or tissue from many plants to be effective; and (4) may be less expensive than phenotypic
selection. Although MAS does not replace the requirement for parent selection, sexual
recombination, and breeding strategies, it can significantly increase the efficiency by
which superior genotypes are selected. For this reason, MAS is considered to be an import-
ant modern enhancement of traditional plant breeding.
The theoretical advantages of MAS may not always be relevant, and it is often argued
that phenotypic selection is faster and cheaper than MAS for many traits. Some of the
factors that can detract from the success of MAS include (1) some breeding facilities
lack the technical facilities and expertise to apply MAS, (2) incomplete linkage between
a marker and a target QTL may reduce the effectiveness of MAS, (3) the marker must
be polymorphic on the parents, and (4) MAS is effective only if the alleles being selected
are important relative to other alleles in the population. This last factor is the key to the
success or failure of every MAS application. It may seem like an obvious statement, but
MAS relies on the ability to predict the value of alleles. The quality of those predictions
rests on many factors, but a key factor is the behavior of an allele in the presence of
other alleles and other physical environments where it has not yet been tested. For
example, a breeder might identify that allele A 1 at locus A has a positive effect on yield.
But this prediction would be made in a limited set of environments, and with a limited
set of germplasm. A breeder who crossed a parent containing allele A 1 with a new
parent containing allele A 4 , and selected for A 1 using a linked marker, might never discover
that allele A 4 is actually better than allele A 1 , or perhaps that allele A 1 causes plants to be
susceptible to a disease that was not present when A 1 was first characterized. For these
reasons, MAS should never be applied independently from phenotypic selection, and
most successful applications of MAS have been as an enhancement to phenotypic selection
rather than as a replacement.
3.5.3 Mutation Breeding
Mutations are genetic modifications that occur in the DNA of plants, producing new alleles
that are different from the alleles that the plant inherited from its parents. Mutations can be
small and localized, or they can cause major structural rearrangements of entire chromo-
somes. Localized mutations include base substitutions and small insertions or deletions.
Because most amino acids are coded by two or more different codons, many base substi-
tutions are “silent,” detectable only through DNA sequence analysis. Most mutations that
occur in noncoding DNA are also silent, although they can sometimes affect gene
expression or chromosome structure. Mutations that cause the transcription of a different
amino acid are more likely to cause phenotypic change, most likely through their influence
on protein folding or their alteration of an active site in an enzyme.
Fundamentally, the success of all plant breeding depends on mutations that have
occurred at some point in the evolution of a species. However, the great majority of
random mutations are deleterious, so breeders rely on a relatively small number of
mutations that have been presorted through natural selection because they provide some
type of selective advantage in at least one environment. Beneficial mutations that arise
naturally are very rare, and most probably go unnoticed. However, it is possible to
artificially induce mutations at frequencies that are much higher than the natural rate.
This can be done through radiation (usually applied to seeds prior to planting) or
through chemical induction.
76 PLANT BREEDING