3 Populations and Pathways
89
outliers over multiple environments. They were able to determine whether the loci
closest to a growth rate QTL were the same as loci showing elevated differentiation in genome-wide scans of natural populations (Campbell and Bernatchez 2004).
They found that eight AFLP loci (so-called “QTL homologues”) closest to the QTL
for growth rate showed values outside the empirically determined 95% confidence
limits for genetic differentiation estimated from 440 AFLP loci. Thus suggesting
that differentiation at these loci was due to selection on nearby growth rate loci.
The authors were able to show that one AFLP locus corresponding to a growth rate
QTL exhibited significantly higher levels of genetic differentiation between ecotypes than expected under neutrality. This exemplifies a similar genetic basis for
adaptations of the normal and dwarf ecotypes in each lake, and how particular portions of the genome are related to dramatic phenotypic changes for a multigenic
trait as growth. Hence, such a study reinforces the potential of a complementary
QTL/genome scan approach towards studies of adaptive divergence. As Campbell
and Bernatchez’s (2004) study did not use a dense linkage map (25 of the 40 presumptive linkage groups were covered), and “only” 440 AFLP markers were used,
it also means that our understanding of the basis of adaptive divergence across sister
species and/or ecotypes may be improved even with the sparse QTL mapping that
is presently possible in marine species.
Hence, although time intensive, costly and challenging, developing linkage maps,
searching for QTLs, and combining information with genome scans arguably represents a comprehensive way of identifying genomic regions contributing to adaptive
variation, especially for multigenic traits (Price 2006, Stinchcombe and Hoekstra
2008). Given this, there are still drawbacks to this approach, which are exemplified
by an example in stickleback.
A microsatellite-based linkage map had established that the portion of the
genome containing a large-effect QTL contributing to adaptive variation in pelvic
morphology between oceanic and lake populations was shown to cover a region
of ≈10 Mb (Shapiro et al. 2004). Associated molecular knowledge including
sequencing of a BAC (bacterial artificial chromosome) covering this region of the
stickleback genome enabled the identification of a gene (Pitx1), which showed
expression differences associated with a reduced pelvis in lake populations (Shapiro
et al. 2004). However, the nature of the precise molecular change driving the phenotypic change – certainly in cis-regulatory modules in this particular case – has yet
to be identified. Such a situation raises at least three questions:
• How is a similar type of question approached without BACs or some substantial
fraction of genome sequence?
• What would happen if this QTL was not a large-effect QTL? It is clear that the
developmental processes of the hind limb and pelvis are under the control of a few
major genes in vertebrates (e.g. Marcil et al. 2003) that translate into large-effect
QTLs. QTLs affecting most ecologically important traits certainly do not belong
to such a category. Hence, such clear genotype-phenotype coupling cannot be
expected in most cases.
89
outliers over multiple environments. They were able to determine whether the loci
closest to a growth rate QTL were the same as loci showing elevated differentiation in genome-wide scans of natural populations (Campbell and Bernatchez 2004).
They found that eight AFLP loci (so-called “QTL homologues”) closest to the QTL
for growth rate showed values outside the empirically determined 95% confidence
limits for genetic differentiation estimated from 440 AFLP loci. Thus suggesting
that differentiation at these loci was due to selection on nearby growth rate loci.
The authors were able to show that one AFLP locus corresponding to a growth rate
QTL exhibited significantly higher levels of genetic differentiation between ecotypes than expected under neutrality. This exemplifies a similar genetic basis for
adaptations of the normal and dwarf ecotypes in each lake, and how particular portions of the genome are related to dramatic phenotypic changes for a multigenic
trait as growth. Hence, such a study reinforces the potential of a complementary
QTL/genome scan approach towards studies of adaptive divergence. As Campbell
and Bernatchez’s (2004) study did not use a dense linkage map (25 of the 40 presumptive linkage groups were covered), and “only” 440 AFLP markers were used,
it also means that our understanding of the basis of adaptive divergence across sister
species and/or ecotypes may be improved even with the sparse QTL mapping that
is presently possible in marine species.
Hence, although time intensive, costly and challenging, developing linkage maps,
searching for QTLs, and combining information with genome scans arguably represents a comprehensive way of identifying genomic regions contributing to adaptive
variation, especially for multigenic traits (Price 2006, Stinchcombe and Hoekstra
2008). Given this, there are still drawbacks to this approach, which are exemplified
by an example in stickleback.
A microsatellite-based linkage map had established that the portion of the
genome containing a large-effect QTL contributing to adaptive variation in pelvic
morphology between oceanic and lake populations was shown to cover a region
of ≈10 Mb (Shapiro et al. 2004). Associated molecular knowledge including
sequencing of a BAC (bacterial artificial chromosome) covering this region of the
stickleback genome enabled the identification of a gene (Pitx1), which showed
expression differences associated with a reduced pelvis in lake populations (Shapiro
et al. 2004). However, the nature of the precise molecular change driving the phenotypic change – certainly in cis-regulatory modules in this particular case – has yet
to be identified. Such a situation raises at least three questions:
• How is a similar type of question approached without BACs or some substantial
fraction of genome sequence?
• What would happen if this QTL was not a large-effect QTL? It is clear that the
developmental processes of the hind limb and pelvis are under the control of a few
major genes in vertebrates (e.g. Marcil et al. 2003) that translate into large-effect
QTLs. QTLs affecting most ecologically important traits certainly do not belong
to such a category. Hence, such clear genotype-phenotype coupling cannot be
expected in most cases.