The effect of location and management on the target drilling rate for winter wheat

Abstract

Previous HGCA-funded project work showed that optimum winter wheat plant populations for yield can be as low as 70 plants/m². Many farmers have reduced seed rates as a result of this work. However, these earlier trials were carried out at two sites only and did not consider other agronomic factors. A single statistical method was used to analyse the results.

The current trials were carried out at six sites over three years. Additional agronomic factors (rotational position, slug control, plant growth regulator use, nitrogen nutrition or foliar disease control) were included to test for specific interactions with seed rate at each site. 

Results were analysed using two methods:

• as in the previous project, a conventional exponential-plus-linear model was used to relate plant population to yield,
• in collaboration with BioSS, Bayesian methodology using both exponential-plus-linear and inverse-quadratic models were used to relate seed rate to yield.

Both models provided equally good fits to the data in the Bayesian analysis but the inverse-quadratic consistently predicted optima up to 30% higher than did the exponential-plus-linear method. Optimum plant populations were similar, or slightly higher, than in the previous project, ranging from 70 plants/m2 in September drillings in southern England, to 250 plants/m2 in Scotland.

Other agronomic effects:

1. Overall, the conventional analysis did not predict a consistent increase in the optimum plant population with latitude, although relatively high optimum plant populations were calculated more frequently at the northern sites. The Bayesian analysis predicts that optimum seed rate must increase by 7 or 8% (or 11 and 28 seeds/m2) per degree of latitude.

2. Low soil mineral nitrogen reserves or delayed spring nitrogen application reduced tillering at most sites, but this affected optimum plant population in only one of the seven experiments. 

3. Slugs ate approximately the same proportion of plants at all seed rates. Optimum seed rates were higher if slugs were not controlled, but optimum plant populations were unaffected because, in this study, slugs only reduced the percentage establishment and did not affect the compensatory ability of plants.
 
4. Optimum plant populations were 30-40 plants lower in rotational positions where take-all infection occurred than in first wheats. Where Latitude seed treatment was used the optimum was 20-25 plants lower than first wheats. Using Latitude seed treatment yielded 0.23-2.59 t/ha more at the optimum plant population than untreated seed.

5. Yield response to foliar disease control (only tested at one site) was significantly reduced at low seed rates, suggesting disease control savings may be possible in sparser crops.

6. No effects of PGR use were found on tillering or compensation for low plant populations; their use should be targeted at high plant population crops where lodging risk is high.