A method is presented for the estimation of nucleotide diversity and g
enetic structure of populations from RAPD (random amplified polymorphi
c DNA) data. It involves a modification of the technique developed by
Lynch and Crease (1990) for the case of restriction sites as survey da
ta. As new elements the method incorporates (i) dominance correction,
(ii) values of asexual reproduction of the populations sampled, and (i
ii) an analytical variance of the number of nucleotide substitutions p
er site. Sampling was carried out at two geographic scales for three a
phid species. At a macrogeographic scale, populations of Rhopalosiphum
padi did not show statistical genetic differentiation. Aphis gossypii
and Myzus persicae, which were sampled at a microgeographic scale, sh
owed a higher genetic differentiation than R. padi, it being statistic
ally significant in M. persicae. The major sources of sampling varianc
e within-and between-populations were found to be nucleotide (i.e., th
e number of alleles used as a function of the number of primers used)
and population (i.e., sample size) sampling. Extremely low estimates o
f nucleotide diversity were obtained for the species studied here. Thi
s result is consistent with previous reports on genetic diversity for
the same or other aphid species which were based on allozyme polymorph
ism, mitochondrial DNA variation and qualitative analyses of RAPDs.