Does interplot interference affect the screening of wheat for yellow rust resistance?

Summary In small plots, adjacent to one another, a representational error can be expected when screening for quantitative forms of resistance to airborne pathogens. The representational error or interplot interference may occur as an underestimation of the level of resistance and/or as an error in the ranking of the entries tested.Four experiments were carried out with wheat (Triticum aestivum) exposed to yellow rust (Puccinia striiformis), three in Kenya, one in Mexico (exp. III).In experiment I 57 entries were compared in an unreplicated trial with three adjacent and one isolated plot situation. The range in the standard deviation of and the ranking order for disease severity (DS) between the 57 entries were the same for all plot situations at all observation dates.In experiment II nine entries from experiment I, representing a wide range of quantitative resistance and having a similar heading date, were compared in adjacent plots consisting of two rows of 10 m in eight replicates and isolated plots of one, six and ten rows of 4 m in three replicates. The range in the standard deviation of and the ranking order for DS between the entries were very similar for the four plot situations.The 10 entries in experiment III differed from those in experiment II, but represented a similar wide range of resistance. Three adjacent plot situations of 0.9×0.5 m, 0.9×2.0 m and 2.7×2.0 m, respectively, were compared with one isolated one with plots of 2.7×2.0 m. The ranking order was not affected, the range in and the standard deviation of the DS in the isolated plots were slightly larger than in the adjacent plots.In experiment IV two mixtures of two entries each were made. Per mixture one entry was fairly resistant (R) the other rather susceptible (S). Within each mixture the entries had a similar heading date. The ratios of the R:S mixtures were 0:100, 50:50, 67:33, 75:25, 82:18 and 100:0. The DS of each entry was the same as its DS in monoculture irrespective of the mixture ratio. There was no mixture effect on DS.The three experiments in Kenya gave no indication of any interplot interference occurring. In Mexico there was a very slight underestimation of the resistance in adjacent plots. The ranking order was always the same irrespective of the test plot situation. The screening of wheat for yellow rust resistance in small adjacent plots is representative for the farmers fields. This is contrary to what has been found in other windborne pathosystems such as barley-barley leaf rust (strong under estimation of resistance), barley-powdery mildew (some under estimation of resistance and different ranking order) and durum wheat-stem rust (fair under estimation of resistance).