INTERGENOTYPIC COMPETITION AND BORDER EFFECT IN BREAD WHEAT AND BARLEY

A coordinated series of experiments aimed to estimate border effect an d intergenotypic competition was carried out on bread wheat and barley in Italy and Germany, The performance of the varieties grown in alter nate rows 14-17 cm apart proved to be strictly dependent on the border ing variety. The varieties grown in adjacent plots consisting of eight rows showed significant border effects: the two external rows yielded up to 40% more than the two innermost, and the two 30 cm end hills fa cing the alleys yielded 29% more than the central part of the plot. It has been suggested that in variety trials the first 30-40 cm at both ends of each plot should be removed mechanically. In spite of the bord er effect detected in the outer rows of the plots, the varieties simil arly exploited the extra space available at the borders, and the space of 30 cm left between adjacent plots seems to be sufficient to dilute the intergenotypic competition at plot level to an extent only margin ally affecting variety performance. In conclusion, grain yield based o n whole plot (8 rows, 6-10 m(2)) appears to provide a good estimate of 'true yield', providing that it is based on the actual area of the pl ot, including the uncropped strip between adjacent plots. By contrast, the reduction of plot size to 1-3 rows is not feasible without introd ucing considerable bias into the estimation of the yield potential of the varieties.