Future Farming Systems - Light, camera, action

By Kay Ansell

Meeting growers" increasing need for accurate, realtime information about the land they are cropping is the first goal of a pioneering, hightechnology project called "Future Farming Systems".

Instead of waiting for results from crop samples sent to a laboratory, the aim is for growers to tap the power of remote sensing to identify and diagnose problem areas and take remedial action - fast.

Nitrogen management - how much to apply, when and where - is an early priority of the project, which will harness information from reflected light (through hyperspectral canopy reflectance) and heat (using canopy thermal imaging) from crops. This data will be integrated with seasonal climate forecasting, paddock zoning and cropping systems models.

Dr Daniel Rodriguez, of Victoria"s Department of Primary Industries in Horsham, says the aim of Future Farming Systems is to combine all this - plus other GRDC-funded information - to improve the efficiency and economic benefits of fertiliser use by taking a systems analysis approach. Primary Industries Research Victoria (PIRVic) is sponsoring the four-year project. Dr Rodriguez"s co-researchers are Dr Peter Fisher, Paul Rampant and Dr Bob Belford.

Dr Rodriguez says developments in precision agriculture, and variablerate and site-specific technologies, are driving the need for rapid, inexpensive methods for identifying and measuring the variability over time and space of a crop"s condition.

The benefits are expected to be environmental, as well as economic: "If you have nitrogen sitting in the soil not being used because the crop is suffering from drought stress, for example, you have the potential for losing money on the one hand, or to have it become a greenhouse gas emission and that is something we want to prevent," he says. "These technologies have huge applications in many areas, but initially crop nitrogen management and spatial grain-quality forecasting will be our key areas of work."

Other potential applications include weed, pest and disease monitoring and control.

Paddock zoning based on soil prospecting can identify so-called "flip flop" areas that are high yielding in some years and low yielding in others, he says. "But to manage these highly variable areas you need to consider the seasonal conditions.

"That is where the seasonal forecasting and the in-crop sensing comes in. They help you decide whether, when and where to apply fertiliser."

These latter aspects would draw on data from two existing GRDC-funded projects: "Tools to reduce the impact of climate variability in south-eastern Australia" and "Strategic Initiative on Precision Agriculture".

Dr Rodriguez recently visited the US Department of Agriculture"s water conservation laboratory in Tempe, Arizona, where he assessed the technology being bought for the project, an investment of around $160,000.

At the Arizona laboratory, Dr Rodriguez consulted research biologist Dr Paul Pinter, Dr Glenn Fitzgerald and scientist Tom Clarke, who are world leaders in the use of remote sensing for crop management.

The project will use three cameras that will capture overhead images. The thermal imaging system will spatially register canopy temperatures to reveal water stress. Another camera will measure the light spectra reflected off the crop"s canopy in particular ranges, while the third will measure hyperspectra canopy reflectance, that is, the full light spectrum.

The light spectra cameras will help reveal the crop"s nitrogen nutrition, among other factors. "Different components in the leaves absorb and reflect light of different wavelengths.

"Changes in the biochemistry of the plant canopy [leaves] can be picked up by the light the canopy is reflecting. It is a very economic and quick method to identify change in the biochemical composition of the plants."

Dr Rodriguez says that more precise and current information about the physiological status of crops that are experiencing, for example, water and nitrogen stresses, would help growers take corrective action and also forecast grain quality.

The latter would give growers the option of harvesting and segregating the premium quality grain in their paddocks to achieve higher returns.

The first experiments have already been designed and will be underway by July this year.

"I expect to have good results after the first two years. We are starting here in controlled conditions in the DPI Plant Breeding Centre and next year we are going into farmers" paddocks and focusing on the high rainfall zone because we see that area as having the most potential for the technology."

Crop canopy sensing techniques can be used to develop maps of water stress, nitrogen status and canopy density when variations in all of these factors are simultaneously present in the landscape.
Pictured here are NIR (near infra red) images of a nitrogen (N=high, n=low) by water (W=irrigated, w=rainfed) experiment. The top image indicates different levels of biomass, while the bottom image shows the sensed nitrogen stress factor displayed as red, the sensed canopy density index displayed as green, and the sensed crop water stress index displayed as blue for the same experiment.

Source: Barnes et al., 2000

For more information:
Dr Daniel Rodriguez, 03 5362 2323, daniel.rodriguez@dpi.vic.gov.au

Illustration source: Coincident detection of crop water stress, nitrogen status and canopy density using ground-based multispectral data; E.M.Barnes, et al 2001. Proc. 5th Int. Conf. Precision Agriculture, Bloomington, MN, USA, July, 2000.