Pyramiding of partial disease resistance genes has a predictable, but diminishing, benefit to efficacy

Abstract

Partial resistance genes often need to be 'pyramided' into a crop cultivar to obtain commercially acceptable levels of disease resistance. Analysis of data from four different wheat mapping populations, segregating for partial resistance to four contrasting foliar pathogens, showed a diminishing benefit to disease control from increasing the numbers of resistance loci in wheat lines. To test whether a general function could describe the efficacy benefit from pyramiding, a simple multiplicative survival model (MSM) was used to predict disease severities on mapping population lines carrying various combinations of two, three or four resistance loci. The effectiveness of each resistance locus was expressed as the disease severity in lines carrying resistance alleles at one locus, as a proportion of the severity in lines carrying no detectable resistance alleles. The predicted severity from any given combination of multiple resistance loci was calculated as the product of the proportional severities for the relevant single loci. A regression line fitted to transformed observed against predicted values explained 93% of the variation, with a slope and intercept not significantly different from 1 and 0, respectively. MSM may therefore provide a simple method to test contrasting types of partial resistance and search for synergistic combinations. The analysis suggests that diminishing returns are a general feature of partial resistance to foliar pathogens in wheat, a finding that is likely to apply to other crop pathosystems. Identifying and combining ever more QTL is likely to provide limited gains. The consequences of these findings for QTL analysis are described.