Youve mapped your yields. Now what? Tia Rund examines the findings of a recent precision farming conference for signs of an on-farm breakthrough.
SCIENCE and its technical spinoffs are still well ahead of understanding how farmers can adjust their crop inputs variably for optimum results – but the gap is narrowing.
Across the world, there is impending agreement that soil quality, particularly the ability to store and provide water to growing plants, is the biggest single influence on yield variation across a field.
Positioning systems already available to the farmer or under development are accurate, reliable and affordable, and there are further developments pending with remote sensing systems which will measure field organic matter by satellite, precision application machinery, and computerised control systems.
But Dr John Stafford, of Silsoe Research Institute, and scientific co-ordinator of the First European Conference on Precision Farming, admitted more development is needed to put all this crop and soil science together with agronomy to work on an individual farm or field.
Research from both the USA and UK reported at the Warwick University conference highlighted the role of soil type and quality. Scientists from the US Department of Agriculture and the University of Missouri-Columbia said the soil quality which caused the most significant variation is its ability to store and provide water, itself a combination of many other measurable factors.
Their trials show that using an electromagnetic induction (EM) sensor trailed behind an ATV can map this ability. Links between yield and individual soil measurements such as pH and organic matter are low, even negative in some cases, argued the Americans. But the EM readings at which maximum yield occurred remained relatively constant for each field over different years and crops, and for differing soil types.
EM sensing could be used to explain and predict product variability with much more detail than traditional soil survey maps, they said.
However, soil scientist Eunice Lord and ADAS colleagues believed not enough is yet known about the effect of soil type on the timing and severity of restrictions on water uptake, and the implications of this for the crops response to other inputs. Is crop water use the cause or the effect of yield variation, they asked.
Studies on contrasting soils at ADAS sites at Boxworth, Cambridgeshire, and Gleadthorpe, Nottinghamshire, showed increased water use correlated with, but was not always the cause of, increased yields.
On two clay soils, yield was positively correlated with soil moisture deficit (SMD) in July. But both yield and SMD were more strongly correlated with early crop growth, and with take-all scores, than with any measured soil properties.
Neutron probe measurements showed equal soil water content across the sites in spring, but moisture abstraction was smaller, and reached a shallower depth, in one field in both years of the experiment. Poor soil structure, take-all disease or both may have contributed to reduced growth, restricting demand for water.
In contrast, on the sands at Gleadthorpe, in a drought year, variation in soil water supply appeared to be the direct cause of variation in yield.
Ms Lord concluded that, before basing input decisions on variable yields, its important to identify whats truly limiting yield, and how far the limitations can be overcome.
WORK headed by Eric Evans at Newcastle University has focussed on optimising lime applications to take within-field variation into account. Across the two study fields in Northumberland, there were substantial areas with uniform pH. The fact that these crossed field boundaries shows current arable practice isnt reducing the pH variation.
Spatially-targeted lime applications are calculated to be more effective than a uniform dressing based on an average pH for the whole field. Using a flat rate, half the area would either not reach the target minimum pH 6.5 or would exceed pH 7.0.
Variations in soil texture and organic matter can be ignored when estimating lime requirement, unless they vary a lot across the field, added Dr Evans. If large numbers of samples are to be taken to estimate the extent of within-field variation, then the cheaper and quicker the soil analysis, the better. Since texture and organic matter content change relatively slowly over time, after an initial assessment lime requirements can be derived from pH measurements alone.
USE like-for-like to get the most from crop yield maps and dont try to associate yield variation in say, oilseed rape one year, with the differences in wheat the next.
Researchers at the Royal Veterinary and Agricultural University in Denmark have reviewed a total of 82 fields at nine locations – eight in England and one in Denmark – to see how useful yield maps are for predicting yield variation.
They concluded that, for each field, its better to look back in time to find an identical crop. For example, for predicting yield variation for wheat, its better to base that on when it last featured in the rotation, rather than looking at the previous years break crop results.