Variation within fields of potentially available soil nitrogen using the hot KCl technique

Abstract

In February 1996, ten fields in the Lothians were chosen which had been ploughed but not yet sown with spring barley. They were sampled on a 30 metre grid and potentially available nitrogen (PAN) was measured by extraction with hot potassium chloride solution.

Differences between fields accounted for 72% of the sums of squares for differences in PAN and only 28% of the sums of squares was contributed by differences within fields. The field means ranged from 52 to 101 kg N/ha and the standard deviation of samples within the same field from 4.9 to 17.1 kg/ha. Some fields had well defined spatial trends but for other fields, the measured values appeared to be randomly distributed. Contour maps were drawn for all fields but may be of limited value for the fields without pronounced trends.

Simulated sampling within each field showed that the classic W-shaped sampling route with 8 or 16 sampling points per field gave values within 2 kg/ha of the field mean and was greatly superior to either of the two diagonal simulated sampling routes. Simulated effects of spatially variable application of nitrogen fertiliser showed that this was unlikely to give sufficient benefits to justify the cost of multiple samples, even in the most variable of these fields. Uniform application to whole fields as indicated by a composite field sample and compared with a single nitrogen rate for all low N status spring barley fields, could however give benefits to the farmer up to £32/ha which would offset the cost of sampling and analysis. In addition, there might be a reduction in nitrate leaching of up to 9%.