Technical Officer, Centre for Conservation Farming, Charles Sturt University.
Wagga Wagga. 2678
Summary: New technology is allowing variations in soil nutrients, moisture, yield and other soil factors identified in the paddock to be mapped with a known degree of accuracy. These maps can then be used to treat specific areas within the paddock depending on its detected requirements. Developments in sensor technology, satellite navigation and computer processing power are close to making this method a commercial reality. Remote sensing methods are being utilised in an attempt to bias sampling areas and thus reduce costs in order to make it affordable for producers in Australian grain production regions.
The importance of the information revolution is becoming more apparent to primary producers. Detailed information is now available at the press of a button - weather data and forecasts, stock market reports and analysis, and news services. Many farmers utilise this technology to help them plan their activities. However the real benefit of this new technology is within the farm boundary - information can now be accessed 'virtually coming up from the ground'.
This type of information gathering has been possible for many years but new sensors, computer hardware and software, and now machinery, are available to make the process a commercial reality.
The single most important feature of this new technology is the development of a simple, compact navigation system called global positioning systems (GPS). GPS provides the means to identify a sampling area with a set of coordinates. This allows the operator to return to exactly that area regardless of changed fence lines or other developments. It also means that variation across a paddock or landscape can be recorded, mapped and used at a later date.
Modern sensors are quicker and have simplified the gathering of samples. For example, in the case of conductivity readings for measuring potential salinity of soil, thousands of readings can be taken and mapped almost automatically.
The Principles of Precision Farming
Precision farming, prescription farming, site-specific agriculture, farming by the foot all basically describe the same process: that is, a small area is treated specifically for its needs which are ascertained by some form of objective measurement. Possibly the easiest way to describe the technique is to describe a software package developed by NASA in 1994.
A four-wheel motorbike mounted with a laptop computer and satellite navigation system drives around the paddock being surveyed to establish the paddock's shape, size and location. With the boundaries mapped, a statistical package is run on the paddock and sampling points are generated. The motorbike is then directed to these points in the paddock by a map on the laptop computer and the satellite navigation system. Samples are taken at these points and placed in bar coded bags.
The bags are swiped with a barcode reader at the site where the sample is collected, thereby giving the sample a number and a location based on the satellite coordinates. The samples are processed and analysed for a number of nutrients and soil features, for example, nitrogen, phosphate, potassium, pH and trace elements.
The results are re-entered into a mapping program and, because the samples have coordinates, a fertility map can be drawn up showing highs and lows throughout the paddock of any of the soil features tested.
The map data are entered into a crop model and nutrient demand in the area is calculated. This is then subtracted from nutrient already present in the soil. The map now generated is the rate needed for the various nutrients, or soil amelioration substances such as lime, for best crop growth. This "deficit" map is entered into the laptop computer on a modified fertiliser spreader or combine which will deliver varying rates of fertiliser, depending on the mapped deficit within any area (Grace, 1994). If the spreader or combine runs back over its track the gates automatically close so that no area is double dosed (Cooke, 1993). The heart of the system is the laptop computer and the global positioning system, both of which are becoming faster, cheaper and more reliable on almost a monthly basis.
A system developed by Massey Ferguson in Britain has GPS systems on a series of headers, tractors and combines. Fertility maps are generated by measuring the yield harvested. A grain sample is measured every 1.2 seconds as it runs through a flow meter in the header. This information is mapped and seeding and fertiliser rates are automatically varied in the combine to take advantage of differing soil types or to average crop performance across the paddock (Massey Ferguson Ltd., 1995).
These systems are designed for high value, highly productive land where large amounts of money can be spent to increase yields. Australian agriculture, in contrast, has large areas of land with lower yield potential where comprehensive sampling programs are uneconomical. What we are hoping to utilise is remote sensing to bias the sampling collection to a number of smaller, but representative, areas that can then be extrapolated to more extensive areas, thus saving money.
These remote sensing systems could be satellite images or aerial video images using near infrared filters to give spectral responses of growing crops, pastures and weeds. Soil samples, tissue or sap samples could be taken and related to a spectral response from a remote image. A demand and deficit map can be generated and areas treated specifically.
A further potential development which, if successful, could be revolutionary on a number of fronts, is the use of this method in the application of pesticides (Cooke, 1993). A method of mapping weed or insect infestations within a paddock is being developed using high resolution aerial video and crop tolerances/per pest number generated. Only areas exceeding the tolerance value will be sprayed. The map in the laptop controlling the spray boom will simply shut off a section of boom, or the whole boom when it is not needed, or when the area has already been covered.
Benefits of this Method
Researchers are aiming for sustainable, profitable, environmental friendly use of resources and taking advantage of and protecting all areas of the farm. Another factor is the public push for environmentally friendly practices (Blackmore, 1994).
An example of how such a system can work are the results of a trial carried out in California and noted by Chancellor (1994) to determine the benefits of "spatially modulated" inputs as compared to blanket applications. The trial involved plots of winter wheat which were sampled at one metre intervals along a central line and also at one metre intervals along a line at either end of the plots, the areas being 280 metres long and 60 metres wide. The samples were analysed for moisture, N and weed populations and a demand and defect map calculated. Spatially modulated inputs were applied and compared to blanket applications and results calculated as a function of input over yield. The benefits were 2%, 12% and 40%, respectively, for water, nitrogen and herbicide.
Though this method has not decreased the quantity of inputs when taken over a paddock or whole farm, the inputs were allocated more efficiently and matched to specific site requirements. The savings and benefits can result from a more efficient allocation of inputs rather than necessarily a net reduction of inputs (Chancellor, 1994).
One opportunity this method has opened up for US farmers is to allocate more precisely their "set aside" areas. They generate a gross margin map for the crop and any areas that fall below the government-subsidised "set aside" or "paid not to grow" rates are put to one side so, in effect, they maximise their profit from cropping areas and "set aside" areas (Blackmore, 1994).
Where a number of factors interact to affect yield, such as pH, moisture content and nutrient uptake, GIS software can layer differing variables (Knight, 1994) and process the data so the limiting factor for any grid cell is known. Farm-based GIS programs are making data analysis easier with every new version of software packages.
The era of full environmental accountability is approaching and it need not be the "horseman of the apocalypse" for farming practices. In many cases it is better book keeping and paddock records. With the auditing of growth variables, allowances can be made for environmental protection, minimising nutrient loss through runoff or leaching and also aid in reducing the impact of spray drift by more efficient application.
There are three common scenarios in which gathered information is processed which influence the final audit result:
yield protection, high physical inputs;
reduced inputs, optimal return, low environmental concern;
reduced inputs, high environmental concern (Blackmore, 1994).
The limiting factors in this technology at present are the sensors which quickly and cheaply sample areas of concern (Cooke, 1993). There is a considerable amount of work being undertaken in the USA developing new methods to obtain accurate soil data, especially with Near Infrared (NIR) techniques. With this method it is possible to measure the spectral reflectance of a soil sample from which its nutrient status can be estimated (Cooke, 1993). This technology has huge potential for rapid soil analysis.
Whilst more information is always an attractive possibility, the collection and processing of this information needs to have a justifiable use.
Precision farming is an evolving science and advances in other technologies (GPS, plant/soil analysis and data processing systems) are offering the farmer of today the technology of tomorrow. With continued sensor and computer development the future adoption of precision farming techniques will have positive benefits for agricultural production in Australia.
One of my visions is a totally automated system that is incorporated within agricultural machinery. This self- contained unit with GPS navigation would automatically take samples, log its position, analyse and store the samples for further analysis and would automatically download to a computer via digital cellular phone. Human labour would be minimal, processing would be automatic and running costs would be low. Most of this technology already exists and this system could be only two or three years away.
These methods outlined are very exciting developments in modern agriculture and it is my feeling that, as this technology evolves, a greater level of efficient and safe land use will result which can only be of benefit to all.
1. Blackmore, S.(1994). Precision Farming: An Introduction. Outlook on Agriculture 23(4) 4, 275-280.
2. Chancellor, W.J. and Goronea, M.A. (1994). Effects of Spatial Variability of Nitrogen, Moisture and Weeds on the Advantages of Site-Specific Applications for Wheat. Transactions of the ASAE 37(3): 717-724.
3. Cooke, L. (1993). Fine-Tuning Agricultural Inputs. Agricultural Research.
4. Grace, J.A., Pearson, R.S. and Gilmore, K.T. (1994). The Applications of GIS Technologies to Prescription Farming. Unpublished. Space Remote Sensing Centre. Stennis Space Center, Mississippi.
5. Knight, J.D. and Mumford, J.D. (1994). Decision Support Systems in Crop Protection. Outlook on Agriculture 23(4), 281-285.
6. Massey Ferguson Group Limited (1995). Datavision Yield mapping System.