An investigation of the reliability of ELISA as a practical test for detecting potato leaf roll virus and potato virus Y in tubers

Enzyme‐linked immunosorbent assay (ELISA), involving mechanical sampling of tubers, was compared with a growing‐on test in which virus was assessed visually, and with ELISA on leaves. The percentage infection of potato leaf roll virus and potato virus Y on samples of potential seed tubers sent for advisory test during the winter, was determined by each method. The tuber ELISA underestimated the incidence of both viruses, and was less accurate than the growing‐on or leaf ELISA. The effect on the reliability of tuber ELISA of the amount and distribution of virus in the tuber, the compromises made to make the test fast and simple, and the quality of anti‐serum and composition of the extraction buffer are discussed.