ADAS High Mowthorpe, Duggleby, Malton, North Yorkshire, Y017 8BP, UK

Background and objectives
In a temperate maritime climate, biotic rather than abiotic stress can be the predominant constraint on wheat yield. Selection of the small number of wheat cultivars that are best able to exploit the favourable water and solar radiation environment leads to genetic homogeneity across the wheat area. Erosion of oligogenic resistance and greater dependence on more durable, quantitative resistance follows. The latter is insufficiently effective to allow yield potential to be achieved without fungicidal intervention. Crop managers are therefore faced with decisions about the amount of fungicide an individual crop requires in order to maximise its economic output. This paper illustrates how the disciplines of crop protection, epidemiology and crop physiology might be integrated to support these disease management decisions.

Results and conclusions
A set of hypothetical steps, which link the cost of fungicide input to the economic value of yield output, have been defined, whereby: (i) increasing fungicide dose is linearly related to input cost; (ii) increasing dose decreases disease severity, according to a dose-response curve; (iii) decreased disease severity increases crop green leaf area index (GLAI, the number of units of planar area of green leaves, per unit of ground that they occupy); (iv) increasing GLAI increases fractional interception of photosynthetically active radiation (PAR), by analogy with Beer's Law [1], increased fractional interception increases crop dry matter accumulation; (vi) yield increases depending on the partitioning of dry matter to grain; and (vii) economic output is linearly related to yield. The curvature of the disease dose-response and Beer's Law functions results in a yield dose-response of monomolecular form, with diminishing return from each additional dose increment. The appropriate dose of fungicide input for an individual crop can be determined experimentally, as the point on the yield response curve at which the value of yield output, minus fungicide cost, is maximised.

As decisions on fungicide requirement have to be made before the yield response can be known, efficient fungicide use depends on the precision with which the response can be predicted [2]. Finite research resources should be focused on improving understanding of variation in those relationships which are responsible for the greatest proportion of the yield response variation, and about which we currently have least predictive ability. We consider each of the steps in turn. (i) Cost per unit dose is known. (ii) Recent research in epidemiology and crop protection has shown how substantial variation in disease dose-response arises from variation in the severity of untreated disease (resulting from variation in the time of onset and rate of the epidemic) and from variation in efficacy of control per unit dose (a function of active ingredient and spray timing). (iii) Crucially, there are insufficient published data describing the effect of disease on GLAI. (iv) Through Beer's Law, the size and morphology of the crop canopy relate tightly to the way interception of PAR decreases with GLAI loss. Early evidence indicated that the amount of dry matter accumulated per unit of light energy intercepted was reasonably conservative, but there is now evidence that some significant variation exists, which remains to be accounted for. (vi) Partitioning of dry matter to grain is now understood to vary significantly, particularly with the balance of growth pre- and post-anthesis. (vii) Grain value is known with reasonable precision at the time of treatment decisions.

This analysis suggests that the pressing task for applied pathologists is to improve prediction of the rate at which the pathogen will deplete the PAR intercepting surface. Advances in crop physiology should aid prediction of the effect of green area loss on yield.

1. Monteith JL, Unsworth MH, 1990. Principles of environmental physics. Edward Arnold, London.
2. Paveley ND, Lockley KD, Sylvester-Bradley R, Thomas J, 1997. Pesticide Science 49, 379-388.