Selection 377 



offspring would on the average be smaller than the offspring 

 from the 51- or 52-cm plant. Offspring of a 54-cm plant which 

 was homozygous for all six contributing genes would be identical 

 with those produced by the 51-cm plant but not with those 

 from the 54-cm plant, which was merely an extreme variant of 

 the next lower class. 



To be successful, selection must be based on the genotype. 

 If two plants have the same genotype but differ because of 

 environmental conditions, it makes no difference whether the 

 larger or the smaller plant is used for seed, for the offspring of 

 each will have the same mean and standard deviation (within 

 the limits of error). A small variant from a plant with more 

 contributing genes will, however, be a better choice for a parent 

 than a large variant from one with fewer contributing genes. 

 Selection acts only on genotypical differences and not on fluc- 

 tuations. Environmental variants are of no value in breeding. 



In our hypothetical problem, we repeatedly selected for re- 

 production the seeds of single plants, and we always kept the 

 offspring from one plant separated from those of other plants. 

 This method is known as line selection. Another method often 

 used is mass selection. Either method may be based upon the 

 nature of the father, the mother, or both parents. In our prob- 

 lem, also, we chose the plants to be used for further reproduction 

 according to the nature of their phenotypes. Other methods 

 are selection according to the nature of the ancestors of the 

 plants used for breeding and selection upon the basis of the 

 phenotypes of their offspring, a method known as "progeny selec- 

 tion or the progeny test. 



Mass selection is probably the oldest method of selecting 

 parents. By this method, a farmer will go through his field 

 and pick a number of his best plants for seed. His selection 

 may be based upon phenotype, pedigree, or performance but 

 probably in the earlier selections was based entirely on pheno- 

 type. If the plant is normally self-pollinated, selection will take 

 both parents into account; but if it is normally cross-pollinated, 

 the male parent will frequently be disregarded and only the 

 quality of the female considered. The seeds of the various 

 plants chosen for reproduction are then sown together, and no 

 attempt is made to keep separate the seeds from individual 

 plants. 



