Crop nutritional management affects soil structure and soil functions at different scales.

Soil organic matter is associated with important biological and physical functions. There are many theories to interpret this association, as yet there is not a fully developed understanding linking soil properties to nutritional management in arable systems.

We used X-ray computed tomography to analyse soil structure at the core and aggregate scale on the Broadbalk long term experiment (Hertfordshire, England). Here we present results of the treatments that have been under continuous wheat for 175 years. Corresponding to treatments that the only difference between the treatments is the nutrient management regime, with the exception of the baseline, or ‘wilderness’ treatment in which the plot was left unmanaged and has returned to mature woodland since 1882. The other nutrient treatments correspond to inorganic fertiliser addition with and without phosphorus, farmyard manure, and no added nutrient.

At core scale (40 µm resolution) we capture macro pore structures that are responsible for convective flow, while the aggregate scale images (1.5 µm resolution) include structures responsible for retention of water by capillary forces.  Therefore, a comparison of images taken at the two resolutions 1.5 µm and 40 µm provides information on how soil partitions between drainage and storage of water, and therefore on the air water balance under different environmental contexts.

The results are presented as a state-space plot of simulated permeability vs. porosity for each treatment. We find that nutrient management resulted in two distinct states at aggregate scale corresponding to water storage potential. Inorganic nutrient management resulted in structures of lower porosity and lower simulated permeability. There was no significant difference between each treatment, or between these treatments and the treatment with no nutrient addition. By comparison, the wilderness and manure treatments had higher porosity and higher permeability, with no significant difference between them.

At core scale, the results are slightly different. Again, the inorganic nutrient management treatments had lower porosity and simulated permeability, with no significant difference between them, and between them and the treatment with no nutrient addition. However, the manure treatment had a significantly lower porosity and permeability than the wilderness treatment. We conclude that long-term cultivation with organic nutrient management results in a similar capacity for water storage and transport to roots than a wilderness control, but that long-term management using a purely inorganic nutrient regime results in a smaller capacity for water storage and a lower transport rate to roots. Organic inputs, roots and plant detritus ploughed into the soil after harvest had no significant impact. Infiltration potential is highest in the wilderness control, lower for the manure treatment, and lowest for the inorganic nutrient management treatment. Again, inputs of organic nutrients from plants had no significant impact. We interpret these findings in terms of a previously hypothesised self-organising feedback loop between microbial activity and soil structure.