MICRO-GEOGRAPHICAL GENETIC STRUCTURE OF INIA GEOFFRENSIS IN THE NAPO-CURARAY RIVER BASIN BY MEANS OF CHESSER´SMODELS - Biology, Evolution and Conservation of River Dolphins Within South America and Asia - page 136

(which measures the deviations from HWE inside each point sample) was obtained for each
one of the microsatellites analyzed as well as the number of positive and negative F IS values
in each one of the localities sampled.
A likelihood assignation analysis of individual genotypes in all the sampling localities
was carried out by using the procedures of Paetkau et al., (1995, 1997) and Waser & Strobeck
(1998).
Two AMOVAs were undertaken. In the first AMOVA, three hierarchical levels were
taken as the source of variation: among sampling localities, among individuals within the
sampling localities and within individuals (F ST , F IS , F IT ). In the second AMOVA, four
hierarchical levels were taken as the source of variation: among rivers (Napo and Curaray),
among sampling locations within the rivers, among individuals within the sampling localities
and within individuals (F CT , F SC , F IS , F IT ). Two out of nine microsatellites were excluded
from this analysis (KWM12a and PPHO 137) because of their low gene variability. The
probability significances were estimated by means of 1023 bootstrap permutations. For these
analyses, the Arlequin 3.1 software was employed (Excoffier et al., 2006)
Finally, the models of Chesser (1991a,b; 1993) were applied to the molecular data
obtained. For this, some definitions are necessary and based on terminology developed by
Cockerham (1967, 1969, 1973):
 = coancestry of parents in the same lineage
= coancestry of random offspring in the same lineage
 = coancestry of parents of different lineages
 = coancestry of random offspring of different lineages
F = coancestry of genes within random individuals
S = lineage number in the overall population
n = number of females for lineage
m = number of reproductive males for lineage
b = number of females within a determined lineage which bred with an specific male
 = probability that females of the same lineage have chosen the same male for breeding,
when more than one male breeds within a lineage.
In this case, I followed the Chesser's nomenclature for F-statistics, with individual
lineages in one overall subpopulation:
F LS = (F IL = (F - F IS = (F - , where F LS is the
proportion of genetic variance found among the lineages which integrated the population
studied, F IL is the gene correlation within individuals in regard to those within the same
lineage. F IS is the correlation of genes within the individuals in regard to those within the
overall population. Henceforth, F LS is equivalent to F ST , F IL is equivalent to F IS , and F IS is
equivalent to F IT by using the original definition of Wright (1951). The parameters, constants
and probabilities defined above were used to estimate transitions of coancestries over
successive generations. The four transitions of coancestries in generation t + 1 we are interest
in are listed below.
1.    mmt )/4 + ( mft )/2 + ((4 -  t /4), where  = (n - 1)/(ns - 1),  = 1/s,