Cost components and total annual cost for four levels of capital expenditure ( ) Level 1 Level 2 Level 3 Level 4 Initial capital outlay 1,500 5,000 10,000 15,000 Cost of capital Depreciation Maintenance Training Annual subscription fees Total annual cost The data in Figure 5.3 show the dramatic effect that the size of the area harvested has on the cost/ha, with the costs/ha becoming asymptotic to the horizontal axis. This shows that the Level 1 system costs less than 4.68 ha -1 for areas in excess of 100 ha with the more expensive Level 4 system, costing ha -1 for areas in excess of 200 ha reducing to 2.85 ha -1 for areas in excess of 1500 ha. Annual cost, ha Area per system, ha Figure 5.3. Annual cost/ha/system of the guidance system for four levels of capital expenditure. Blue Level 1, 1500; Red Level 2, 5000; Green Level 3, 10,000; Black Level 4, 15,000. With CTF systems in grassland it is not only the harvester that needs to be equipped with machine guidance systems but also the tractors drawing trailers and conducting other field operations. The data in Figure 5.3 enable the reader to determine the total cost for the number of systems required for their operation and to maybe have different levels of technology for the harvester and tractor based operations for a given harvested area. For simplicity however the data in Figure 5.4 assume they will have the same level of technology as the harvester and that three tractors are needed to conduct the transportation of the crop or conduct other field operations. Hence the costs/ha in Figure 5.4 are four times that of Figure 5.3. This shows that the Level 1 system costs less than ha -1 for areas in 59

60 excess of 100 ha with the more expensive Level 4 system, costing ha -1 for areas in excess of 200 ha reducing to ha -1 for areas in excess of 1500 ha Annual cost, ha Area per system, ha Figure 5.4. Comparison of: a. the annual cost/ha of the guidance systems for a harvester and three tractors for four levels of capital expenditure: Blue Level 1; Red Level 2; Green Level 3; Black Level 4, with: b. the benefits for: a. The 2-harvest/45% tracked area ( 38 ha -1 ) (dot long dash) and b. The 3-harvest/15% tracked area ( 98 ha -1 )(dot short dash) systems. 5.4 Cost benefit analysis By comparing the costs and potential benefits shown in Figure 5.4 it can be seen that: 1. The break-even area for the most sophisticated (Level 4) guidance system is 450 ha for the 2 cut system with a 45% trafficked area; this reduces to a break-even area of 50 ha for the least sophisticated (Level 1) guidance system. 2. The break-even area for the most sophisticated (Level 4) guidance system is 175 ha for the 3 cut system with a 15% trafficked area; this reduces to a break-even area of 54 ha for the Level 2 guidance system. 3. It is interesting to note that if it was possible to achieve a 35% trafficked area with four Level 1 guidance systems it would be cost effective at areas above 28 ha. (4 x / 68 ha -1 = 27.5 ha). (Total annual cost ( )/Annual economic benefit ( ha -1 )). 4. It should be noted that none of the associated benefits of guidance and auto-steer have been considered in this study. These include more grass cut per pass 60

61 (particularly with Levels 3 & 4 guidance and with wider machines), straighter rows and less stress on the drivers. 5.5 Case studies Using the data above it is possible to undertake an analysis of individual operations, for example: 1. A farm with 200 ha of grass with 3 cuts where CTF systems can reduce the trafficked area from 80% to 40% and is contemplating the purchase of a Level 3 navigation system for the harvester and two Level 2 systems for the supporting tractors. a. Figure 5.2 shows a benefit of 60 ha -1 for the 3 cut system with a 40% trafficked area. b. Figure 5.3 shows for an area of 200 ha the cost of a Level 3 navigation system is ha -1 and 6.63 ha -1 for each of the Level 2 systems, giving a total of ha -1. c. The break-even area is (from Table 5.3) the total cost = = 5700 divided by the benefit of 60 ha -1, giving 95 ha. d. The net benefit of conversion to CTF with a trafficked area of 40% gives rise to 60 ha -1 less ha -1 = ha -1. e. This equates to 6300 for the enterprise. 2. If the area of grass in the above example, was limited to 100 ha, only marginally above the break-even area, the cost/ha of the navigation systems would double giving a total of and a net benefit of 3.00 ha -1 and a total of 300 for the enterprise. In this case reducing the technological investment to three Level 2 systems reduces the cost to ha -1, resulting in a net benefit of ha -1 worth 2025 for the enterprise. 3. In the case of contractor harvested systems a financial agreement would have to be made between the farmer and the contractor. Where some of the benefit of any additional yield/revenue to the farm has to offset the cost and operational complexity experienced by the contractor services. Therefore the negotiation ranges between yield benefit of 60 ha -1 to the farm for a 40% tracked area, 3 cut system and a base contractor cost of ha -1 where the contractor is harvesting in excess of 1500 ha of controlled traffic grassland. In Denmark where there is a history of CTF grassland management, (Appendix I), contractors use their capability to offer CTF systems to gain competitive advantage in 61

62 the market (Chamen, 2016). The actual cost per ha will be less than that quoted for contracting services, when the guidance systems can be amortized over larger land areas with other crops/field operations where the navigation systems provide further benefits. 5.6 Conclusions The following conclusions are made assuming that: the average yields for 2 and 3 cut managed grassland harvest systems in the UK are 12 t ha -1 and 16.6 t ha -1 of dry matter respectively, with a value of 72 t -1 and conventional or normal traffic management covers 80% of the field area. The effect of the removal of the traffic would increase forage yields by 13%. It was also assumed that the additional cost of the adoption of CTF is due only to the cost of the navigation systems and that any improvements in machine matching would be part of the normal longer-term replacement policy, as suggested in Chapter Reducing the trafficked area from 80% to 45% increases the yield by 0.53 t ha -1 and 0.73 t ha -1 for the 2 and 3 cut systems respectively. 2. Reducing the trafficked area from 80% to 15% increases the yield by 1.00 t ha -1 and 1.36 t ha -1 for the 2 and 3 cut systems respectively. 3. These yield increases are currently valued at between 38 ha -1 and 98 ha -1. Which are in agreement with the suggested benefits from earlier studies in Scotland and Sweden. 4. A 1% reduction in the trafficked area increases the benefit of CTF at a rate of between 1.10 ha -1 and 1.50 ha -1 for the 2 and 3 cut systems respectively. 5. Based only upon the additional costs of the guidance systems to implement CTF: the low accuracy ( mm) light-bar, manual steered system can be purchased for approximately 1,500 system. The cost for a fully integrated, high accuracy (20mm) real time kinematic navigation system is approximately 15,000 system. There may need to be as many as 4 systems to equip the harvester and the accompanying tractors. 6. Four low accuracy systems cost less than ha -1 for areas in excess of 100 ha; four fully integrated, high accuracy systems cost less than ha -1 for areas in excess of 200 ha reducing to ha -1 for areas greater than 1500 ha/harvest. 7. The break-even area for four low accuracy, light-bar, manual steered systems is 50 ha for a 45% trafficked area CTF system with 2 cuts /year. This increases to 450 ha for four fully integrated, high accuracy real time kinematic navigation systems. 8. The break-even area for four low accuracy, light-bar, manual steered systems is 28 ha for a 35% trafficked area CTF system with 3 cuts a year. This increases to 78 ha 62

63 for four low accuracy, assisted steering navigation systems and 250 ha for four fully integrated, high accuracy real time kinematic navigation systems. With a trafficked area of 15% the break-even area for four fully integrated, high accuracy real time kinematic navigation systems is 175 ha. Overall the above analysis shows that providing the navigation systems are wisely selected for the size of the operation then controlled traffic farming in grassland can be cost effective. The data given are aimed at allowing the benefits, costs and break-even areas of individual systems to be estimated and investment choices to be made. The data also provide a basis for the negotiation between a farmer and contractor when considering the benefits and costs of grassland CTF systems. 63

64 6. Conclusions 1. Previous studies in the UK have shown that wheel traffic damage can reduce grass yields in the range of 5% to 20%. Detailed analysis of the most robust data indicates that the absence of wheel traffic could improve the mean yield by c.13%. This has been seen by the results in this study, which indicate a benefit of 8% after only the second and third silage cut being CTF. 2. The wheel damage can potentially impact on the majority of the field area, as a previous pilot study by Harper Adams University showed, that with normal (N) random traffic management practices 65% of the area of the field had been trafficked in just one cut. 3. A range of commercially available equipment has been identified that would form the basis of potentially practical CTF systems for forage production on farms of different sizes. These systems were considered alongside the requirements for arable production in cropping rotations and were based upon mower widths of 3, 4, 5, 9 and 12 m and gave rise to reducing trafficked areas from 40% to 13% for grassland only and 48% to 18% when considering arable operations in the rotation. 4. As with arable CTF systems, farmers need to embrace the idea and then plan future machinery acquisition accordingly. In very few instances is it feasible to make a complete change to a low trafficked area CTF system overnight, it is a process of planning and gradual transition until all compatible equipment becomes available as part of the normal machinery replacement policy. Therefore the expectation is that early grassland CTF systems might have relatively high trafficked areas that would reduce when machines were replaced, hence for the purposes of the economic analysis the only extra costs would be those associated with the vehicle guidance systems. 5. The experimental work showed the soil was affected by the number of passes of vehicles. The greatest single increase, as a percentage of the overall soil bulk density increase, was from zero to two passes on both N (54%) and the CTF (45%). The CTF field showed the greatest overall increase in soil bulk density by the time the number of vehicle passes had reached 6+ with an 18% increased compared to 15% for the N field. These increases in soil bulk density were reflected in similar increases in the penetrometer (soil resistance) and visual evaluation of the soil structure (VESS) scores. 6. Total DM yield increase for the CTF system over the second and third silage cuts, when it would have had a beneficial effect, was 13.5% (0.80 t ha -1 ). 7. The increased traffic from the increased number of vehicles passes from the experimental work also effected the dry matter yields from these areas with a reduction in yield for the 6+ passes compared to the 0 passes for both the N (19.5%; P<0.01) and for the CTF (29.9%; P<0.01) with the second silage cut and a reduction of 36.7% for the N (P<0.01) and 38.2% with the CTF (P<0.001) for the third silage cut. 64

65 8. The experimental work showed that the use of existing equipment together with an autoguidance system reduced the trafficked area covered by the CTF system (30.4%) by 57% compared to the N system (87.4%). The area covered by the CTF would have been reduced still further to 19.4%, if the second slurry application had been able to be spread at the same working width as the other machinery operations. 9. Based upon a 13% increase in yield from the absence of wheel damage and the average dry matter yields for 2 and 3 cut harvest systems in UK grassland with a value of 72 t, the following conclusions can be made: a. Reducing the trafficked area from an assumed 80% for random traffic to 45% for CTF increases the yield by 0.53 t ha -1 and 0.73 t ha -1 for 2 and 3 cut systems, respectively. Similarly reducing the trafficked area to 15% increases the yield by 1.00 t ha -1 and 1.36 t ha -1 for 2 and 3 cut systems, respectively. b. These yield increases are currently valued at between 38 ha -1 and 98 ha -1 and are in agreement with the suggested benefits from earlier studies in Scotland and Sweden. 10. Based upon the assumption that: it is only the cost of the guidance system that is needed to implement a CTF system and that four guidance systems are required to equip the harvester and the accompanying tractors. Then the following conclusions can be made: a. Four low accuracy, manual steered systems cost less than ha -1 for areas in excess of 100ha; four fully integrated, high accuracy systems cost less than ha -1 for areas in excess of 200 ha reducing to ha -1 for areas greater than 1500 ha for each harvest. The low accuracy systems will inevitably increase the theoretically anticipated trafficked area due not only to their lower accuracy but also their inability to reposition vehicles precisely on the same traffic lanes. b. The break-even area for four low accuracy, manual steered systems is 28 ha for a 35% trafficked area CTF system with 3 cuts a year. This increases to 250 ha for four fully integrated; high accuracy real time kinematic navigation systems. c. With a CTF trafficked area of 15% the break-even area is 175 ha for four fully integrated, high accuracy real time kinematic navigation systems. The above analysis shows that providing the navigation systems are wisely selected for the size of the operation then controlled traffic farming in grassland can be cost effective. The data given are aimed at allowing the benefits, costs and break-even areas of individual systems to be estimated and investment choices to be made. 11. The data from the economic analysis also provide a basis for the negotiation between the benefit to a farmer and the costs to a contractor when considering the implementation of grassland CTF systems. 65

66 12. There is sufficient evidence to support the hypotheses that reduced coverage of a silage field with a planned system of controlled traffic will reduce the area extent of compaction and therefore increase yield and reduce variation compared to a field with no restriction on traffic. 66

67 7. References AHDB Dairy Report, (2015). Soil Compaction and Soil Loosening for Maintaining Grassland Productivity, AHDB Dairy, Stoneleigh, UK. Alakukku, L. (1999). Subsoil compaction due to wheel traffic. Agricultural and Food Science in Finland, 8, Alvemar, H. (2014). Controlled traffic for grass silage production An economic evaluation for dairy farmers. Degree thesis No 889. Swedish University of Agricultural Sciences. Uppsala. Are, M., Reintam, E., Selge, A., Sanchez de Cima, D. (2015). The after-effect of soil compaction on soil properties and grassland productivity. Mulla tallamise järelmõju mulla omadustele ja rohumaa saagikusele. Ed. Alaru, M.; Astover, A.; Karp, K.; Viiralt, R.; Must, A. Agronoomia 2015 (10 15). Tartu: Ecoprint. (In Estonian, with English summary). Ball, B.C., Batey, T. and Munkholm, L.J. (2007). Field assessment of soil structural quality a development of the Peerlkamp test. Soil Use and Management, 23, Baver, L.D., Gardener, W.H. and Gardener, W.R. (1972). Soil Physics, 4 th Ed. John Wiley and Sons, New York. Bell, B. (2008). Farm Machinery. 5 th ed. Ipswich: Old Pond, 326p. Chamen, W.C.T. (2011). The effects of low and controlled traffic systems on soil physical properties, yields and the profitability of cereal crops on a range of soil types. Unpublished PhD thesis, Cranfield University. Chamen, W.C.T. (2016). Personal communication. CTF Europe, Maulden. Chyba, J., Kroulík, M., Krištof, K., Misiewicz, P.A., Chaney, K. (2014). Influence of soil compaction by farm machinery and livestock on water infiltration rate on grassland. Agronomy Research, 12, Cottey, D. (2016). Personal communication, Claas UK, Saxham. Davies, B., Eagle, D., Finney, B. (1993). Soil Management. 5 th Ed. Farming Press, Ipswich 280p. Douglas, J.T. and Crawford, C.E. (1989a). Effects of wheel-induced compaction on grass yield and nitrogen uptake, Departmental Note 19, Scottish Centre of Agricultural Engineering, Penicuik. 67

68 Douglas, J.T. and Crawford, C.E. (1989b). The response of ryegrass to soil compaction and applied nitrogen, Departmental Note 21, Scottish Centre of Agricultural Engineering, Penicuik. Douglas, J.T. and Crawford, C.E. (1989c), Soil compaction and novel traffic systems in ryegrass grown for silage: effects on herbage yield, quality and nitrogen uptake. Proceedings of The XVI International Grassland Congress, 4-11 October 1989, Nice, France, Douglas, J.T. and Crawford, C.E. (1991). Wheel-induced soil compaction effects on ryegrass production and nitrogen uptake. Grass and Forage Science, 46, Douglas, J.T. and Crawford, C.E. (1993). The response of a ryegrass sward to wheel traffic and applied nitrogen. Grass and Forage Science, 48, Douglas, J.T., Campbell, D.J., Crawford, C.E. (1992). Soil and crop responses to conventional, reduced ground pressure and zero traffic systems for grass silage production. Soil & Tillage Research, 24, pp Douglas, J.T., Crawford, C.E., and Campbell, D.J. (1995). Traffic Systems and Soil Aerator Effects on Grassland for Silage Production. Journal of Agricultural Engineering Research, 60, Douglas, J.T., Koppi, A.J., Crawford, C.E. (1998). Structural improvements in a grassland soil after changes to wheel-traffic systems to avoid soil compaction. Soil Use and Management, 14, Dugdale J. (2015). Personal communication. Elonen, P. (1986). Soil compaction a problem in Finnish arable farming. Ed. I. Håkansson, J. von Polgár & K. Rask. Reports from the Division of Soil Management, No. 71. Jordpackning ett problem i Finsk åkerodling. Rapporter från jordbearbetningsavdelningen, No. 71. Swedish University of Agricultural Sciences, Uppsala. (In Swedish) Farmers Weekly. 25 th March 2016 and 1 st April Reports on Forage Harvesting Equipment. Frame, J. (1983). The effect of wheel tracking on red clover (Trifolium pratense) swards. BGS Occasional Symposium No 14. British Grassland Society, Frame, J. (1985). The effect of tractor wheeling on red clover swards. Research and Development in Agriculture, 2, Frame, J. and Merrilees, D.W. (1996). The effect of tractor wheel passes on herbage production from diploid and tetraploid ryegrass swards. Grass and Forage Science, 51,

69 Frost, J.P. (1985). Some Effects of Machinery Traffic on Grass Yield. Machinery for Silage. Ed. J.K. Nelson & E.R. Dinnis. BGS Occasional Symposium No 17. British Grassland Society, Frost, J.P. (1988a). Effects on Crop Yields of Machinery Traffic and Soil Loosening. Part 1. Effects on Grass Yield of Traffic Frequency and Date of Loosening. Journal of Agricultural Engineering Research, 39, Frost, J.P. (1988b). Effects on Crop Yields of Machinery Traffic and Soil Loosening. Part 2. Effects on Grass Yield of Soil Compaction, Low Ground Pressure Tyres and Date of Loosening. Journal of Agricultural Engineering Research, 40, Galambošová, J. and Godwin, R.J. (2012). Elements of Precision Farming. Slovak University of Agriculture in Nitra, 135 p. Godwin, R.J., Richards, T. E., Wood, G. A., Welsh, J. P., Knight, S. (2003). An economic analysis of the potential for precision farming in UK cereal production. Biosystems Engineering, 84, Godwin, R.J. (2015). Precision Farming. Ingenia, 64, Royal Academy of Engineering, London Godwin, R.J. and Spoor, G. (2015). Choosing and evaluating soil improvements by subsoiling and compaction control. In Ball, B. C. and Munkholm, L. J. (eds). Visual Soil Evaluation: Realising Potential Crop Production with Minimum Environmental Impact. CABI, Wallingford, UK. Håkansson, I., McAfee, M., Gunnarsson, S., (1990). Effects of traffic during harvest on yield of grass leys. Results from field trials on 24 Swedish sites. Verkan av körning med traktor och vagn vid vallskörd. Resultat från 24 försöksplatser. (No. 78). Department of Soil Sciences, Uppsala. (In Swedish, with English summary). Hansen, S. (1996). Effects of manure treatment and soil compaction on plant production of a dairy farm system converting to organic farming practice. Agric. Ecosyst. Enviroment, 56, Hargreaves, P.R., Bonnett, S., Baker, K., Ball, B. (2012). DairyCo Soil Compaction Progress Report. SRUC Research & Harper Adams University College. Hargreaves, P.R., Ball, B., Baker, K. (2014). DairyCo Soil Compaction Progress Report: Grass Yield and Soil. SRUC Research & Harper Adams University. Hargreaves, P.R. (2016). Personal communication, SRUC, Dumfries 69

70 HGCA (2002). Precision farming of Cereals: Practical Guidelines and Crop Nutrition. Home Grown Cereals Authority, London (Now: AHDB Cereals and Oilseeds, Stoneleigh). James, D. (2016). How controlled traffic farming can help improve grass yields. Farmers Weekly, 18 March 2016, 38. Jorajuria, D., Draghi, L., Aragon, A. (1997). The effect of vehicle weight on the distribution of compaction with depth and the yield of Lolium/Trifolium grassland. Soil & Tillage Research, 41, Jørgensen, R.N., Green, O., Kristensen, K., Gishum, R., Sørensen, C.G. (2009). Estimating impact on clover-grass yield caused by traffic intensities. Ed. C. Lokhorst and P. W. G. Groot Koerkamp. Precision livestock farming 09. Wageningen Academic Publishers, Kjeldal, M. (2013). Contractors improve yield and feed quality by use of CTF in forage grass. First International CTF Conference, Toowoomba, Australia. Kroulík, M., Misiewicz, P., Chyba, J., White, D. (2014). Field traffic intensity for forage harvesting in the UK. Harper Adams University Project Report 099. McIntosh, J.L. (1969). Bray and Morgan soil test extractants modified for testing acid soils from different parent materials. Agronomy Journal, 61, Nix, J. (2000). Farm Management Pocketbook. 31 st (2001) Edition. Imperial College at Wye. Nix, J. (2015). Farm Management Pocketbook. 46 th Consultants Ltd. (2016) Edition. Agro Business Negi, S.C., McKyes, E., Raghavan, G.S.V. and Taylor, F. (1981). Relationships of field traffic and tillage to corn yields and soil properties. Journal of Terramechanics, 18, PDA (2011). Soil Analysis Key to Nutrient Management Planning PDA Leaflet 24. Potash Development Association, York, UK. Pedersen, H.H. (2012). Mosegaarden, 12 m CTF forage grass. [Online] CTF Europe. Available from: [Accessed 24/04/2015]. Raghavan, G.S.V., McKyes, E., Stemshorn, E., Gray, A., Beaulieu, B. (1977). Vehicle compaction patterns in clay soil. Transactions ASAE, 20, Rasmussen, K.J. and Møller, E. (1981). Regrowth after pre-wilting of grassland crops. II. Soil compaction in connection with harvest and transport. Genvækst efter fortørring af græsmarksafgrøder. II. Jordpakning i forbindelse med høst og transport. Tidskr. Planteavl. 85, As cited in: Alvemar, H. (2014). Controlled traffic for grass silage production - 70

71 An economic evaluation for dairy farmers. Degree thesis No 889. Swedish University of Agricultural Sciences, Uppsala. Reintam, E., Krebstein, K., Sanchez de Cima, D., Leeduks, J. (2013). Impact of soil compaction on productivity of rye grass and lucerne. Mulla tallamise mõju karjamaa raiheina ja hübriidlutserni saagikusele. Ed. T. Kangor; S. Tamm; R. LindepuuAgronoomia 2013 (22 27). Jõgeva: AS Rebellis. (In Estonian, with English summary). Robertson, G.P., Sollins, P., Ellis, B.G., Lajtha, K. (1999). Exchangeable ions, ph and cation exchange capacity. In (Eds): G.P. Robertson, D.C. Coleman, C.S. Bledsoe and P. Sollins. Standard Soil Methods for Long-term Ecological Research. Oxford University Press. Smith, E.K., Misiewicz, P.A., Girardello, V., Arslan, S., Chaney, K., White, D.R., Godwin, R.J. (2014). Effects of traffic and tillage on crop yield (Winter Wheat Triticum aestivum) and the physical properties of a sandy loam soil. ASABE Paper No , St. Joseph, Michigan, USA. Stewart, L., Copland, T., Dickson, J., Douglas, J. (1998). Economic evaluation of traffic systems for arable and grass crops on an imperfectly drained soil in Scotland. Journal of Sustainable Agriculture, 12, University of Nebraska (1999). Management Strategies to Minimise and Reduce Soil Compaction. G A. Volden, B., Sveistrup, T.E., Jørgensen, M., Haraldsen, T.K. (2002). Effects of traffic and fertilization levels on grass yields in northern Norway. Agricultural and Food Science in Finland, 11, Warner, N.L. and Godwin R.J. (1988). An experimental investigation into factors influencing the soil injection of sewage sludge. Journal of Agricultural Engineering Research, 39,

72 Appendix I CTF in Forage Grass An AgroTech and CTF Europe Workshop Report 9 10 June 2008 Summary Motivation for CTF comes from both farmers experiences and from research. Traffic damages grass through compaction of the soil and by direct tyre damage to the sward. Further losses in yield come from uneven cutting height due to rutting and costs are increased by reduced economic life of grass leys. Several farmers in Denmark have experienced higher yields and more even crop growth as a result of using CTF in forage grass. Rutting and tearing caused by traction were considered the major causes of damage to grass rather than compaction of the top and subsoil, which were of lesser concern or effect on yield. Research at Bygholm is continuing to show the damaging effects of wheel tracks that increase with tracked area, pressure and number of passes. Results have also shown that a larger tracked area at low pressure may do more damage per unit area than narrower tracks at higher pressure, providing rutting is avoided. RTK GPS guidance is essential for efficient operation of CTF, resulting for example in 12 m wide systems tracking just 13% of the field area. Achieving CTF in practice is through a gradual alignment of all machines in the system and matching of wheel tracks that generally keep vehicles within 3 m overall width on the road. Farmers practising CTF in grass do so only during years when the fields are cropped with grass, but would like to extend it into forage maize, where severe soil damage occurs during harvesting. Farmers who are in the process of conversion to or have converted to CTF, are increasing in their enthusiasm for the system. A 12 m working width and a 2 m track width were suggested as the standards that manufacturers should seriously consider for grass machines. Nine metres is an alternative that could match the width recommended for combinable crops. 72

73 Introduction A number of Danish dairy farmers have experienced significant yield increases and more even crops by using CTF. Driving across grass fields causes damage not only by compaction of top and subsoil but also by physical damage to the crop. Tracks created by random traffic also cause uneven cutting height which again reduces yield. Most forage grass in Denmark is grown in rotation with other crops. Fields are typically in grass for three years before ploughing for other crops. The aims of the workshop were: To present experiences with practical and research implementation of CTF in forage grass. To discuss optimal implementation of CTF in grass. The programme for the workshop was a mix of presentations and field visits that included scientific research, practical application, field observations and both static and moving demonstration of the machinery involved. The workshop started at the Bygholm Research Centre of the University of Århus, located in Horsens. Presentations Hans Henrik Pedersen, while introducing the presentations and the workshop as a whole, said that some of the sessions would be in English and others in Danish, for which there would be ongoing translations. (Hans Henrik s presentation) Hans Henrik Pedersen, Danish Agricultural Advisory Service CTF in Forage Grass Hans Henrik, who said that he had been in the advisory service longer than he cared to remember, had first come across GPS in the early 1990s, when it was a hot topic, but little used other than for yield mapping on grain harvesters. More recently however, auto-steering had become the main driving force for GPS and in 2005 he had gone to Australia to find out more about it and what they were using it for. This was when he first heard about CTF and how happy farmers were in Australia, being able to use it for sowing between rows as well as for CTF. This introduction to CTF led to the organisation of a seminar on CTF in Denmark in November 2005, which although not well attended by researchers, had good support from farmers, including a significant number from the UK. CTF, he said, is all about soils, getting them healthy enough that roots get right down into them and are not constrained by compaction. Soils are where the money is and to find out about them we should always have a spade! As a result of that workshop, CTF Europe was born and became a reality at Agritechnica in At that time we were four colleagues, one each from Denmark, UK, Germany and the Netherlands, but have recently been joined by a new partner in Slovakia. CTF Europe is a commercial network with well in excess of 100 members, all of whom have access to all of the partners. Four newsletters are published each year in what will now be 73

74 five languages and members have access to seminars across Europe. The CTF website has a lot of good background information about CTF as well as reports on all workshops. AgroTech was formed by the Danish Agricultural Advisory Service in 2007 and is a Danish government-authorized service institute working in the field of knowledge transfer between biology-based natural science and technology. AgroTech cooperates with private companies, public research institutions and the Danish Agricultural Advisory Service. Hans Henrik, whose principal expertise lies with satellite-based guidance systems, then went on to provide an overview of the different systems available (Table 1). Table 1. Categories of GPS guidance systems. Type Track to track accuracy (cm) Absolute accuracy (cm) Approx. receiver price ( ) Licence fee, /annum GPS DGPS (Egnos, Beacon) ,000-2 frequency DGPS (SF2, Ominstar HP) ,000 1,000-1,500 RTK ,000 - For CTF, the absolute or repeatable positioning accuracy of RTK is essential. Other systems can be used, but these require a mark on every pass across the field to achieve repeatable positioning. The motivation for CTF in forage production came from both farmers experiences and from research. Research showed that yield was reduced significantly by driving on grass and that the penalties increased with increasing number of passes and the pressure applied. Percentage reduction in yield ranged from 10 to 60% (Rasmussen, 1981). Most farmers growing grass also grow maize and in this respect both satellite guidance and CTF could be a means of improving crop management. Slurry is often injected before sowing maize, and a key aim with this is to inject the slurry exactly below where the seed will be sown. With RTK this is perfectly feasible and AgroTech have this year used this to measure the accuracy of seed and slurry placement. Slurry is often applied with large tankers driving in dog walk mode and in the case of the machines measured, this led to around 50% of the area being wheeled (4.5 m out of 9 m). In the wheeled area, injection depth was cm, while in the loose non-wheeled area, the depth was around 17 cm, which meant that slurry was placed too far from the maize seed. The effect of compaction was also very evident during injection, when there was very little soil movement where the wheels had run. Surprisingly, delegates reported that they had rarely seen wheeling effects in maize and yet this is a crop that is considered highly susceptible to compaction. In terms of CTF for grass alone, a 12 m system would track around 13% of the area with 800 mm wide tyres, but this would be increased to 26% if intermediate tracks were necessary. Finally, Hans Henrik outlined the challenges/questions that we were facing with CTF in the context of forage production. These were: 74

75 Should we have standard working widths and if so, what should they be? o 8, 9, 10, 12 or 15 m? What wheel track (gauge) should we use? o 2.0, 2.25, 2.75 or 3.0 m? Chopper and forage wagons should these be towed or driven behind in the same track or on the adjacent track? Will fields remain level without traffic? Can work capacity be maintained with CTF? Ole Green, University of Århus The Department of Agricultural Engineering at the University or Århus specializes in working on technical solutions in agriculture and this is particularly true for CTF, on which they have been working since In spring 2007 they started a grass trial to assess how different types of traffic effect yield. This includes timing of traffic, pressures on the soil and number of passes. Over the past 18 months, they have managed to simulate 5 years of traffic. They have 840 plots that include 35 replications of the treatments, which are applied in a unique way by driving temporarily off the permanent traffic lanes. Even though this appears to be random application of the treatments, RTK guidance and curve tracking means that repeated passes can be made in exactly the same place. The figure above shows very clearly that clover really does not like to be driven on! 75

76 The treatments include: 6 and 12 t axle load 1 and 2.5 bar tyre pressure time of year Results (Figure 1) were contrary to expectations, but were immediately explicable by the fact that the lower pressure system had increased the area of grass impacted by a factor of around 2.5 due to a larger and wider tyre footprint. Ole said that the results were allowing them to gradually improve the output of a model to predict the effect of different traffic systems on grass yields. Figure 1. Yield losses of grass relative to those from completely non-trafficked plots. The vertical scale is the percentage yield loss and the horizontal is the date of driving. University of Århus field visit We visited the field where Ole was carrying out his research and saw for ourselves the methodology used and the noticeable effect of the wheelings, despite this having been a very dry season in Denmark (Figure 2). 76

77 Figure 2. Part of the experiment at the University of Århus investigating the damage caused by driving on grass used for forage production. Note the absence of clover in the wheel tracks. Jørgen Sønderby, Organic Dairy Farmer, Bjerringbro and Bendt Jensen, Organic Farming Advisor, Randers The Bjerringbro farm extends to 350 ha and has 180 dairy cows, but Jørgen also grows organic vegetables (leeks, cabbage and carrots) as well as grain crops. His medium term grass leys are sown either in autumn or spring using ploughing followed by harrowing to level the land and a 7 m wide 12 t roll for a final firming. About ten years ago, if treated well, the grass would remain productive for 6-8 years, but recently production has fallen from 7-9 t/ha in the first year or two, to just 5-6 t ha -1 by year three, meaning much higher production costs. This change has coincided with using a contractor who has large self-propelled machines and harvester wagons carrying perhaps 20 t. The one thing that hadn t changed though was the size of the tractors hauling the wagons, which remained at about 150 hp. At first he couldn t figure out what they were doing wrong, but then he noticed that strips with poor grass coincided with the path of the haulage wagons. Initially he thought this was the result of compaction, but then he, together with his agronomist Bendt Jensen, concluded that it was not only compaction that was causing the problem, but more importantly, it was grass being torn out by the traction of the towing tractors. If larger tractors had been used, traction damage might not have been so great. At around this time auto steer systems were starting to come onto the market and he was considering these for his vegetable cropping, but thought, why not try it for grass. So, they looked around for matching widths of machine, considering 8 m, 12 m and 16 m, but there was nothing that would provide complete matching. In the end, they opted for 12 m that matches their slurry spreader at 24 m and hired a 12 m rake. The only machine that does not match is their mower, so grass is not cut on the fixed tramlines. Compaction and traction from the mower is not a major issue, but they would like to avoid grass from the mower 77

78 falling into the tracks that are driven over while raking (Figure 3). This results in some grass being left on the ground as well as a loss in quality of that which is picked up. The auto steering is a John Deere system using SF2 correction which means that the positions drift and have to be re-computed. This is difficult during raking because the tracks are not visible beneath the grass. They are now in their second year of an experiment looking at the response of four grass mixtures (e.g. Danish mixture 42 with red and white clover plus ryegrass) to traffic and no traffic. They already have three years of test results for alfalfa which returned 8000 kg ha -1 of dry matter in the first year and 10,000 kg ha -1 in the second and third years. In this year of drought, the alfalfa has continued to grow, whereas grass stopped around 25 May. They have observed only minor differences in yield between tracked and non-tracked plots, but they are using relatively light equipment (20 t slurry wagon and 9 t tractor compared with the contractor s machinery that might weigh 51 t), but on sloping land they can show the difference in traction damage between uphill and downhill operation. Jørgen said that they have seen a cm difference between upslope and downslope crop 12 months after the damage occurred. They continue to work towards a fully integrated system and are now looking to achieve a 12 m CTF system for maize. Jørgen s presentation and a supporting article are available through the links provided. Figure 3. Large capacity and good chopping are essential for efficient operation (Jørgen Sønderby). Thorkil Schrøder, JF-Stoll JF-Stoll is a Danish company specializing in grassland machinery such as mowers, mower conditioners, tedders and rakes. In Thorkil s department they test functionality, conditioner 78

79 performance and capacity and machine stability and manoeuvrability. A constraint on their designs is transport width, which they must keep to 3 m or less, both as self propelled or transported machines. (A range of machines is shown here) Cutting widths of mowers, which come in a triple format, range from m (Figure 4) and have transport wheels centred at 2.75 m, but these could be adjustable. One uncertainty they have in terms of CTF is the cut overlap needed to ensure complete cover. Presently, JF- Stoll is working closely with Jørgen Sønderby in his trials, and Thorkil showed some graphic evidence of the traction effects on the productivity of grassland. Figure 4. The 14.5 m mower made by JF-Stoll and used by Erling Kjær (Day 2 visit). Tim Chamen, CTF Europe. An introduction to CTF Tim started by illustrating the fact that machines were now seven times heavier than when they were in the 1940s and that they compacted around 100% of the area every season. This damage can be clearly seen in forage production where yellow lines appear in grass fields around five days after harvesting. CTF is simply a case of confining compaction to the least possible area of permanent traffic lanes. This avoids damage on most of the soil damage that is easily visible just below cultivated surfaces. Without compaction, yields increase while tillage and machinery costs go down and environmental damage is also significantly reduced. A number of controlled traffic systems are emerging for European agriculture, including OutTrac, TwinTrac and AdTrac, all of which offer different opportunities for using existing equipment in clever ways. At the centre of all CTF systems however is a good satellite guidance system that must provide repeatable positioning and a high level of accuracy. In many cases satellite guidance can be justified purely on overlap savings, but when it is also used for CTF, the economics are far more favourable. Tim then went on to describe the Unilever Colworth CTF project, which is aimed at finding out if there are any problems with CTF at field level what works well, what is a problem and are the benefits and economics achieved in practice? Now coming into its fifth year, the Colworth project has not met with any major problems and has provided information on 79

80 improvements to soil conditions and how to manage the wheel tracks. The complete presentation is provided at this link. Day 2 Farm visits Bjarne Andersen, Dairy Farmer, Outrup Bjarne started with CTF in 2006 after he noticed the amount of damage that wheel tracks were causing in his 110 ha of forage grass, which is used to feed a 330 dairy cow herd. However, he soon discovered when he asked for a 12 m wide system, that manufacturers were not keen to build wide machines! His system is therefore still in the development stages, but the more he learns about CTF, the more enthusiastic he becomes. He applies 70 t/ha of slurry and will soon be acquiring a new slurry tanker that not only spreads on the 12 m width that he wants, but will have narrower tyres and a wheel track that matches his other equipment. He can now harvest the crop at 12 m centres (by towing the forage wagon behind the harvester), but is waiting for a 12 m mower to complete the system (his present unit is 8.5 m). Further development will include integration of maize into the CTF system where he will inject slurry next to the seed. A CTF system means that he won t have to plough and will save a lot of fuel. He also expects to be able to work in a wider range of conditions; generally contractors have to stop working after just 5 mm of rain, even if the land has been deep ripped. We then visited two of his fields that had been in controlled traffic, one of which had been in CTF for three years (Figure 5). Although one could pick out the linear effect of the mower in places, this was far less obvious than the permanent tracks that had accommodated the slurry tanker, tedder, rake and harvester. Figure 5. This field had been in CTF for three years and the main tracks at 12 m spacing are obvious, but those left by the mower at 8.5 m centres were only visible in some places. 80

81 Erling Kjær, Rødding Erling was born on this farm and has spent 30 years in farm contracting. Forage grass was his main business for the first 25 years, taking this from seeding to harvest. He was the first to start offering a service for maize production and now covers 1500 ha, of which 300 ha are organic. The main difficulty with organic is keeping the crop clean from weeds. Five years ago he started with slurry and straw baling. He now has ten permanent employees, who spend the winter building houses. For slurry application he used to have a self-propelled Terragator but found the width (8 m) and tank size (16 m 3 ) to be too small and as there was no wider slurry injector on the market, he built his own. Similarly, Erling has built a 16 row (12 m) maize planter (Figure 6). Figure 6. The 16 row (12 m) maize planter. With his forage equipment he has now matched everything, other than the mower, to work on a CTF system at m, the latter being the width of his slurry injector. This machine folds to a road width of 4.2 m, which he considers too wide, although it is no wider than many others. So far he has had no difficulty with the permanent wheel tracks, but he does use wide tyres, with 1.5 bar inflation pressure in the wagons and 1.2 bar in the tractors. The wagon wheels are driven. Complete matching of the system will be achieved when he takes delivery of a 14.5 m JF-Stoll mower later this year. Track width is based on 2 m although the harvester does have a track width of around 3 m. His clients have not asked for CTF, but Erling decided that he wanted to do the job better and the only extra cost has been the guidance system that is used on the slurry spreader. Whether farmers will pay extra for his services relies heavily on whether he can demonstrate that he is providing them with extra crop yield and quality. If his clients decide to cut their grass with their own machinery, he needs the guidance system to find the tracks from his previous pass. Three or four cuts are made each year. At the moment, maize does not fit into his CTF system as he has a 12 m, 16 row maize planter coupled with a 6 m harvesting head. He said that the soil was spoiled during maize growing and harvesting and the only solution was to plough after this crop. The next page provides a cameo of his equipment. 81

82 Discussions I can only report on the discussions that took place within my earshot and those that I noted! Hopefully, many of you will have had rewarding interactions with other participants and will have taken these away with you. Design of forage transport systems Transferring cut and chopped forage to a wagon within a CTF system was widely discussed and presents some challenges. As we heard from Ole, he has used a tractor and trailer directly behind the self-propelled (SP) harvester (see photo), but this will not be attractive to all. A unit attached to the harvester is more attractive, but with big wagons, the towing attachments on SP harvesters often aren t strong enough. But, this can be overcome (Photo 1. Erling s machine). Representatives from the Danish importers of both John Deere and Claas SP harvesters said that these could be delivered with appropriate hitch points for large wagons. Providing wagons can be changed within 30 seconds, this was considered an acceptable method, although drivers may dislike this option because they have to get out of the cab to make hydraulic and electric connections. An alternative solution suggested was for the forage harvester to drive in between the fixed tracks and to blow grass into wagons running on the fixed tracks. Damage from the harvester may be minor if rutting is avoided. Standardization Agricultural machines should be produced at integer rather than fractional widths. For forage grass, 12 m was considered suitable but mowers and rakes may need some extra width to avoid skips. If combinable crops need to be integrated, a 9 m system may be more suitable. A wheel track width of 3 m is in most cases unsuitable due to road travel with wagons. A 2 m track width means that the overall width of tractors and wagons may exceed 2.55 m by only a small amount. SP harvesters and slurry tankers on the other hand typically exceed 2 m and therefore have an overall width greater than 2.55 m. Road width This is a considerable constraint on CTF systems, particularly in Germany where the 2.55 m limit is more stringently adhered to than in some other European countries. Transport containers Surprisingly these do have an impact on CTF. An example is the 12 m cutting platform offered by Claas which only has a cutting width of 11.5 m. This is because the unit has to fit into a standard transport container whose inside dimensions are 12 m. If special containers have to be used, these are more costly. 82

83 Conclusions Achieving CTF in forage production systems is perfectly feasible and practitioners are experiencing rising yields, largely as a result of avoiding grass damage caused by the tearing associated with high tractive loads, particularly on slopes. Achieving complete matching of machines takes planning, attention to detail and dedication and is enhanced by suppliers such as JF-Stoll who are prepared to supply customised machines. However, some degree of standardization should be considered to make it easier for manufacturers to supply appropriate equipment. Adoption of RTK GPS guidance systems is a pre-requisite for cost-effective and practical implementation of CTF. The farmers who have set out to achieve CTF have become more enthusiastic for the system following time and experience. Acknowledgements The organizers are indebted to the Danish farmers who gave so willingly of their time and so enthusiastically shared their experiences with us. We thank them and their supporting staff sincerely for their hospitality. Thanks are also due to Ole Green and the University of Århus for hosting the first part of our meeting and to JF-Stoll for their presentation. M ichelin Cargo B ib 91 0/5 0R26.5 M ich elin Cargo Bib 1050/50R32 Photo 1. New Holland FX60 forage harvester and wagon, showing strengthened towing point on the harvester (insets) and the tyre sizes (Erling Kjær photo cameo) 83

84 Samson 14.5 m slurry injection system pulled by Fendt 936 Vario on 1100/45R46 tyres. All up train weight was in the order of 59 t. Tanker tyres - six Trelleborg 850/50R