Detecting soil nitrogen supplies by canopy sensing


Cereals & Oilseeds
Project code:
01 December 2007 - 30 November 2008
AHDB Cereals & Oilseeds.
AHDB sector cost:
£17,980 from HGCA (Project RD-2008-3563)
Project leader:
R. Sylvester-Bradley1 , J.J.J. Wiltshire1 , D.R. Kindred1 , D.L.J. Hatley2 , & S. Clarke3 . 1 ADAS UK Ltd, Boxworth Cambridge CB3 8NN 2 ADAS UK Ltd, Bentinck Farm, Rhoon Road, Terrington St Clement, King's Lynn, Norfolk PE34 4PW 3 ADAS UK Ltd, Gleadthorpe, Meden Vale, Mansfield, Nottingham NG20 9PF



About this project


This report describes a second year of research that tested whether soil nitrogen supplies to cereal crops can be detected using canopy sensors; the first season was reported in HGCA Project Report No. 427.

Nitrogen fertiliser experiments on cereals were established at four sites in 2006-7. In the following year, commercial cereal crops (wheat, oats or barley) were grown and, at each site, plot positions as used in the previous year were marked out for testing with a reflectance sensor.

Reflectance was measured four times during tillering, between December and May (dependant on site), using a Crop Circle instrument (provided by Soilessentials Ltd) which measured reflectance at 880 nm (near-infrared, NIR) and 590 nm (orange). A Normalised Difference Vegetation Index (NDVI) was calculated to give a measure of vegetation cover. Soil mineral N (SMN) data were obtained for the sites. No fertiliser N was applied in 2008 and total N uptake at harvest was taken to represent the 'soil N supply' (SNS). Data were interpreted for relationships between canopy reflectance and soil N.

The best level of tillering and ground cover was achieved at High Mowthorpe, which was sown early. Boxworth and Terrington crops were smaller, and the crop at Rosemaund was very small. High levels of N applied in 2007 had large effects on SMN at Boxworth and Terrington, but maximum amounts were smaller at High Mowthorpe or Rosemaund and maximum SMN and SNS levels were small.

Use of the sensor successfully detected the differences in SMN residues at Boxworth and Terrington, especially below 120-140 kg/ha, as was found in the previous year's experiments. The relationships improved with later assessment of NDVI. Change in NDVI between assessment dates showed that canopies always grew during the 2007/08 winter but change in NDVI was less useful for predicting SNS than absolute values of NDVI.

Merged data from both seasons of the study showed that NDVI signals overwinter could be interpreted according to their differences from the theoretical NDVI of an unlimited crop. It was concluded that young canopies can signal soil N status where SMN is less than 120-140 kg/ha. Effects were more certain as crops grew, so canopy sensing for soil N supplies should prove more useful as the season progresses.