To select candidate populations of wild species to be given priority f
or conservation, genetic criteria gained from the study of molecular m
arkers may be useful. Traditionally, diversity measures such as expect
ed heterozygosity or percentage of polymorphic loci have been consider
ed. For conservation we propose instead that priority should be given
to measures of allelic richness. To standardize the results of allelic
richness across populations, we sued the technique of rarefaction. Th
is technique allows evaluation of the expected number of different all
eles among equal-sized samples drawn from several different population
s. We also show how the contribution of each population to total diver
sity can be partitioned into two components. The first is related to t
he level of diversity of the population and the second to its divergen
ce from the other populations. For conservation purposes the uniquenes
s of a population-in terms of its allelic composition-may be at least
as important as its diversity level. These new descriptors are illustr
ated by means of isozyme and chloroplast DNA data obtained for an enda
ngered tree species, the argan tree of Morocco (Argania spinosa (L.) S
keels). With these analyses the conservation value of the argan tree p
opulations, especially those of two isolates present in the north of t
he country, can be better appreciated. The methods proposed to identif
y priority areas for conservation of the genetic resources of the arga
n tree are compared to those sometimes advocated in the case of reserv
e design, where one of the goals is to maximize species richness.