P. MILATA
Leading Farmers CZ, a.s., Klánovická 485/43, 198 00 Praha 14, Česká republika

ABSTRACT
Precision farming technologies enable spatially variable management of inputs application for plant production (fertilisers, seeds, irrigation) as well as soil processing. Various methods of variable application of plant nutrients, mainly in the form of chemical fertilisers, are the most spread commercially. Mainly three principles are used as a basis for setting of variable fertiliser rates: soil characteristics (mineral nutrients content, pH, organic matter content, soil types etc.), yield maps and remote sensing. Little information on return on investments in above stated technologies is available in literature.

This paper evaluates results of usage of a technology for variable application of nitrogen - Hydro N Sensor and compares them to available data on potential benefits of other variable fertilisation technologies. After variable application of N according to Hydro N Sensor the yield of winter cereals was increased compared to yield achieved after uniform application of the same amount of N in exact field trials in Europe (112 sites within 3 years) by 189 kg/ha (2.41%), of which in Germany in 2001 (25 sites) by 270 kg/ha (3.4%), in the Czech Republic in 1999 by 196 kg/ha (3.52%). IRR of investment in this device on a large farm in the Czech Republic oscillates around 100%.

INTRODUCTION
Precision farming (precision agriculture, precision ag, site specific farming, site specific crop management …), in Czech (literary translation) ”přesné zemědělství” arises in 1990s in connection with a possibility to record precisely the position of agricultural machine or sampling spot in a field through technology of satellite navigation (GPS). It is generally related to plant production although the principles of precision farming maybe used more expressively in animal husbandry (individualisation of care of single animals) and started earlier there. Of many definitions of precision farming I would quote the following one: ”Farm management system based on information and technologies that identifies, analyses and manages variability of individual fields to achieve optimal profitability, sustainable development and soil resources protection.”

Precision farming in the most general form can be characterised as a system based on effective utilisation of data describing soil, field, growth or other elements of the environment and their interactions aiming to support and optimise decision making processes at final product production. Principles of precision farming are not yet applicable in the same level of availability, reliability a profitability in individual cultivation operations.

From the information and technological point of view precision farming relies on steady developing methods of data collection and processing.

The ways of data acquisition can be split as follows:
- ground
- soil sampling, growth sampling
- yield monitors
- smart scouting
- remote sensing
- contact sensing
- airborn
- satellite
- on-line
- off-line

The most spread method is perhaps sampling so far (maps of soil content of P, K, Mg, etc., N Tester, or other traditional analytical techniques). The sampling method has, though, its practical and cost constraints and level of its evidence in relation to field variability and yield is a moot point. The methods of yield monitoring and also remote sensing struggle intensively in the present. Smart scouting and contact sensing are rather at the initial stage of their development and commercial use.

DISCUSSION
Split of precision farming technologies according to area of their utilisation:
- soil processing
- crops sowing and seeding
- nutrition and fertilisation of crops including diagnostics
- plant protection
- irrigation
- harvesting of crops

The most spread area of precision farming is fertilisation including diagnostic along with yield monitoring, mainly due to variable application of P, K, Mg and liming according to spatially referenced soil samples. Second main subgroup is variable application of nitrogen either according to soil properties (soil types maps, humus content maps, contact sensor for electromagnetic resonance) or according to characteristics of plant biomass identified by remote sensing. Further more variable application of P, K, Mg according to their off-take by precrop struggles. The nutrients off-take is quantified from yield maps. Variable liming based on contact pH sensors is under development.

Systems of precision farming in the area of field crops fertilisation that are currently commercially available:
- P, K, Mg, Ca fertilisation
- variable or zone fertilisers application according to spatially correlated soil samples (millions of ha/year in the USA)
- variable fertilisers application according to off-take of nutrients by precrop identified from yield maps (cannot be used for liming, in Europe tens of thousands ha per year, in the USA perhaps hundreds of thousands ha/year)
- N fertilisation
- variable application according to Hydro N Sensor - remote sensing (hundreds of thousands ha/year in Europe, very little outside Europe), moreover this technology produces maps chlorophyll content and biomass density
- variable N application according to soil types maps or humus content maps (perhaps hundreds of thousands ha per year in the USA)
- variable N application according to electromagnetic resonance - contact on-line sensor - beginning of commercial utilisation

Economic evaluation of above stated systems of variable application of nutrients is difficult because of lack o background information for calculation and in case of some systems due to expected higher effect after more years of usage than they were used so far.

Background data are available for economic effectiveness evaluation of variable nitrogen application technology based on remote on-line sensing of growth colour - Hydro N Sensor. The benefits of this technology are yield increase, production quality increase, decrease of growth lodging, higher yield uniformity, possibility to decrease applied rates of plant growth regulators and other.

The usage of Hydro N Sensor brought yield increase in winter cereals in comparison to uniform application of the same rate of N in exact field trials (112 sites) in Europe in the years 1999 to 2001 on average by 189 kg/ha (2.41%), of which in Germany in 2001 (25 sites) by 270 kg/ha (3.4%), in the Czech Republic in 1999 (13 sites) by 196 kg/ha (3.52%). When compared to fertilisation according to general farm practice in the Czech Republic in the years 2000 and 2001 the yield of winter wheat was increased by 260 or 460 kg/ha respectively. If only the effect of Hydro N Sensor on winter cereals yield increase is considered and compared with the purchase price of the device and operational costs, the value of internal rate of return (IRR) of this investment for a farm cultivating 1000 ha of winter cereals in a period of 5 years lies between 85 and 220% with current grain prices. That means that the investment is repaid within 0.5 to 1 year.

Of the other available information on economic benefits of variable fertilisation I will use for instance the data of German company Agricon that suggests that after variable application of P and K on the basis of yield maps and soil samples profit increase of 12 EUR/ha was achieved in two years average, while J. Schmerler suggests that variable application of fertilisers and variable rates sowing on a farm in Germany (Brandemburg) on acreage of 3900 ha in the years 1995 to 1998 after having spent DM 49 per ha on average, revenues increase (or operation cost decrease) achieved 60 DM/ha in winter wheat, 52 DM/ha in spring barley, 103 DM/ha in grain maize, 31 DM/ha in silage maize, 42 DM/ha in sunflower and 39 DM/ha in beans.

Numerous available information on the benefits of precision farming, namely P and K fertilisation are presentations of companies running these technologies and are not based on exact field trials but on theoretical constructions which scientific background is unclear. Often not the influence on yield increase is emphasised (since it is generally known that that this influence is, at least several first years of P and K fertilisation, low), but the possibility of fertilisers savings is stressed. Majority of these calculations, namely the savings of P and K fertilisers on the basis of variable application according to soil samples in comparison with uniform application is in principle not correct, though.

CONCLUSIONS
By evaluation of available information on technology for variable application of nitrogen - Hydro N Sensor through IRR method it was concluded that under conditions of large Central European farms with winter cereals acreage around 1000 ha high IRR between 85 and 220 % is achieved, that means that the repayment period of the investment is in very favourable range 0.5 to 1 year. This result is quite exceptional. When economic return on investments in other variable fertilisation technologies was examined, it was usually concluded that there is not enough data available to prove economic reason of them.

Industry and research so far do not concentrate much on economic aspects of precision farming technologies. The influence of the fact that precision ag is a fashionable issue cannot be fully excluded and, rather than clearly defined benefit, the struggle to be on the cutting edge of the technological development is the driving force. One cannot either exclude approach like: farms in EU are rich, why should not they buy the systems. The wide spread of variable application of P,K in the USA is the result of well developed marketing strategy of suppliers and distributors of these systems, the evidence of the economic benefits is a subject of discussions and estimations as it is with all elements entering biological systems. In commercial practice it is possible to find approach where only the mapping of variability of some characteristic is the result. The question then is what is the benefit of the maps that cannot be used for variable application of inputs, records of variable inputs, limiting of negative influence on the environment etc.

Generally it can be suggested that although we have the technological ability to apply the inputs variably, due to difficulties to determine economically optimal rates of inputs, the benefits from majority of precision farming technologies will be limited until there are not carried out trials specifying how the economic optimum of inputs dosage changes with field variability. Hence the more attention should be paid to real and proved benefits of such a technology when its purchase is considered.

When company decides on orientation towards precision farming it is necessary to consider as well more general coherence including expected development in data utilisation in farm production and its administration. Already now it is obvious that relevant state institutions in the Czech Republic prepare a platform which will be based on GIS. Ortophotomaps being prepared by state institutions should become a fundament of this system into which farmers, their suppliers or other specialised firms will insert and record key data on fields and growths. These data will create knowledge database needed for decision optimisation, but they will serve also state institutions administrating farm subsidies as reference and checking source of data. From this shot it is obvious that those farms and contractors that will start to learn precision farming and therefore the data management will be ahead of their competitors. This jump will be based on their experience with functioning of integrated GIS platform, by possibility to influence and actively participate in amendments of this platform and namely by continuous acquisition of relevant data on their growths and fields in standardised format, in case of contractors on their customers. The access to this data, possibility and ability to utilise them when selling own products and services will become then an important competitive advantage.

REFERENCES
Bullock, D.S. (1999) The economics of precision farming: A primer for agronomists designing experiments In Precision Agriculture '99: Papers presented on the 2nd European Conference on Precision Farming, J.V. Stafford (ed.), Sheffield (UK), pp. 937-946

Johansen, C. (1996) Precision Agriculture Buyers Guide

Mc Bratney, A.B., Whelan, B.M. (1999) The Null Hypothesis' of precision agriculture In Precision Agriculture '99: Papers presented on the 2nd European Conference on Precision Farming, J.V. Stafford (ed.), Sheffield (UK), pp. 947-957

Milata, P., Fiman, P. (2001) Precisní zemědělství - mýtus nebo realita?, www.leadingfarmers.cz

More, S. H., Wolcott, M.C. (2001) Mapping and interpreting electrical conductivity in production fields, www.agctr.lsu.edu

presentation (2001) of Agri Con GmbH, Im Wiesengrund 4, Jahna, Germany

presentations (1998 - 2002) of Hydro Agri GmbH & Co. KG, Hanninghof 35, Duelmen, Germany

Schmeler, J. (1999) Cost/benefit analysis of introducing site-specific management on a commercial farm In Precision Agriculture '99: Papers presented on the 2nd European Conference on Precision Farming, J.V. Stafford (ed.), Sheffield (UK), pp. 959-967

www.precision.com.au/Press_Releases/Rapt_with_VRT.htm (2001)