Biology Reference
In-Depth Information
loci affecting the trait. The number of loci involved generally can be determined
only with elaborate and specially designed experiments. Several assumptions
are made when estimating heritability: 1) all loci affecting the trait act
indepen-
dently
of one another and 2) the loci are
unlinked
(located on different chro-
mosomes). Another assumption 3) is that environment affects all genotypes in a
similar manner. These three assumptions are not always justified. Thus, heritabil-
ity estimates are difficult to interpret, although they are useful for predicting
response to selection under specific environmental conditions.
Heritability estimates usually are made by regression-correlation analyses
of close relatives (parent-offspring, full sibs, half sibs), experiments involving
response to selection, or analysis-of-variance components. Traits with high heri-
tability respond readily to selection with an appropriate selection method. The
magnitude of the response to the selection, that is the differences in mean values
between parent and progeny generations, provides an estimate of heritability
in the narrow sense (h
n
2
). This estimate is valid only for the population being
examined, under the test conditions used, for the behavior observed, and for the
method of measurement used.
Heritability of most insect behaviors is relatively high, probably because many
arthropod behaviors are highly stereotyped (
Ehrman and Parsons 1981
). For exam-
ple, the heritability of locomotor activity of
D. melanogaster
has been estimated to
be 0.51, and heritability of mating speed of male
D. melanogaster
has been esti-
mated to be 0.33. Heritability for honey production from honey bees ranged from
0.23 to 0.75, depending upon the experimental conditions and colonies tested
(
Rinderer and Collins 1986
). Italian honey bees are less able to remove the parasitic
mite
Varroa
than Africanized bees, and the heritability of this ability was estimated
to be 0.71 (
Moretto et al. 1993
). Heritability of the length of the prereproductive
period in
Helicoverpa armigera
, which is when migratory flight occurs in this noc-
tuid moth, ranged from 0.54 to 0.16 (
Colvin and Gatehouse 1993
). Heritability of
host selection behavior by
Asobara tabida
, a parasitoid of
Drosophila subobscura
,
ranged from 0.03 to 1.0 depending upon the test method used (
Mollema 1991
),
illustrating that test conditions are crucial to behavior analysis.
11.4.3 Some Polygenically Determined Behaviors
Behavior often is a continuous variable, controlled by multiple genes with small
additive effects (
Plomin 1990, Heisenberg 1997
). The task of teasing apart the
respective roles of genes and environment requires statistical analysis (
Doerge
2002
).
Drosophila
behaviors determined by multiple genes include locomo-
tor activity, chemotaxis, duration of copulation, geotaxis, host-plant preference,
mating speed, phototaxis, preening, and the level of sexual isolation within and
