Powdery mildew and leaf rust caused large yield losses in spring barley grown near Christchurch, New‐Zealand, in two seasons. Disease present during early growth stages was as damaging to yield as disease late in the season. Moderate leaf rust severities after anthesis were most damaging when combined with earlier mildew epidemics. Later growth did not compensate for reduced yield potential induced by early infection. This was attributed, at least in part, to an effect on leaf size, and therefore on green leaf area, at later growth stages. There was a closer relationship, by regression analysis, of yield to green leaf area than to disease severity in three cultivars. The three cultivars. which differed in yield potential and disease resistance, were not equally sensitive to disease. It is proposed that high yielding cultivars may be the most sensitive to yield constraint by disease.

Recently it has become evident that the estimates of crop losses by biotic and abiotic agents are not adequate for present-day requirements. With the increasing awareness of the need for more accurate and reliable crop loss estimates within both administrative and scientific disciplines, there has been an enhanced interest in the precision of disease measurement.

Reduced yields caused by powdery mildew and leaf rust in two seasons were associated with reduced plant growth. Combinations of early, late and full epidemics in one season, and 12 epidemic combinations in the second, were designed to identify crop sensitivity to disease by comparing growth and development with healthy plants. Early epidemics reduced ear number by increasing tiller death, and reduced grain number by effects on spikelet, floret or grain abortion, depending on the type of epidemic. Epidemics later in crop growth increased floret and grain abortion and also reduced grain weight. There was no compensation by later‐determined components for reduced growth and delayed development at earlier growth stages. Plants infected at early growth stages were more sensitive to late infections, seen as effects on the later‐determined components, than plants which were healthy initially. Interactions occurred between epidemics at different times and are likely to occur between diseases and other constraints.

No abstract