Slug forecasting in cereals

Abstract

In a series of field experiments, slug damage to winter wheat seeds was found to be directly related to the biomass of slugs in the soil (determined by soil sampling) and inversely related to the percentage of fine soil aggregates in the seed-bed and to depth of sowing. Sowing at 40-50 mm depth compared to 20-25 mm provided protection equivalent to a broadcast application of molluscicide pellets and in one year resulted in a 9% yield increase. Best protection from slug damage was achieved by drilling at 40-50 mm depth combined with an application of molluscicide pellets broadcast immediately after drilling. A broadcast application of pellets was more effective than pellets drilled with the seeds.

The percentage of winter wheat seeds killed by slugs was monitored, together with various risk factors, at 93 sites throughout the UK. Estimates of slug populations at each site were obtained from traps baited with methiocarb pellets, and it was found that the peak number of slugs trapped during the period from July until just before cultivation was more important than any other risk factor or combination of factors in accounting for differences in seed damage between sites. This factor accounted for 26% of the variability in seed damage. A combined function of the number of slugs trapped at drilling and the percentage of fine soil aggregates in the seed-bed accounted for 21% of the variability in seed damage. Both (i) the peak number of slugs trapped before drilling, and (ii) the combined function of number of slugs trapped at drilling and the percentage of fine soil aggregates in the seed-bed, have potential for identifying fields with a negligible risk of slug damage to wheat seeds. However, further research is needed to improve the accuracy of forecasting slug damage. In particular, other forms of trap should be tested for their potential use in providing more accurate estimates of slug numbers and biomass at drilling, and techniques of assessing seedbed conditions should be improved.