Abstract:
In order to better understand the evolutionary patterns revealed by molecular
markers, we studied genetic diversity and differentiation of populations and races of
Douglas-fir using randomly amplified polymorphic DNA (RAPD) markers of nuclear and
mitochondrial origin. We conducted a range-wide survey of RAPD diversity and
differentiation, and compared RAPDs and published allozyme data at the population and
race level. Southern hybridization using enriched organelle DNAs as probes was used to
screen RAPDs for cytoplasmic DNA markers, allowing us to distinguish between RAPD
bands of nuclear, mitochondrial, and putative chloroplast origin. RAPD markers of mitochondrial origin were found to be frequent among Douglas-fir RAPD profiles. In our range-wide survey, nearly half (45 %) of all polymorphic RAPDs scored exhibited maternal inheritance; the chioroplast genome is known to exhibit paternal inheritance in Douglas-fir. Hybridization of RAPD blots with
mitochondrial genome enriched DNA identified 16% of a total of 237 RAPD fragments as being of mitochondrial origin. Failure to distinguish between nuclear and mitochondrial RAPD origins could seriously bias estimates of population genetic
parameters. When a nuclear origin was assumed for all bands, differentiation among
populations and races was nearly 1.5-fold greater then when only nuclear RAPDs were
analyzed (GST=O.47 and 0.34, respectively). We infer that the large number of
mitochondrial RAPD markers we observed are indicative of a large, complex, and
polymorphic mitochondrial genome in Douglas-fir. Nuclear RAPD markers were considerably more polymorphic within populations than mitochondrial RAPDs (Hs=O.22 and 0.03, respectively), however mitochondrial RAPDs were much more highly differentiated than nuclear RAPDs (GSTO.91 and 0.34, respectively). Both markers indicated similar phylogenetic relationships among populations and races. Estimates of diversity and differentiation were higher for RAPDs than for allozymes, though not significantly so (P>0.05; H5O.22 and 0.16, HT=O.32 and 0.22, and G1=0.34 and 0.29, respectively, for RAPDs and allozymes). However, simulations showed that Hs and HT for dominant biallelic markers such as RAPDs are expected to be approximately half that of a codominant, multiallelic marker with
equivalent levels of underlying molecular diversity to that of our allozyme data set. This suggests that RAPDs reflect considerably higher amounts of diversity at the molecular level than do allozymes. The simulations also suggest that our estimates of population differentiation (GST) for RAPD markers are essentially unbiased at the sample sizes used.
