The Influence of the Length and Number of Rows on Experimental Error in Field Trials With Carrots and Onions

In accumulating the immense amount of data now available from the various research institutions on variety, fertilizer, and other studies of vegetable crops there has been no uniformity of practice regarding plot size. In some cases it has been the practice of workers to employ large-sized plots while others have believed that a small plot is entirely sufficient. It has been especially unfortunate that until recently but few workers have realized the uncertainty of results obtained from single plot treatments. In extensive plot work with cabbage, carrots, celery, lettuce, onions, and other vegetable crops on peat soils in New jersey, New York, and Indiana during the past twelve or fifteen years it has become perfectly apparent to the author that single plot treatments are of value only as rough indications of crop tendencies under varying conditions.In the work referred to the plots varied in size from three square rods to sixteen square rods, or from about a fifty-third to a tenth of an acre. A cursory examination of the mass of data obtained during the course of the work was sufficient to show that greater variation in yield and behavior existed between adjacent similarly treated large sized plots than was the case with the smaller areas. The apparent superiority of the smaller plot size naturally led to the inquiry as to how small the plot could be and still give adequate results. The smaller plot seemed desirable because a greater number of treatments and replications can be made on a given area, and at less expense.