Root system management in winter wheat: practices to increase water and nitrogen use

We aimed to quantify some of the key root-shoot relationships and establish the extent to which agronomy can impact on, or potentially manage, root systems in limiting environments. We targeted crops with small to medium-sized leaf canopies (3 to 8 units of green area index (GAI)) that we expected to be more vulnerable to the extremes of soil conditions and agronomic conditions selected, especially under shallow-rooting conditions. Controlled environment experimentation was used to interpret the field observations and extend the range of environments.

The key messages from the agronomic treatments in this study are: (1) Adjusting seed rate (and hence plant population density) had a large, and significant, effect on the size of the root system. There was also a potentially large, though less consistent, effect of seed rate on increasing root distribution at depth. (2) Adjusting the nitrogen fertiliser programme within a range typical of current practice did not have any consistent benefits for root growth and yield, though there is limited evidence for sitespecific benefits in root growth and distribution. (3) There were no agronomically important variations in total root size or root distribution in two contrasting varieties. Nevertheless, we feel that some rooting characteristics could be targeted in crop improvement programmes.

Total root length (TRL) varied from 7 to 28 km m-2, whilst root length density (RLD) ranged from 3 to 6 cm cm-3 in the plough layer to generally less than 1 cm cm-3 below 40 cm. For moderate to good yields, the relationship between TRL and ear number appears to be conservative at between 30 to 35 m of TRL per ear. TRL per unit of GAI and TRL per tonne of grain increased with an increase in yield. Reducing seed rate affected ear number per m-2 more than it did TRL or GAI, and TRL per ear increased significantly with a reduction in seed rate. TRL was weakly associated with the total amount of nitrogen available as either residual soil nitrogen or applied fertiliser nitrogen. Nitrogen offtake was more strongly correlated with RLD in the plough layer (0-20 cm soil depth) than with TRL. At low to moderate yielding sites total root size was small (less than 10-15 km m-2) across a wide range of leaf canopy sizes. As the average site yield increased towards moderate to high yields there was evidence for a trade-off between root growth and shoot growth, such that any given yield could be achieved by further increments in either the root system or the leaf canopy.

We developed simple methods of assessing roots in early spring and related results to actual values of TRL and RLD. At mid-tillering there was a good correlation between root counts at the face of a soil pit and the actual RLD. These simple measures were also related to yield and could be developed for growers and agronomists as measures of root and soil condition that would be particularly useful for identifying remedial action for following seasons, rather than changes in the current season.