Biology Reference
In-Depth Information
havior, and clearly its ef ects were dependent on the environment. It
was also clear that the pollen-hoarding QTL must interact in a complex
network with ef ects on multiple foraging traits.
6.5
Pln4
and Mapping the Interactions
of Pollen-Hoarding QTLs
h ree pollen-hoarding QTLs had been identii ed and verii ed to af ect
individual foraging behavior. Two of the QTLs demonstrated ef ects
that were counter to our expectations—the high hoarding and foraging
ef ects were inherited from the low strain. We had two hypotheses to
explain these results:
1. h e alleles for
pln1
and
pln3
were i xed (identical within strains
but dif erent between them) by chance in the high and low strains, not
by our selection. A high allele for each was i xed in the low strain, and
a low allele for each was i xed in the high strain. h eir overall pheno-
typic ef ects on the strains were negative, but we could not get rid of
them. h is can occur in a small breeding population, a consequence of
ge ne tic drit (see Section 4.2.2). But we did not think that this was the
case because we had outcrossed (a breeding technique where one breeds
selected stock to parents outside the selected breeding population in
order to introduce new genetic material) and reselected and continued
to get consistent results. Alleles with ef ects of 33 percent and 10 per-
cent of the total phenotypic variance should be easy to select in or out
of a breeding population. h e outcross should have of ered an opportu-
nity to get the alternative alleles for the two QTLs into the strain gene
pools and select out the alleles giving opposite ef ects.
2. h e ef ects of the alleles were conditional on the other genes with
which they interacted. In this case, the ef ect of the low allele acting on
its own might be high pollen hoarding or foraging, but in combination
with low (or high) alleles at other loci the ef ect could be toward the low-
pollen-hoarding/foraging phenotype. h is is a characteristic of epistatic
interactions between genes known as allele-conditional ef ects.
We needed to look at the interactions of the QTLs in detail. h is
time Olav Rueppell, a postdoctoral fellow in my lab, took on the task.
