3 Simpósio Internacional de Agricultura de Precisão

Transcription

1 PRECISION FARMING FOR CEREAL CROPS MANAGEMENT GUIDELINES R J Godwin, G.A.Wood, J.C.Taylor Cranfield University at Silsoe, Silsoe, Bedford MK45 4DT, UK Abstract. The results of a 6 year study to develop practical management strategies for spatially developing nitrogen are reported. The results show that there are significant advantages from spatially varying nitrogen application rate in real time using the crop density of winter wheat in the spring period. Average economic benefits of 22/ha have been found together with a one third reduction in the residual nitrogen. From these results a strategy has been developed to aid farmers in varying the application rate of nitrogen given the crop status at selected growth stages. Yield maps where shown to be of little value for deciding upon the next seasons nitrogen application; there are however, valuable indicators for the replenishment of phosphorous and potassium. An economic analysis shows that benefits from precision farming are attainable for relatively small areas of land but the critical factor is the inherent field variability and the potential improvements to yield that are possible. Keywords. Precision farming, nitrogen, fertilizer, cereals, NDVI, yield maps, management guidelines. 1. Introduction Precision Farming is the term given to a method of crop management by which areas of land or crop within a field are managed with different levels of input in that field. The potential benefits are the economic margin from crop production may be increased by improvements in yield or a reduction in inputs, and the risk of environmental pollution from agrochemicals applied at levels greater than optimal can be reduced. These benefits are excellent examples of where both economic and environmental considerations are working together. This paper provides an overview of a 6-year study, with the aim of developing practical guidelines for the variable spatial application of nitrogen fertilizer. These have been fully reported elsewhere Godwin et al (2002 a and b) and in the Special Edition of Biosystems Engineering 84, (4), (2003) which contains a series of papers by the authors and their associates on a program of work funded by Home Grown Cereals Authority, AGCO and Hydro Agri (now Yarra). To this, recent work on variable fungicide application has been added together with the results of active reflectance methods for assessing crop variability. The duration of the study extended between in the fields detailed in Table 1, which were selected to provide a range of soils typical of approximately 30% of the land used for

2 arable production in England and Wales. These fields had predominantly been in cereals for several years prior to the experimental work Table 1: Field details and location Field name Location Soil Series* Cropping Pattern Far Sweetbrier Old Warden, Bedfordshire Hanslope Winter Wheat/ Oilseed Rape Onion Field Trent Field Houghton Conquest, Bedfordshire Goodworth Clatford, Hampshire Denchworth/Oxpasture/ Evesham Andover / Panholes Twelve Acres Hatherop, Gloucestershire Sherborne / Moreton / Didmarton Continuous Winter Wheat Continuous Winter Barley Continuous Winter Wheat Far Highlands Old Warden, Bedfordshire Wickham / Evesham Winter Wheat *after: Jarvis et al. (1984) and Hodge et al. (1984) At the outset it was agreed that the reasons for any underlying field variation needed to be established prior to managing the crop in a spatially variable manner. Hence, uniform 'blanket' treatments were applied in the 1995/6/7 seasons. Yield maps for these seasons, provided an indication of crop yield variation both in space and time. The effects of variable inputs were studied on all fields for the final three seasons. A number of fields planted with uniform seed rate were subjected to variable inputs of nitrogen. An additional two fields, Onion Field and Far Highlands, had variable nitrogen inputs applied across a range of seed rates that had been sown to create different crop canopy structures. Over the past decade the technology has become commercially available to enable the farmer to both spatially record the yield from a field and vary both seed and fertiliser rates on a sitespecific basis. Significant advances have also been made (Miller & Paice, 1998) to permit the spatial control of weeds on a site-specific basis by varying the dose rate of herbicides depending upon the weed density. However, the benefits of either an increase in yield and/or a reduction in fertilisers and agrochemicals have to be offset against the costs of investing in specialist equipment to enable yield maps to be produced and variable applications to be implemented. A range of potential benefits has been reported, from various combinations of different variable application rate practices. Earl et al. (1996) postulated a potential benefit of

3 33.68 ha -1 could be possible combining variable nitrogen application and targeting subsoiling to headlands for a crop of wheat in the UK, when wheat prices were 125 t -1. Measured benefits in the range of to ha -1 (-$6.37 to $42.38 ac -1 ) were reported by Snyder et al. (1998) on irrigated maize in the USA. Schmerler & Basten (1999) measured an average benefit of ha -1 (60 DM ha -1 ) when growing wheat on a farm scale trial where both seed and agrochemical rates were varied. Studies conducted by James et al. (2000) investigated the benefits of using historic yield data as a guide to varying nitrogen application, for winter barley on a field with both clay loam and sandy loam soil types. Data from the 1997 harvest reported by Godwin et al. (1999) indicated that a modest benefit of could be possible. Earl et al. (1996) estimated the costs of yield map production and the ability to apply fertiliser on a site-specific basis to be ha -1 for an arable area of 250 ha, at 7% interest rate amortised over a 5-year period in the UK. Studies in the USA, by Snyder et al. (1998) estimated the cost of yield mapping and variable rate equipment, for nitrogen application, for two fields of 49 and 64 ha as 8.50 ha -1 ($11.88 ha -1 ). Schmerler and Basten (1999) reported costs of ha -1 (49 DM ha -1 ) for a 7,100 ha German farm. The major reason for the higher figures was the cost of the equipment to variably apply herbicides in addition to the seed rate and fertiliser. This paper: (i) (ii) (iii) (iv) reports on the results obtained from a series of agronomic studies. examines the increase in revenues that have been achieved through the use of precision farming practices during a three-year study of 5 fields in cereal production in Southern England (Godwin et al., 2002 a and b, 2003). estimates of the costs of upgrading farm equipment, at the time of purchase, to a level that enables precision farming techniques to be practised. compares the costs/benefits and analyses the potential returns from adopting precision farming technology for given farm sizes and levels of variability, and demonstrates the how data has been made available to farmers to decide if adopting precision is economically viable for their farm.

4 2. Determining the Inherent Variability 2.1 Crop yield Typical variations in crop yield are presented in Figure 1, which shows that there is some similarity over the three-year period. The spatial trend map or average yield map (Blackmore, 2000) for the period shows that, the yield range for this particular field is in excess of ±20% of the mean. The higher yielding zones occur to the west and the lower yielding zones to the east of the 100% (or mean) contour. The variation in yield for the 4 main fields averaged between ±25 percent of the mean yield with a range of ±20% to ±33% Yield (% of grand mean) High Ave. Low Soil types Fig. 1. Average yield map for yield at Trent Field, The fields were initially surveyed at a commercial detail level of approximately 1 auger hole/ha to provide an overview of soil textural and profile variation. These were complemented by "targeted" profile pit descriptions. The location of the profile pits were selected to encompass (i) the range of yields observed in the yield maps of 1994/95 and 1995/96, (ii) (iii) the density of the crop from aerial digital photography captured in May 1996, and soil maps based upon auger sampling at 100 m grid spacing. Further studies with both soil coring apparatus (to a depth of 1 m) and electromagnetic induction (EMI) equipment increased the resolution to define soil textural boundaries (James et al, 2003). The latter technique is particularly useful for differentiating between soil textures as shown in Figure 2, where the higher levels of conductivity indicate higher

5 moisture content soils which if conducted at field capacity would indicate greater clay content. Objective techniques, using cluster analysis, have been developed which enable potential management zones to be determined using historic yield and EMI data (Taylor et al. 2003). Differences in soil nutrient levels have been identified between the management zones and, hence, form a basis for targeted sampling of soil nutrient status. Northing (m) EMI Apparent Soil Conductivity ms/m 3 to 6 6 to 8 8 to to to Easting (m) Fig. 2. EMI conductivity Trent Field 2 nd February Soil fertility and crop nutrition Detailed analyses of macro- and micro-nutrients in both soil water extract and plant tissue were conducted at approximately 50 m grid spacing together with soil ph. These indicated that there was variation in nutrient levels in each of the fields (Taylor et al; 2003). However, with the exception of isolated areas with low ph, the levels were above the commonly accepted limits. 2.4 Crop canopy Variations in crop canopy occur both in space and time in the same field. In order to obtain consistent and reliable data for monitoring crop development for 'real time' management and to explain field differences, a light aircraft was equipped with two digital cameras fitted with red (R) and near infra red (NIR) filters (Wood et al, 2003). Field images obtained from aerial digital photography (ADP) from a height of 1000 m give a pixel resolution of 0.5 m x 0.5 m. 0

6 Normalised Difference Vegetation Index (NDVI) values were estimated from the following equation: NIR R NDVI = NIR + R The resulting images, such as Figure 3, show the effect of variations in crop development immediately prior to the first application of nitrogen. These images are (i) immediately valuable in discerning patterns of field variability, and (ii) provide detailed spatial data on crop tillers/shoot density. These data, when calibrated against detailed agronomic measurements at targeted locations, were used in near "real time" to estimate crop condition and potential nutritional requirements. The extension of this principle to farm scale operations, using 8 sampling points, provides an effective calibration between the crop indicators and NDVI. The cost of this technique, for 3 flights/year, has been estimated at 7/ha (Godwin et al., 2002b and 2003b). Shoot Density (shoots m -2 ) > < 600 Fig. 3. Normalised Difference Vegetation Index (NDVI) image of Trent Field

7 3. Variable Application of Nitrogen 3.1 Field layout The main aim of this work was to develop an experimental methodology that could be employed by farmers to determine an optimal application strategy for a given input in any particular field. To achieve this, it was important to use standard farm machinery for the experiments. The proposed design comprised a series of long strips, which ran through the main areas of variation within each field, an example of which is presented in Figure 4, where the treatment strip is interlaced with the field standard (Welsh et al., 2003a, b). The treatment strips were half the width of a tramline, nominally 12m. The fertilizer was applied using a pneumatic or liquid fertiliser applicator. These strip widths allowed the experiments to be harvested by the combine harvester without harvesting the zones affected by the tramline wheel marks. The combine was equipped with a yield sensor, with a mean instantaneous grain flow error of 1% (Moore 1998). 3.2 Historic yield and shoot density studies These treatment strips were established to test the following strategies: (i) (ii) increasing the fertiliser application to the higher, or potentially higher, yielding parts of the field whilst reducing the application to the lower yielding parts. reducing the fertiliser application to the higher, or potentially higher, yielding parts of the field whilst increasing the application to the lower yielding parts. However, before these strategies could be implemented, the high, average and low yielding zones had to be identified. Two methods were used: (i) historic yield data, as shown in Figure 2.and (ii) shoot density data, estimated from NDVI data, as shown in Figure 3. Using this approach, experimental strips (Figure 4) were established to give the following treatments: Historic Yield 1 (HY1). High yield zone received 30% more nitrogen; average yield zone received the standard nitrogen rate; and the low yield zone received 30% less nitrogen. Shoot Density 1 (SD1). High shoot density zone received 30% more nitrogen; average shoot density zone received the standard nitrogen rate; and the low shoot density zone received 30% less nitrogen. Historic Yield 2 (HY2). High yield zone received 30% less nitrogen; average yield zone received the standard nitrogen rate; and the low yield zone received 30% more nitrogen. Shoot Density 2 (SD2). High shoot density zone received 30% less nitrogen; average shoot density zone received the standard nitrogen rate; and the low shoot density zone received 30% more nitrogen.

8 Standard N rate strips were located adjacent to each of the variable treatment strips to allow treatment comparisons to be made, since classical experimental design and statistical analyses with replicated plots was not possible. Uniform Treatments Variable Treatments Low Yield Standard N Rate Standard N Rate + 30% Standard N Rate - 30% Standard N Rate Historic Yield 1 (HY1) Historic Yield 2 (HY2) Standard N Rate Shoot Density 1 (SD1) Shoot Density 2 (SD2) Standard N Rate Variation High Yield 12 m 3.3 Crop canopy management studies Fig. 4. Plan of field experiments The methodology for these studies was developed over three years in Onion Field, but was extended to include Far Highlands in the final season. Seed rates of 150, 250, 350 or 450 seeds/m 2 were used to establish 24 m wide strips of wheat with a range of initial crop structures. The strips were then subdivided into two 12 m wide sections, along which one received a standard field rate of nitrogen fertiliser (200 kg N/ha), and the other a variable amount dependant upon crop growth. Observations were made in near real time using the aerial digital photographic technique and crop canopy measurements described earlier. Appropriate flights were made prior to each of the three nitrogen application timings in the February to May period, and crop growth (shoot populations at tillering and canopy green GS30-31 and GS33) compared with benchmarks from the HGCA Wheat Growth Guide (1998).

9 4. Results 4.1 Historic yield and shoot density studies An example of the yield distribution along the variable treatment yield strips is presented in Figure 5 for the HY1 and HY2 strategies. The effect of increasing (160 kg N ha -1 ) and decreasing (90 kg N ha -1 ) the nitrogen application rates to the high and low yielding zones in comparison with the field standards can be clearly seen. This shows that for Trent Field in 1997/98 there were advantages of adding fertiliser to both the high and low yielding zones and penalties for reducing the rate. The results in Table 2 indicate that there are no economic benefits from HY1 and HY2 in Trent Field or Twelve Acres. The reason for this is due to the reduction in nitrogen application rate causing a significant yield loss in both the high and low yielding zones, which are not compensated for by savings in nitrogen costs. Yield (t ha -1 ) High yield zone All 125 Ave. yield zone 160 HY1 Standard HY Low yield zone Distance along strip (m) Fig. 5. Combine yield of Historic Yield treatments (HY1 & HY2) compared with a standard application along the treatment strips in Trent Field. Shaded areas are transition zones and are deleted from the analysis (after Welsh, 2003a)

10 Table 2: Economic consequences ( /ha) of 3 years of alternative nitrogen management scenarios for all fields in comparison to a standard application rate Strategy Trent Field Twelve Acres Far Sweetbrier Mean HY HY * -9.53* SD SD *contains data from 1998/99 and 1999/00 only Managing the crop using maps of the relative shoot density from NDVI data provided a positive benefit when more nitrogen was applied to the areas of low shoot density, and less to the high density areas (SD2). The success of this, however, depended on the actual shoot populations present, which differed between seasons. This occurred because there was little variation along the strip with a low shoot density, which from hindsight using the principle of canopy management would respond best to a uniform application of nitrogen. Overall, the shoot density SD2 approach which uses a real-time assessment of the crop canopy/ structure to control the nitrogen requirement appeared to offer the greatest potential for crop production. Nitrogen strategies based on historic yield maps (HY1 and HY2) showed no or very little benefit. Yield maps are, however, a valuable tool for: (i) (ii) the replenishment of potassium and phosphorous removed by the previous crop, and identifying the size of the zones needing particular attention from other field factors identifying the size of the zones needing particular attention from factors such as the impact of water-logging, ph and uneven fertilizer application. 4.2 Canopy management studies The results presented in Table 3 are a comparison of both the yield and the economic (Gross Margin) performance of the variable and uniform nitrogen strips. Also shown are the mean of the variable nitrogen application rate and the uniform rate for both fields. These show that regardless of seed rate in Onion Field both the yield and the gross margins for the variable nitrogen strategy exceeded those for the uniform practice. The similar data from Far Highlands show yield benefits at the lowest seed rate only. The other 3 seed rates show a small reduction in yield, which was economically compensated for by lower nitrogen application.

11 Table 3 Nitrogen application rates (N) (kg N ha -1 ), Yield (Y) (t ha -1 ) and gross margin (GM) ( ha -1 ) comparisons between variable and uniform nitrogen application strategies (after Wood et al., 2000b) Target Seed Rate (seeds m -2 ) Plant population (plants m -2 ) Onion field N Y GM N Y GM N Y GM N Y GM Variable N Uniform N Difference Far Highlands Plant population (plants m -2 ) N Y GM N Y GM N Y GM N Y GM Uniform N Standard N Difference The financial benefits showed that in seven of the eight comparisons the variable N management out performed the uniform application. The maximum advantage to variable N management was 60/ha that was produced from a combination of higher yield (+11%) and a slightly lower total N input compared to the standard N approach. Overall yield benefits were greatest where the mean application rate of the variable nitrogen strips was approximately that of the field standard. On average, for the two fields, the overall benefit of the variable nitrogen strategy was 22 ha -1. An analysis of the responsive areas to variable nitrogen in both the shoot density and canopy management studies indicate that between 12% and 52% of all fields responded positively and depending upon field and season 5. Environmental Implications Whilst this work did not specifically address environmental implications of nitrogen usage patterns was possible to draw some conclusions on the possible impact of precision farming on the nitrogen balance. This was achieved by calculating the potential off-take of nitrogen in the variable treatment compared to the standard from the strip mean grain yields, the average

12 fertiliser N application rates, and the grain and straw nitrogen contents by assuming the straw yield was equal to 65% of grain yield. N deficit N surplus kg hā Uniform N Variable N [100] 250 [143] 350 [177] 450 [200] 28 Bayes 95% confidence level 13 Seed Rate [Plant population] (seeds m -2 [plants m -2 ]) The plant populations in Onion Field were generally low and in the lowest seed rate (which produced only 100 plants/m 2 ) both the uniform and variable nitrogen programmes had nitrogen off-takes which were significantly lower than the amount applied. This resulted in a surplus at the end of the season, see Figure 6. Fig. 6: Surplus or deficit of applied nitrogen relative to off-take in grain and straw at Onion Field in 2000 (after Wood et al., 2002b) However at the three higher plant populations the off-take from the variable N applications were higher than applied N resulting in a net reduction in N balances. Averaged over the four seed rates, the N surplus for the variable treatments was 18.5kg/ha compared to 28kg/ha for the uniform treatments. This represents a 34% reduction in the net amount added to the soil from the uniform application and this could have considerable environmental significance. A similar analysis was conducted for Far Highlands by assuming the grain and straw nitrogen contents were similar to Onion Field, in this case the average saving from the variable N treatments compared to the uniform N treatments was 32.5kg/ha. 6. Potential benefits from adopting precision farming The potential benefits considered in this paper arise from managing crop canopy in real time by varying nitrogen in two fields, growing winter wheat (Wood et al., 2003b). The difference in the economic performance of variable and standard nitrogen application for a range of seed rates is summarised in Table 1. Table 4 shows that the benefits of variably applying nitrogen in comparison with uniform application, based on the standard farm practice, range from - 1 ha -1 (Far 195 plants m -2 ) up to 60 ha -1 (Onion 200 plants m -2 ). The mean benefit over all seed

13 rates in Onion Field was ha -1. The overall mean improvement in Onion Field and Far Highlands was 22 ha -1, marginally less than reported in Godwin el al. (1999). Table 4 Economic comparisons of variable and uniform nitrogen application rates Target Seed Rate (seeds m -2 ) Establishment (plants m -2 ) Onion Field Gross margin ha -1 Variable N Uniform N Difference Establishment (plants m -2 ) Far Highlands Gross margin ha -1 Variable N Uniform N Difference In addition to the benefits already mentioned, other factors listed in Table 5 need also to be considered. Table 5 Other economic considerations Factor Implication Penalty or Benefit Water-logging Economic penalty Up to 195 ha -1 Fertiliser application errors Economic penalty Up to 65 ha -1 ph Economic advantage Up to 7 ha 1 Herbicide application* Economic advantage Up to 20 ha -1 * -1 ** Fungicide application** Economic advantage Up to 22 ha * after Rew et al. (1997), ** after Secher (1997)

14 During the course of this project three examples of the above additional benefits of precision farming were recorded. The cost of water logging, shown as up to 195 ha -1, was calculated using a potential yield reduction of 3 t ha -1, experienced at one of the trial sites following a wet period in the winter of 1998/99 (Wood et al., 2000). This could have been rectified for a one off cost of 50 ha -1 (Nix, 2000) for mole draining the site and clearing blocked drain outlets that could have an economic life in excess of 5 years. Uneven distribution of fertiliser resulted in a yield penalty of up to 1 t ha -1. The cost of failing to rectify problems involving ph levels was estimated to be up to 7 ha -1. The collection of these data using yield-mapping techniques enables simple cost/benefit analyses to be conducted to ascertain the scale and extent of the problem(s), from which estimates of the cost of correction can be made to compare with the potential long-term benefits. Economic benefits resulting from the site-specific control of herbicide (Rew et al., 1997 and Perry et al., 2001) and fungicide (Secher, 1997) application are included in Table 5. The reported cost savings for herbicides range between 0.50 and ha -1, and were achieved by targeted application using patch-spraying techniques. A statistically significant yield increase of 0.3 t ha -1, equivalent to a revenue increase of ha -1, has been achieved by varying fungicide application rate according to crop canopy density. Albeit a more recent study by the authors in Bedfordshire, Hampshire and Lincolnshire showed that, whilst there were regional differences in the desired dose rates, when using remote sensors to apply more or less fungicide according to estimates of within field variability of canopy density, this proved of little agronomic value and as a result no economic worth. It can be seen from Table 5 that the potential economic penalties of normal field management problems can outweigh the highest increase in benefit achieved from spatially varying nitrogen fertiliser. It is, therefore, imperative these problems are addressed prior to considering the use of spatially varying nitrogen. 7. Estimation of the costs of precision farming systems 7.1 Precision farming monitoring and control systems A full precision farming (PF) system comprises hardware and software to enable variations in crop yield to be mapped and crop related treatments to be variably applied on a site-specific basis. In reviewing the literature it is apparent the cost of practising precision farming techniques is dependent on: (i) (ii) (iii) the level of technology purchased, i.e. a full or partial system, depreciation and current interest rates, and the area of crops managed.

15 To determine the realistic cost for UK conditions an analysis was conducted, based on prices quoted by main suppliers of precision farming equipment, in January It was apparent that precision-farming systems could be categorized into four main classes. Namely: Class 1. Class 2. Class 3. Class 4. Comprises a fully integrated system from an original equipment manufacturer. Comprises a full system from a specialist manufacturer. Comprises a full system, which is a combination of specialist and OEM. Comprises a basic system from an OEM. Most new combine harvesters can be fitted with yield mapping hardware, however, the degree of integration between the yield mapping system and other components of the combine operating and performance monitoring system varies between manufacturers. The systems range in functionality from fully integrated yield mapping and combine performance monitoring systems, that can be removed from the combine and fitted to tractors or sprayers and include sub-metre DGPS (Class 1 at 11,363) through to low cost partial systems that provide full yield mapping functionality but reduced application rate control functions (Class 4 at 4,500). The remaining two classes comprise, Class 2 (at 14,100) is a full precision farming system produced by specialist manufacturers, and Class 3 (at 16,150) is an addition of parts of Classes 1 and 2 which comprise an OEM integrated yield and combine performance monitor with components from specialist manufacturers to be mounted in either tractors or spray vehicles for variable application rate control. This has the added advantage that the parallel systems enable both harvesting and application control to be undertaken at the same time. The basic system (Class 4 at 4,500) uses a non-differential GPS to provide position information to +10 m. This provides the operator with the capability to produce yield maps of a slightly lower resolution than those produced using full precision farming systems, but probably sufficiently accurate for most management tasks. Variable application rates are achieved through changing the tractor forward speed whilst maintaining a constant material flow from the applicator in use. The speed control is achieved by the operator manually attempting to match a target speed displayed on the on board vehicular computer screen. This provides a limited range over which the application rate can be varied, dependant on the tractor transmission type, but does permit farmers to make initial ventures into precision farming management without a large capital outlay. 7.2 Assumptions used and the basis of the cost calculations The costs are based on the following assumptions: (i) one set of variable-application crop treatment equipment, i.e. PF-system, can farm an identical area to that harvested by the combine,

16 (ii) (iii) operations involving variable application equipment are not conducted at the same time as combine harvesting (with the exception of Class 3), for multiple PF-systems the total area would be divided equally between units. The basis of the additional costs associated with purchasing precision farming equipment are summarised in Table 6 for all systems. Table 6 Summary of the cost of precision farming equipment PF equipment Cost Full system Class 1 Class 2 Class 3 Basic system Class 4 Initial capital cost Cost of capital 8.5 % 8.5 % Depreciation all equipment 13% for 5 yr replacement 13% for 5 yr replacement Maintenance Combine 3.5% for 150 hrs use pa 3.5% for 150 hrs use pa Tractor 8% for 1000 hrs use pa 0 Seed drill 7.5% for 150 hrs use pa 0 Fertiliser distributor 7.5% for 150 hrs use pa 0 Training 60 pa ( 300 over 5 yr) 60 pa ( 300 over 5 yr) 7.3 Annual cost per unit area This has been calculated for a range of arable areas that could be managed using a single PFsystem i.e. the vehicle mounted computer used to record yield when fitted to the combine harvester and control application rate when fitted to the tractor. Fig 7 shows the total annual cost ha -1 for the range of systems. This shows that the cost of the basic system (Class 4) at 5 ha -1 for an area of 250 ha is significantly cheaper than the full systems. Values for the full systems range between 12 ha -1 and 18 ha -1. Doubling the area to 500 ha reduces the cost to 2.50 ha -1 and 6-9 ha -1 respectively. These figures show very clearly the effect of the area per PF-system on the annual cost of the operation, with the costs becoming asymptotic to the horizontal axis as the farmed area increases.

17 Annual cost ( ha -1 ) Area per PF-system (ha) Class 3 Class 2 Class 1 Class 4 8. Other costs Fig 7 Total cost of four different Precision Farming systems 8.1 Soil texture and chemical analysis Soil texture can be determined by traditional manual surveying techniques from auger samples on an approximate 100 m grid basis or the more recently developed electromagnetic induction techniques (James et al. 2003). These costs are based upon a cost per hectare as given in Table 7 and should be viewed as a one-off investment. Table 7 Typical fixed area costs One-off cost ha -1 Annual Cost sample -1 Soil surveying (manual) 25 P, K, ph, Mg 9 Soil surveying (E.M.I.) P, K, ph, Mg + Cu, B 20 Available N a) upper, middle and lower samples from 0.9 m deep core 100 b) 0.9 m core bulked together 40 Soil nutrient status is, however, determined upon a cost per sample; current sampling and analysis costs for a range of nutrients are given in Table 7. This indicates that nitrogen analysis is expensive if undertaken annually, with one sample per hectare, and explains why there is great interest in targeting the samples needed for this and similar analyses based upon the management zone concept reported in Taylor et al. (2003).

18 8.2 Crop canopy status The costs to assist in the management of the crop canopy in near real time using crop reflectance data relate to either aerial digital photography (ADP) or tractor-mounted radiometers (TMR). The hardware required for obtaining remotely sensed data comprises a pair of digital cameras for use in ADP mounted in a light aircraft (Wood et al, 2003a) or a tractor mounted radiometer (Boissard et al., 2001). The annual depreciation and maintenance costs for these, summarized in Table 8, have been calculated using the same assumptions as used in the earlier sections. The TMR may also be hired. For the farm scales in the UK it is most likely that a service provider, agronomy consulting group or a syndicate of farmers would make the substantial investment for the digital camera system for ADP. The cost would, therefore, be spread over a large area. Table 8 Cost associated with acquiring crop reflectance data TMR Purchased TMR Hired ADP Cameras Hardware cost ( ) 10,000-15,000 Annual costs ( ) 13% 1,300-1, % Cost of 8.5% 850-1,275 Rental charges - 4,000 - Total annual cost ( pa) 2,500 4,000 3,750 Cost of ground calibration ( ha -1 ) In order to estimate the cost per hectare of acquiring the crop reflectance data using ADP it has been assumed that:(i) each 3 hour flight could cover up to 3,650 ha and that each field would need to be photographed prior to each application of nitrogen at 3 growth stages, and (ii) it is possible to make 2 flights per day. The cost of data collection, is presented in Fig 8 this includes the cost of the plane, pilot, cameras and the technicians to perform the image calibration in the field. It can be seen that the cost is almost independent of the area flown above 1000 ha, and at 1500 ha (a typical day s work for collecting the ground calibration data) would cost 7 ha -1. The cost of the tractor-mounted radiometer (TMR) is more likely to be borne by an individual farmer or a small syndicate of farmers. The TMR could provide similar, but lower resolution, data to ADP for use in producing fertiliser application plans. The cost per ha has been calculated, as a function of the area managed per radiometer, using the data in Table 8.

19 The difference between owning ( ha -1 ) and renting ( ha -1 ) a TMR being 3 ha -1 at 500 ha, however, to be competitive with ADP it would be necessary to manage an area in excess of 1500 ha Cost ( ha -1 ) Area per flight (ha) Fig 8 Cost of acquiring crop reflectance data using aerial digital photography Currently the authors and their research students, Havrankova (2006), Morris (2006) and Wilson (2006), are evaluating the advantage of a new radiometer (Crop Circle, Francis et al., (2005) and Sudduth et al., (2005)) which can be vehicle mounted and contains its own light source. The capital cost of which is less than tractor mounted radiometer referred to above. 9. Breakeven analysis The breakeven analysis has been based on a benefit of 15 ha -1. This has been calculated by subtracting the 7 ha -1 cost of acquiring ADP data from the 22 ha -1 benefit achieved by varying nitrogen application according to crop needs assessed using real time monitoring of the canopy in Onion Field and Far Highlands. In order to estimate the area per PF-system required to break even the mean benefit of 15 ha -1 has been compared with the cost of the four different classes of PF-system shown in Fig 7. It can be seen from Fig 9 how the increase in system cost increases the area per PF-system required for breakeven at an economic return of 15 ha -1. This shows that for a low cost basic system precision farming can be economically viable for areas in excess of 78 ha, rising to 308 ha for the most expensive system.

20 Cost ( ha -1 ) Area per PF-system (ha) Class 3 Class 2 Class 1 Class 4 15 ha -1 Fig 9 Breakeven area per systems for a return of 15 ha Sensitivity analysis of field variability The scale of any benefit obtained from adopting precision farming practices will ultimately depend on the magnitude of the response and the proportion of the field (%) that will respond positively to variable management. The increase in yield required to break even for different levels of field variability has been calculated using the costs based on a Class 3 system and grain at 65 t -1 is shown, as an example, in Fig 10. The proportion of the field (%) responding positively to variable nitrogen management is based upon data from the shoot density and crop canopy studies.

21 Yield increase required in variable parts of the. field to achieve 65 t -1 (t ha -1 ) Area per PF-system (ha) Percentage area of the field likely to produce a positive response to variable inputs 10% 20% 30% 50% Fig 10 Field variability sensitivity analysis Class 3 ( 16,000) system It can be seen from Fig 10 that for an area of 250 ha, and response areas of 30% a minimum yield increase of c1.0 t ha -1 would provide a breakeven return. Figures similar to those above have been produced for all the Classes of precision farming system and incorporated into Table 9 for practical on farm use. This requires farmers and/or agronomists to answer the following questions and then refer to Table 9 to estimate the potential for their farm. 1. How large is the cereal area farmed by a precision farming system? Choose the area up to 1000ha. 2. What is the proposed capital investment in the precision farming system? Choose the system cost between 4500 and What percentage of the total cereal area has the potential for improvement? Estimate an area between 5% and 30% from agronomic assessment of field and yield map variability. 4. Am I likely to exceed the breakeven yield benefit shown in the table from applying precision farming techniques to those areas? To illustrate this, it can be seen from Table 9 that cereal areas of: ha with a system costing 4500, where 10% of the area would respond positively, is economic with a yield increase of 0.24t/ha on that area.

22 2. 500ha with a system costing 11000, where 20% of the area would respond positively, is economic with a yield increase of 0.48t/ha on that area. Table 9 System (investment) Basic level entry ( 4,500) Practical Guideline to estimate the potential economic benefit of precision farming (HGCA, 2002) Level of yield increase (t/ha) needed to justify investment* % area 250 ha 500 ha 750 ha 1,000 ha responding 5% % % % Fully 5% integrated 10% single unit 20% ( 11,000) 30% Multiple units for combine and tractor ( 16,000) *based on wheat at 65/t 11. Conclusions 5% % % % Yield maps are indispensable for targeting areas for investigation and treatment by precision farming practices and subsequent monitoring of results. They provide a valuable basis for estimating the replenishment levels of P and K fertilisers, however, they were not found to provide a useful basis for applying spatially variable nitrogen. 2. The spatial variation in canopy development within a field can be estimated using an ADP technique for real-time agronomic management. This technique can be extended from field scale to farm scale for crops of similar varieties and planting dates. The technique can be used as a basis for determining the most appropriate application rate for nitrogen, and as a guide for herbicide and plant growth regulator application. 3. The application of nitrogen in a spatially variable manner can improve the efficiency of cereal production through managing variations in the crop canopy. Between 12% and 52% of the area of the fields under investigation responded positively to this approach. In 2000 seven out of eight treatment zones gave positive economic returns to spatially variable nitrogen with an average benefit of 22 ha -1.

23 4. Simple nitrogen balance calculations have shown that the spatially variable application of nitrogen can have an overall effect on reducing the nitrogen surplus by approximately one third. 5. Based upon nitrogen and cereal prices at 0.30 kg -1 and 65 t -1 respectively, and for equipment prices in the UK in January 2001 the benefits of the variable rate application of nitrogen provided an average improvement of 22 ha The annual costs per hectare of the systems vary between 4.67 ha -1 and ha -1 for the basic and most expensive system respectively for an area of 250 ha. 7. The benefits outweigh the additional costs of the investment for cereal farms greater than 80 ha for basic systems and ha for the more sophisticated systems 8. The costs of detailed soil analysis prohibit collection from a dense grid of data points and targeted sampling is recommended. 9. Common field management problems can result in significant crop yield penalties and should be corrected prior to spatially variable application of nitrogen. 10. The benefits obtained from precision farming practices depend upon the magnitude of the response to the treatments and the proportion of the field, which will respond. As a result of these studies a flow chart has been produced to assist cereal farmers in the decision making required for variable nitrogen application (HGCA, 2002) the outline of which is given in Appendix1. Acknowledgements The authors would like to thank the sponsors of this work, Agco Ltd., BASF, Home-Grown Cereals Authority and Hydro Agri and the contributions made by their collaborators, Arable Research Centres and Shuttleworth Farms. We would also like to thank Jana Havrankova, Jim Wilson, David Morris and Robert Walker for their assistance. Thanks must also be extended to Messrs Dines, Hart, Welti, Wilson and Wisson who allowed us to use their fields. References Blackmore, B. S The interpretation of trends from multiple yield maps. Computers and Electronics in Agriculture, Elsevier, Vol 26, No.1 pp Boissard; P., Boffety; D., Devaux; J.F. Zwaenepoel; P., Huet; P., Gilliot; J.-M. Heurtaux; J., Troizier, J Mapping of the Wheat Leaf Area From Multidate Radiometric Data Provided by On Board Sensors. In: Proceeding of the 3 rd European Conference on Precision Agriculture, Earl, R., Wheeler, P N., Blackmore, B S., Godwin, R.J Precision Farming the Management of Variability. Journal of the Institute of Agricultural Engineers, 51(4),

24 Francis, D.D., Schlemmer, M.R., Schepers, J.S., Shanahan, J.F., Luchiari, A Performance of a Crop Canopy Reflectance Sensor to Assess Chlorophyll Content. Godwin; R.J., James; I T., Welsh; J P., Earl, R. Managing spatially variable nitrogen a practical approach ASAE/CSAE-SCGR Paper No St. Joseph Mich., USA. Godwin; R J., Wood; G A., Taylor; J C., Welsh; J P., Knight; S., Blackmore, B S. Management guidelines for precision farming: Nitrogen. 2002a. ASAE Meeting Paper No St Joseph, Mich., USA. Godwin, R. J., Richards, T. E., Wood. G. A., Welsh, J. P., Knight, S. Economic Analysis of Precision Farming Systems. 2002b. ASAE Meeting Paper No St. Joseph, Mich., USA. Godwin; R J., Wood; G A., Taylor; J C., Knight; S., Welsh; J P Precision Farming of Cereal Crops: a Review of a Six Year Experiment to develop Management Guidelines. Biosystems Engineering. 84, (4), Godwin, R. J., Richards, T. E., Wood. G. A., Welsh, J. P., Knight, S An Economic Analysis of the Potential for Precision Farming in UK Cereal Production. Biosystems Engineering. 84, (4), Havrankova, J Ground based remote sensing systems for determine canopy nitrogen in winter wheat. Unpublished MSc by Research Thesis, Cranfield University at Silsoe, Silsoe, Bedford MK45 4DT, UK. HGCA.The Wheat Growth Guide HGCA, London. HGCA. Precision farming of cereals - practical guidelines and crop nutrition HGCA, London. Hodge, C.A.H., Burton, R.G.O., Corbett, W.M. Evans, R. and Seale, R.S Soils and their use in Eastern England. Soil Survey Bulletin No.15, Harpenden. James; I.T. Earl; R., Godwin. R.J On farm development of a variable rate nitrogen fertilizer strategy. Paper 00-PA-012.AgEng 2000, Warwick,UK. James; I.T., Waine; T.E., Bradley; R.I., R.J., Taylor; J.C., Godwin; R J Determination of Soil Type Boundaries using Electromagnetic Induction Scanning Techniques. Biosystems Engineering. 84, (4), Jarvis, M.G., Allen, R.H., Fordham, S.J., Hazelden, J., Moffat, A.J. and Sturdy, R.G Soils and their use in South East England. Soil Survey Bulletin No.13, Harpenden, UK. Miller, P.C.H., Paice, M.E.R Patch Spraying Approaches to Optimise the Use of Herbicides Applied to Arable Land. Journal of the RASE, 159,

25 Moore, M.R An investigation into the accuracy of yield maps and their subsequent use in crop management. Unpublished Ph.D. Thesis Cranfield University, Silsoe, Bedford, U.K. Morris, D Precision Farming of Cereals in Northern Ireland. Unpublished MSc by Research Thesis, Cranfield University at Silsoe, Silsoe, Bedford MK45 4DT, UK. Nix, J. Farm Management Pocketbook 31 st edition, Imperial College Wye, UK. Perry, N.H., Lutman, P.J.W., Miller, P.C.H., Wheeler, H.C A Map Based System for Patch Spraying Weeds (1) Weed Mapping. In: Proceedings BCPC Crop Protection Conference Brighton Weeds 2001, Brighton, UK. Rew, L.J., Miller, P.C.H., Paice, M.E.R The Importance of Patch Mapping Resolution for Sprayer Control. Aspects of Applied Biology, 48, Schmerler, J., Basten, M Cost/Benefit Analysis of Introducing Site-Specific Management on a Commercial Farm. In: Precision Agriculture 99, The Second European Conference on Precision Agriculture Secher, B.J.M Site Specific Control of Diseases in Winter Wheat. Aspects of Applied Biology, 48, Optimising Pesticide Applications, Snyder, C., Havlin, J., Kluitenberg, G., Schroeder, T Evaluating the Economics of Precision Agriculture. In: Proceedings of the Fourth International Conference on Precision Agriculture, Madison, WI, USA Sudduth, K., Kitchen, N., Scharf, P., Palm, Harlan., Shannon, K. and Hummel, J. (2004). Hardware and Software Systems for In-Season Variable-Rate Nitrogen Application. USDA Agricultural Research Service, University of Missouri, USA. Taylor; J.C. Wood; G.A., Earl; R., Godwin, R.J Soil factors and their influence on within field crop variability II: Spatial analysis and determination of management zones. Biosystems Engineering. 84, (4), Welsh, J.P., Wood, G.A., Godwin, R.J., Taylor, J.C., Earl, R., Blackmore, S., Knight, S. 2003a. Developing Strategies for Spatially Variable Nitrogen Application in I: Barley, Biosystems Engineering. 84, (4), Welsh, J.P., Wood, G.A., Godwin, R.J., Taylor, J.C., Earl, R., Blackmore, S., Knight, S. 2003b. Developing Strategies for Spatially Variable Nitrogen Application in II: Wheat, Biosystems Engineering. 84, (4), Wilson, J Variable seed spacing for uniform size of seed potatoes. Unpublished MSc by Research Thesis, Cranfield University at Silsoe, Silsoe, Bedford MK45 4DT, UK. Wood, G.A., Welsh, J.P., Godwin, R.J., Taylor, J.C., Knight, S., Carver, M.F.F Precision Farming: seed-rate and nitrogen interactions. HGCA Crop Management for the Millennium Conference, Cambridge.

26 Wood, G.A., Taylor, J.C., and Godwin, R.J. 2003a. Calibration Methodology for Mapping Within-Field Crop Variability using Remote Sensing. Biosystems Engineering. 84, (4), Wood, G.A., Welsh, J.P., Taylor, J.C., Godwin, R.J., Knight, S. 2003b. Real Time Measures of Canopy Size as a Basis for Spatially Varying Nitrogen at Different Seed Rates in winter Wheat. Biosystems Engineering. 84, (4),

27 Appendix 1 Nitrogen management for Winter Wheat (HGCA, 2002)