Modern European wheat cultivars are usually selected for high yield under intensively managed agricultural systems. An aim of this study was to test the hypothesis that older cvs or landraces may harbour beneficial alleles not present in elite germplasm that would give them an advantage under non-inversion tillage or reduced N input conditions. There are few data to suggest specific characters (e.g. root system traits) that would define ideotypes for high performance in low input systems.

We tested 10 winter wheat accessions, a subset of the WHEALBI panel, which included old cultivars, landraces and modern elite lines. Two field experiments at Cambridge, UK (2016/16 and 2017/17) in a split-split-plot design tested the effects of tillage (plough, deep non-inversion [DNI], shallow non-inversion [SNI], N level ("40 vs. "180 kg applied N/ha) and genotype. To assess the distribution of root mass in a subset of treatments, soil cores were collected, cut into 25 cm sections, bulked, weighed and washed to obtain roots, which were dried and weighed. In 2016 there were significant differences in grain yield between genotypes (G), N levels and tillage practices. Elite cvs had a greater percentage yield increase to N inputs than landraces, which had smaller NUE. Similar results were obtained in 2017. There were significant differences between genotypes grouped by Class in root mass at different depths through the soil profile in 2016 and 2017. In 2016, there were no significant differences between tillage practices, although there was a trend for higher root mass under SNI. In 2017 elite cvs had greater root mass than landraces, but there was no significant effect of tillage on root mass distribution. There were more roots with SNI compared to plough in upper layers, and root mass at depth positively correlated with NUE and grain yield.

In summary, landrace accessions yielded less than the commercial cultivars, regardless of fertilisation level and tillage practice. The lack of consistent genotype x N level interactions indicates that landraces or older cvs were not more efficient in N use than modern elite cvs, or performed better under one tillage treatment vs. another. However, this study examined a very limited set of lines. Further studies to explore genotypic diversity for root growth dynamics under different tillage approaches are needed to investigate whether variety choice is also an important management consideration when choosing the most appropriate tillage practice for a given farming system.