In 6 experiments the effects of various factors on plant-weight variation (sum of squares of deviation from the mean shoot weight) at harvest were examined [UK]. These factors were seed size, sowing depth, time of seedling emergence, size at different stages of development, within-row spacing, density and row position within a bed. In a drilled experiment, up to 25% of the variation in mature leek size could be attributed to variation in time of seedling emergence. In the experiments which used transplants, 15% of the variation in the mature crop could be associated with variation in size shortly after seedling emergence from peat blocks. Variation in row position within a 4-row bed never accounted for more than 11% of the weight variation at harvest, and variation in within-row spacing never accounted for more than 4% of weight variation. The coefficient of variation (c.v.) of leeks growing at 22 plants/m2 was 44%; the c.v. of those growing at 50 plants/m2 was 61%. Using large seeds or sowing deeply gave a quicker emergence and a higher percentage emergence than using small seeds or sowing shallowly. These factors each never accounted for more than 7% of the weight variation in mature leeks. Most of the plant-weight variation could not be attributed to any of the measured factors in any of the experiments. The c.v. of leeks which had emerged on a specific day and had been grown in a uniform environment, was between 50 and 66%, which was similar to the c.v. observed in the transplant and drilled experiments. Selecting for large transplants reduced the c.v. to 28% in 1 experiment; this effect was only partly attributable to the greater mean plant size resulting from this treatment.
