Behavioral Genetics and Animal Domestication - Genetics and the Behavior of Domestic Animals - page 57

selection or the like. The requirements at this stage are only the behavioral
phenotypes for all individuals and knowledge of their relatedness (the
pedigree). Heritability studies can further dissect traits down to distinguish
between broad-sense and narrow-sense heritability, with the genetic compo-
nent further dissected down to additive and dominance components (additive
in this case refers to those loci which give a cumulative effect, whereas
dominance represents the interactions between alleles at the same locus)
( Falconer and Mackay, 1996; Lynch and Walsh, 1998 ).

Quantitative Trait Loci (QTL) and Association Mapping
QTL analysis is nowadays a very well-established technique, which became
increasingly popular in the 1990s ( Lander and Botstein, 1989 ). In recent
times it has enjoyed something of a renaissance, first with the use of micro-
satellite markers, and now with SNP markers and resequencing technology.
Microsatellite markers are regions of the DNA that are non-coding and con-
tain small motifs (two, three or four bases long) that are repeated anywhere
from three to a hundred or more times. The nature of these repeats means
they are much more likely to expand than other regions (they are typically
in non-conserved regions, hence this expansion is rarely a problem), thereby
generating a large degree of variability, and making them excellent as molec-
ular markers. In contrast, SNPs are any single-base changes that are present
in the genome. They are obviously by far the most prevalent marker in
the genome, and the ability to identify these markers has truly opened up the
world of genomics, not least with the ability to reliably saturate any given
genome with a wealth of markers. These markers are finally enabling the
step to gene discovery to be performed (previously very much a weakness
with this approach).
At its most basic, this is the crossing of two distinct populations that
differ for one or more traits of interest. As such, this technique is particularly
amenable for domestic animals, where extremely large phenotypic differ-
ences frequently occur between wild and domestic populations ( Andersson
and Georges, 2004 ). The inter-cross individuals are then genotyped for multi-
ple markers spread throughout the genome, with the genotype information
correlated with the phenotypic data. This enables the identification of the
number of loci affecting the traits that differ between the two populations,
their effect size and their genomic location. The loci can be defined as their
additive and dominance components, whilst pleiotropy and epistasis (multi-
ple genes interacting with one another) can also be identified. Two-way epis-
tasis is the only realistic form of detectable epistasis, due to the massive increase
in the number of tests performed (essentially an exponential increase).
As mentioned, the major issue is one of resolution of the detected loci
(with very large confidence intervals generally the norm) and the sample
sizes required for

the detection of smaller effect loci ( Beavis, 1998 ).