Plant breeding is always a numbers game. […] In addition, we are looking for rare alleles, so the more plants we try, the better. These rarities may be new mutations, or they can be existing ones that are neutral—or are even selected against—in a wild population. A good example is mutations that disrupt seed dispersal, leaving the seeds on the heads long after they are ripe.
2013 May-June, David Van Tassel, Lee DeHaan, “Wild Plants to the Rescue”, in American Scientist, volume 101, number 3