Mr. Dudash et al., 5 GENERATIONS OF ENFORCED SELFING AND OUTCROSSING IN MIMULUS-GUTTATUS- INBREEDING DEPRESSION VARIATION AT THE POPULATION AND FAMILY LEVEL, Evolution, 51(1), 1997, pp. 54-65
The focus of this study was to examine the consequences of five sequen
tial generations of enforced selfing and outcrossing in two annual pop
ulations of the mixed-mating Mimulus guttatus. Our primary goal was to
determine whether purging of deleterious recessive alleles occurs uni
formly between populations and among families, and thus gain insights
into the mode of gene action (dominance, overdominance, and/or epistas
is) governing the expression of inbreeding depression at both the popu
lation and family levels across the life cycle. Inbreeding depression
was detected across the five-generation breeding program in both popul
ations for germination success, total flower production and adult abov
eground biomass. No inbreeding depression was detected for date of fir
st flowering. The serial breeding program minimized selection and may
have allowed the random fixation of deleterious recessive alleles. Thu
s, at the population level this experiment is consistent with (1) domi
nance with weak selection on partially deleterious recessive alleles;
(2) random fixation of deleterious alleles; and (3) overdominance beca
use we observed a steady state of performance following inbreeding and
outcrossing in these two populations of M. guttatus. At the family le
vel, however, significantly different maternal line responses (materna
l line x pollination treatment and maternal line x pollination treatme
nt x generation) provide a mechanism for the invasion of a selfing var
iant into the population through any maternal line exhibiting purging
of its genetic load. These family level differences in inbreeding depr
ession across the five generations suggest that dominance rather than
overdominance underlies the expression of inbreeding depression. No ev
idence of epistasis was detected at the population level. A significan
t interaction occurred, however, between the inbreeding coefficient, f
(2), and maternal families for all traits examined suggesting that epi
stasis could be playing a role in the expression of inbreeding depress
ion among maternal lines.