Wild Phaseolus vulgaris is distributed between northern Mexico and nor
thern Argentina. Analysis of phaseolin and molecular markers (isozymes
, Restriction Fragment Length Polymorphisms or RFLPs) indicate that th
is gene pool consists of two major groups, Mesoamerican and Andean, an
d a third intermediate gene found in northwestern South America. Previ
ous to this study, only four accessions of wild P. vulgaris beans from
Bolivia had been collected and their genetic relationship, with other
wild beans from Latin America was not known. Due to the problem of in
tense erosion in some areas of Bolivia, it was our objective to survey
and document Phaseolus spp. in this area before their extinction. We
conducted a collection expedition in May 1994 in the departments of Co
chabamba, Chuquisaca and Tarija. This resulted in collections of four
populations of P. augusti, two of cultivated P. lunatus and two mixtur
es of cultivated P. vulgaris. The first mixture was made of ''k'opurus
'' or beans consumed after toasting, and represented an addition of 17
accessions to the Bolivian collection. The second mixture was made of
''porotos'' and resulted in the addition of 10 new accessions. Seven
germplasm collections of wild P. vulgaris were found, which allowed us
to increase the number of known populations of wild common bean for B
olivia. Another accession was found as a wild-weed-crop complex. Seven
of these wild P. vulgaris accessions along with another accession fro
m Bolivia collected previously, and a number of P. vulgaris accessions
from Mexico (17), Guatemala (3), Colombia (10), Ecuador (6), Peril (1
7) and Argentina (16) were analyzed with RAPDs. The use of 14 random p
rimers and one SCAR (Sequence Characterized amplified Region) resulted
in 90 bands, of which 83 were polymorphic. This data was used to cons
truct a dendrogram which shows clear separation into three clusters, c
orresponding to each of the gene pools and on intermediate group. The
Bolivian wild P. vulgaris beans grouped with the accessions of souther
n Peru and Argentina into the Andean gene pool. RAPD analysis of genet
ic diversity correlated well with genetic diversity obtained with othe
r markers. Moreover, the ease of analysis allowed us to obtain a large
number of bands which was conducive to greater sensitivity and identi
fication of geographic subgroups and accessions of hybrid origin.