Biology Reference
In-Depth Information
studies showed that metamorphic and paedomorphic forms could be segre-
gated according to Mendelian expectations. In particular,
Humphrey (1967)
showed that when
A. mexicanum
/
A. t. tigrinum
hybrids were backcrossed to
A. mexicanum
, there is 1:1 segregation of phenotypes. Later,
Tompkins
(1978)
and
Gould (1981)
argued from Humphrey's single cross that
A. mexicanum
was an example of single gene, macromutational evolution.
To further test the single gene hypothesis,
Voss (1995)
crossed domesticated
A. mexicanum
and
A. t. tigrinum
to create F
1
hybrids and like
Humphrey
(1967)
, backcrossed these to
A. mexicanum
. The ratios of metamorphs
and paedomorphs arising from two relatively large backcrosses were consis-
tent with single locus control, thus supporting the classical idea of a single
mutation underlying the evolution of paeodomorphosis. In these crosses,
the
A. t. tigrinum
and
A. mexicanum
alleles were dominant and recessive in ef-
fect, respectively. Subsequently, DNA was isolated from individuals of these
crosses, amplified fragment length polymorphisms (AFPLs) were typed, and a
genetic linkage map was constructed. This map was used to roughly locate the
position of the major gene (
met1
) for paedomorphosis (
Voss & Shaffer, 1996
).
To determine if
met1
arose independently in the domesticated stock, the back-
cross design was repeated using
A. mexicanum
collected fromLake Xochimilco
(
Voss & Shaffer, 2000
). The segregation of phenotypes did not fit a single gene
model, and the segregation of AFLPs marking the
met1
region exhibited
segregation distortion, suggesting an epistatic effect between
met1
and other
loci on viability. Still,
met1
genotypes did associate with metamorphic and
paedomorphic phenotypes, indicating a difference in genetic background
between wild-caught
A. mexicanum
and the domesticated stock.
Voss and Smith (2005)
made additional
A. mexicanum/A. t. tigrinum
backcrosses and reared over 500 offspring to rigorously quantify the genetic
effect of
met1
from wild-caught
A. mexicanum.
These crosses showed that
met1
not only explained discrete variation in the expression of metamorpho-
sis and paedomorphosis but also continuous variation in metamorphic
timing. Approximately 20% of the individuals that inherited two
met1
alleles
from wild
A. mexicanum
were paedomorphic. The remainder of these ho-
mozygotes delayed metamorphic timing by an average of 36 days, relative
to heterozygotes. These results showed that
met1
determines the expression
of metamorphosis and paedomorphosis by altering metamorphic timing.
Within the context of hybrid crosses, alleles from metamorphic and paedo-
morphic species decrease and delay the time to metamorphosis, respectively.
Variation in the number of paedomorphic individuals generated from
wild-caught versus domesticated
A. mexicanum
crosses reflect differences
