NIAB, Huntingdon Road, Cambridge CB3 OLE, UK

Background and objectives
Seed health testing can be a crucial although frequently neglected factor in the management of seedborne disease. The main reasons for this are firstly, that the basic technologies employed in most seed health testing have changed little over the years and sometimes suffer from difficulties with accuracy, sensitivity and speed, and secondly, that there can be uncertainty, under some circumstances, as to how the data from seed health testing are to be used effectively for management of disease.

Testing methods have developed in order to overcome some of their intrinsic disadvantages, leading to more reliable and cheaper seed health data, and there has been considerable but not exclusive interest in the contribution that DNA and immunological techniques can make to this process [1]. Over the previous 10 years these two diagnostic approaches have developed in parallel and often competitively; DNA through nucleic acid hybridization to PCR, and serology through polyclonal to monoclonal and recombinant antibody technologies. Much exciting progress has been made.

Results and conclusions
Relatively less progress has been made on issues related to the use of seed health data, irrespective of their reliability and the methods employed to acquire them. To a certain extent the interpretation of seed health data depends on the particular circumstance. For example, for quarantine purposes it is essential to know the factors which affect the economic risk posed by the introduction of a seedborne pathogen, more than the meaning of a particular inoculum level in a seed lot. Usually a null standard is de rigueur. However it is necessary, when attempting seed treatment management, to know the level of infection in a seed lot and its agronomic implications. Thus different imperatives will drive the choice of seed health test method and the supplementary information needed for proper contextual interpretation of its result. Until this problem is addressed more emphatically, the full potential of seed health testing will not be realised.

This paper attempts to illustrate the circumstances under which seed health testing can contribute to disease management. It will identify what further data are required to interpret and use the information on seedborne disease loading in order to effect appropriate management action. It will review technological developments in diagnostic methodology and indicate their utility for particular circumstances and any attendant disadvantages.

1. Reeves JC, 1995. In Skerritt M, Appels R, eds, New Diagnostics in Crop Sciences. CAB International, Wallingford, UK, pp. 127-149.