MC HARE¹, MD BAKER¹ and DW PARRY ²

¹ Harper Adams Agricultural College, Newport, Shropshire TF10 8NB, UK; ²HRI East Malling, West Malling, Kent ME19 6BJ, UK

Background and objectives
The effect of Fusarium ear blight (FEB) on the grain yield of winter wheat is not clear. There are two components of yield that could be affected by ear infection; the number of seeds per ear and the weight of individual seeds. Cassini, in a review of Fusarium diseases of cereals in Western Europe [1], proposed that reduced yield following Microdochium nivale ear infection was caused solely by a reduction in thousand grain weight (TGW). In contrast to M. ;nivale, the effect of FEB caused by Fusarium culmorum on the yield of winter wheat is much clearer. A reduction in yield, associated with a decrease in the weight of individual wheat seeds was demonstrated in field trials in the Netherlands following inoculation of wheat ears with F. ;culmorum conidia. The mean loss of yield of 9.7% was almost exclusively attributable to a reduction in TGW [2]. It is known that infected seed is an important source of inoculum for seedling blight, but the relationship between seed infection, seed weight and disease symptoms is not. The aim of this study was to investigate the relationship between natural M. ;nivale and F. ;culmorum seed infection resulting from FEB, seed weight and subsequent Fusarium seedling blight symptoms.

Materials and methods
Frequency distributions of seeds by weight were produced for three commercially grown seed lots of winter wheat cv. Riband infected by M. ;nivale and three seed lots of winter wheat cv. Avalon which had shown symptoms of FEB when grown in field plots. Seven-hundred seeds were taken at random from each seed lot, individually weighed and placed in one of eight weight categories: 10-20 ;mg, 21-30 ;mg, 31-40 ;mg, 41-50 ;mg, 51-60 ;mg, 61-70 ;mg, 71-80 ;mg and 81-90 ;mg. The number of seeds infected by either M. ;nivale or F. ;culmorum in each weight category was determined by isolating the pathogen from individual seeds. Following treatment in sodium hypochlorite seeds were placed on antibiotic amended potato dextrose agar to recover M. ;nivale or moist filter paper to recover F. ;culmorum. Frequency distributions of M. ;nivale or F. ;culmorum infected seeds by weight were then produced. Seedling emergence was assessed in two controlled environment pot experiments according to randomized block design. In one experiment M. ;nivale infected seed was used and in the other, F. ;culmorum. Seeds from the heaviest and the lightest weight categories from each of the three M. ;nivale infected lots plus a non-infected lot were sown with or without Rappor (50g/l guazatine, DowElanco). In the second experiment two seed lots, one heavily infected by F. ;culmorum and one without infection were used and treated as for the M. ;nivale infected seed. Environmental conditions were set to 8øC for 8 ;h of light and 6øC for 16 ;h of darkness and constant 20øC with 16 ;h of light for experiments with M. ;nivale and F. ;culmorum respectively. Complete seedling emergence was assessed when no further seedlings had emerged on five successive days.

Results and conclusions
For M. ;nivale infected seed the frequency distributions by weight for each seed lot as a whole was not severely skewed and was similar to that for infected seeds. However, this was not the case with F. ;culmorum infected seed. As the percentage of infected seed in each lot increased the distribution by weight for the seed lot as a whole became skewed and infected seeds were frequently lighter than uninfected seeds. Seedling emergence was clearly reduced by M. ;nivale infection in the absence of seed treatment, emergence being significantly lower from heavy seed than light seed. F. ;culmorum also reduced seedling emergence but with fewer seedlings being produced from light seeds. In this instance, chemical control was only observed with large infected seeds. Results showed that the removal of small seed from a lot infected by M. ;nivale may not increase emergence but a seed treatment may provide effective control. With F. ;culmorum, removal of small seeds may be beneficial but chemical control may only be effective on large seeds.

References
1. Cassini R, 1981. Fusarium: Diseases, Biology and Taxonomy. University Park and London: The Pennsylvania State University Press, pp. 58-63.
2. Snijders CHA, Perkowski J, 1990. Phytopathology 80, 566-70.