Improvement of nutrient management through effective use of precision agriculture technologies in the southern Australian grains industry.

Author: Dr Allan Mayfield

In this GRDC project SPAA has promoted and supported the uptake of PA technologies to improve the economic returns to growers. This has been by providing independent information on PA at expos, conferences and training workshops and in SPAA Magazines and Newsletters for growers, advisers and the commercial PA industry. It has also been through grower cooperators doing on-farm PA trials at sites linked to grower groups in south eastern Australia. Results of research trials in a range of cropping environments across SA over several years have demonstrated the value of zoning to assist with the identification of yield limiting factors and to manage these limitations for a better economic return. In this project the research paddocks were zoned according to variations in grain yield, EM conductivity and elevation. This zoning was then used to target soil sampling sites and for the location of on-farm trials to assess treatments to improve economic returns. Key factors reducing economic returns included insufficient N or P fertiliser in some zones and excessive N or P fertiliser in other zones. In the latter cases reducing the rate of fertiliser did not reduce grain yield. A whole farm economic analysis of the value of PA on the six cooperating farms showed an average annual return of $18/ha each year from the use of guidance and variable rate inputs. Achievement of this type of economic return, together with on-going support to growers, advisers and the industry will continue to increase the uptake of PA further. Every additional 5% of the cropped area in SA managed with these technologies is expected to increase returns to growers by an estimated $3.6m/year. Output 1. To increase the effective use of PA: Since the start of this project the use of guidance and autosteer systems has increased greatly from less than 10% in 2003 to an estimated 40% now (according to major PA companies). This uptake is likely to continue to increase even further. SPAA's extension role has contributed to this increase through providing independent information at field days, expos, grower meetings and in newsletters and magazines. In contrast, the low use of variable rate management within paddocks is estimated by SPAA and PA companies at only 2%. This is due to the technical complexity of compiling data and determining causes of variability as well as some uncertainty of the economic value of variable rate management. This project has generated information and increased PA skills of growers and advisers to help overcome this technical complexity and uncertainty. On-farm trials provided an understanding of the underlying reasons for yield variability and responses to treatments in different zones. They were also used in an assessment of the economic value of precision agriculture on these farms. Output 2. Results of large scale field experiments to assess the economic value of variable rate inputs. Six field sites covering a wide range of cropping environments across south eastern Australia were established to assess the economic value of variable rate treatments. Treatments were chosen were to address major limiting economic factors in cropping profitability for those paddocks, based on preliminary data and previous experience. The sites and general treatments were 1. Urania, Lower Yorke Peninsula - nitrogen nutrition in wheat & barley 2. Yeelanna, Lower Eyre Peninsula - nitrogen nutrition in wheat & barley 3. Buckleboo, Upper Eyre Peninsula - subsoil ripping & fertiliser placement 4. Waikerie, Murray Mallee - cereal types and phosphorus replacement 5. Rupanyup, Wimmera - phosphorus replacement and using Yield Prophet according to zones 6. Donald, Wimmera - ryegrass control All sites had strong links with local grower groups. These were, respectively, Yorke Peninsula Alkaline Soils Group, Edillilie Landcare Group and Yeelanna Agricultural Bureau, Buckleboo Farm Improvement Group, Mallee Sustainable Farming Group and BCG and Wimmera Framing Systems (for both Victorian sites). Treatment details and results are in the Additional Pages. EM mapping helped to define zones with different responses to nitogen fertiliser in cereals at Sandilands and Yeelanna but responses in the different EM zones varied between seasons and cereal types. At Urania, responses to N fertiliser in wheat were greatest in the lower EM (lighter) soil whereas in another year responses in barley were greatest in the higher EM ( heavier and saltier) soil type. At Yeelanna, barley grown in the soil with the higher EM value and greater available soil N was less responsive to applied N fertiliser than where grown in the lower EM soil with less available soil N due to lodging. EM mapping was also used to define areas responsive to deep ripping. In a trial at Buckleboo wheat yields increased where the soil was deep ripped in the lower EM (lighter soil) areas but not in the higher EM (heavier soil) areas. Deep ripping is an expensive operation and therefore this finding will assist growers to identify areas where economic responses to deep ripping are likely to be worthwhile. There was no grain yield response to the deep placement of fertiliser in either soil type at this site. In another lower rainfall site, Waikerie, phosphorus responses in barley were greatest in the lighter soil type (sand hill) than on the heavier and shallower soil type (flats). A phosphorus replacement strategy, according to the variation in crop yield the previous year, was effective in a wheat crop at Rupanyup in 2005 because of the lack of response to P fertiliser in any of the replacement zones. Control of annual ryegrass in wheat in a trial at Donald was improved by adding other herbicides to trifluralin and/or by increasing the sowing rate of wheat where the ryegrass densities were high but not where they were low. Output 3. Assessment of the N-Sensor for variable rate N application. The mean grain yield of wheat in 10 trials over two years was 25kg/ha (0.7%) higher where post emergent nitrogen fertiliser was spread according to crop reflectance using the N-Sensor compared with where the same total amount of fertiliser was spread at a constant rate. The N-Sensor was useful for mapping variability of biomass across paddocks and so could be used to produce a data layer to be combined with other data layers when producing zone maps. The N-Sensor also satisfactorily mapped patches of tillering ryegrass in a young canola crop but not ryegrass at the heading stage in a podding lentil crop. Other sensors, including the GreenSeeker and the Cropcircle, also produced suitable maps of weed patches in young crops. This technology will provide opportunities for variable rate management of patches of weeds in young crops or before sowing.

(0.1mb)