With an increased focus on lower input farming due to rising costs and concerns about sustainability, there is growing interest in the potential of sowing mixed varieties to generate higher and more consistent yields than the equivalent single-variety sowings.  Data from 14 mixture trials in winter barley, spring barley and winter wheat between 2005 and 2016 in which crops were subject to two levels each of fungicide and fertiliser – standard practice and reduced, were analysed.  Overall, barley showed greater benefit from sowing mixtures than wheat; the behaviour of the winter crops mixtures can be categorised into two groups.  The first group showed greatest benefit when only one need was fully satisfied: either fungicide or fertiliser.  The second group showed greatest benefit when either all needs were met or none, but not when only one was satisfied.  The latter finding has informed current trials to elucidate mechanisms driving differences, which will help in the selection of mixtures to perform well with reduced input cultivation.

Image

Modern cereal varieties tend to be selected for under conditions of high nitrogen and fungicide input.  It has been found that in some situations complementary mixtures in barley can lead to higher and more stable yields than the equivalent single variety sowings.  Understanding the mechanisms by and conditions under which greatest benefit is gained would help select candidates for inclusion in better mixtures, productive under lower input cultivation. 

Data from 14 field trials for winter wheat, winter barley and spring barley were analysed.  The trials all had the same scale and density of sowing and were carried out in similar locations using the same machinery.  Each trial had the same two treatments: fungicide and fertiliser, each with two levels: conventional and reduced.  Mixed effect models incorporating spatial correlation were used to obtain a prediction for each variety and mixture under each treatment combination.  These predictions were used to calculate, for each treatment combination in each trial, the difference between the yield from the mixture and the mean yield from the contributing varieties.  These mixture effects were then analysed using mixed models to estimate the mean mixture effect under each treatment combination. 

Previous work has hinted at the possibility that cereal mixtures fall broadly into two categories.  The first category shows greatest benefit when only one need is fully satisfied: either fungicide or fertiliser.  The second category shows greatest benefit when either all needs are met or none, but not when only one was satisfied.  One explanation might be that, in the first category, each variety in the mixture is either highly dependent on fertiliser but resistant to pathogens, or has low fertiliser needs but is susceptible to pathogens whilst in the second category the varieties may have less extreme fertiliser needs and susceptibility.  A Henderson Dip statistic was used to calculate the strength of evidence for these categories, assessing whether they might just have happened by chance.

The analysis found mixtures to be beneficial in spring and winter barley, but less so in winter wheat.  Data from winter barley trials supported the hypothesis that there are two categories of mixtures and to a lesser extent so did winter wheat data.  The stronger barley effects are consistent with the greater variation in morphology seen between barley varieties. 

These findings are informing current trials of barley mixtures under reduced input cultivation by elucidating the mechanisms driving mixture benefits and thus helping in the selection of varietal combinations for low input agriculture.  

This work was done in collaboration with Adrian Newton at The James Hutton Institute and was funded under the Scottish Government's Strategic Research Programme for environment, agriculture and food.