We determined DNA fingerprints and pathotypes on international differe
ntials for 151 isolates of the rice blast fungus Pyricularia grisea th
at were collected from 15 rice cultivars at a blast disease breeding n
ursery in Santa Rosa, Colombia. The pathogen population expressed an e
xtraordinarily high number (39) of pathotypes. DNA fingerprinting, usi
ng the P. grisea repetitive DNA sequence MGR586, identified 115 haplot
ypes that were partitioned into six discretely distinct genetic lineag
es. The mean genetic similarity within lineages was high, ranging from
92 to 98%. The mean similarity between lineages ranged from 37 to 85%
. Each lineage was associated with a specific subset of cultivars and
expressed a generally nonoverlapping subset of pathotypes. Typically,
the pathotypes within a lineage had a closely related infection spectr
um, differing by single compatibility differences on a particular subs
et of the international differentials. The cultivar and pathotype asso
ciations indicated that the lineages express historically divergent vi
rulence spectra. MGR-DNA fingerprint analysis also deciphered the line
age associations of 31 isolates whose pathotypes could not be scored,
i.e., the isolates were avirulent on all international differentials.
MGR-DNA fingerprint analysis provides the genealogical framework for e
valuating pathogen variation that is essential for understanding disea
se epidemiology and pathotype evolution. This kind of analysis may be
particularly useful for selecting germ plasm sources for more durable
blast-resistance breeding.