So far we have considered characters determined by a single gene with two alleles, occurring in sharply contrasting states, which can have a major affect on the fitness of the organism. In some cases we are justified in modelling selection in this manner, but in many cases, probably the majority, we are not. It is possible to expand the basic theory to consider characters determined by two gene loci, but this approach is no longer useful when we consider characters that are determined by many genes. In these cases we may observe a general relationship between parent and offspring, which suggests that there is an underlying genetic basis to the trait, but we usually do not know how many genes are involved or how they interact. In addition, we may also be aware that the environment influences the trait to some extent. Consequently, in order to study these traits we examine their variability, and attempt to dissect this variation into its genetic and environmental components. This type of analysis is called quantitative genetics.

We can consider three types of quantitative traits (Hartl and Clark 1989):

1. Meristic traits in which the phenotype is expressed in discrete, integral classes. Examples include litter size or number of seeds produced per individual, number of flower parts, and kernel colour in wheat.

2. Continuous traits in which there is a continuum of possible phenotypes. Examples include height, weight, oil content, milk yield, human skin colour, and growth rate. In practice, similar phenotypes are often grouped together into classes for the purposes of analysis.

3. […]