CHAPTER IV 

 FIELD PLOT TECHNIC 



In carrying out crop-breeding studies the number of varieties 

 and strains has been greatly multiplied. Vilmorin's isolation 

 principle, whereby the value of any selection is determined by the 

 breeding nature of the progeny, has been universally adopted. 

 The field is then the plant-breeder's laboratory and the question 

 of correct field technic is of the utmost importance. 



The difficulties, of making all conditions of similar nature for a 

 large number of strains or varieties which must be tested, are 

 very numerous. The method used must be such that the per- 

 formance will be a correct indication of the comparative value 

 of the strains being tested when grown under farming con- 

 ditions. The purpose of the present chapter is to discuss field 

 plot technic for such disturbing factors as soil heterogeneity and 

 climatic conditions. 



SOIL HETEROGENEITY 



The field selected for the comparative trials should be repre- 

 sentative of the soil and climatic conditions under which the crop 

 will be grown. The land must then be cropped in such a manner 

 that it is kept in a uniform state of good productivity. In order 

 to do this, it is necessary to observe some one of the standard 

 rotations. It is a good practice to have one or more bulk crops 

 rotated with the breeding plots in order to keep the land uniform. 

 If only one area of land is available there is then no choice and the 

 investigator must see that this field is treated in the best possible 

 way. If more than one field is available, it is possible to deter- 

 mine which is more nearly uniform by a correlation of contiguously 

 grouped plots as outlined by Harris (1915). 



Harris' Method of Estimating Soil Heterogeneity. By 

 Harris' method the coefficient of correlation is used as an index 

 of soil uniformity . This statistical constant measures the degree 

 of correlation between contiguous plots grouped in a certain way. 

 If the variation in yield from plot to plot is simply due to random 

 sampling, there will be no correspondence between contiguously 

 grouped units. On the other hand, if the field is "patchy" 



51 



