Electrophoretic separation of allelic variants of enzymes (allozymes) has been used to study the population genetics of a wide variety of species for many years (see Avise [1974]). Although electrophoretically-evident variation is only a small portion of the genetic variation of a gene, many enzymes have been shown to exhibit high levels of allozymic variation. Planktonic crustaceans seem particularly variable [ Nelson and Hedgecock, 1980].
More recently, genetic variation at the molecular level [i.e., structural traits of nucleic acids such as DNA (deoxyribonucleic acid)] has been used in population genetic studies. A variety of technical approaches have been used to assay molecular variation for population genetic studies. One means of investigating molecular variation is through digestion of the purified DNA with restriction enzymes that cut the DNA strand wherever a recognition site sequence (usually 4 to 8 base pairs) occurs. The resultant restriction fragment length polymorphisms (RFLP) yield a molecular fingerprint that may be unique to a particular individual [ Bruford et al., 1992]. Of great interest as potential indicators of population structure are nuclear regions called variable number tandem repeat (VNTR) loci (see Quellar et al. [1993]). Unlike allozymes, the sequences do not encode functional gene products; these non-coding regions are frequently highly variable [ Jeffreys et al., 1985]. It is now also possible to rapidly assess molecular variation at the highest level of detail: that of the nucleotide base sequence of the DNA molecule. Direct sequencing of selected portions of genes without an intermediate cloning step allows assessment of numerous individuals in population genetic studies [ Innis et al., 1988].
Many molecular studies of the population genetics of marine organisms have targeted mitochondrial DNA (mtDNA) (see, e.g., Avise et al. [1987]), which is found in a cytoplasmic organelle and encodes genes involved in central metabolic processes. MtDNA has the distinct advantage that it is inherited only from the maternal parent and without recombination in most organisms, unlike nuclear genes which experience recombination each generation. Since the time it may require for zooplankton to cross an oceanographic domain (such as the North Atlantic) may be far longer than the life-span of individual zooplankton, it is advantageous to use as tracers genetic characters that are identical from one generation to the next. Mitochondrial traits thus provide conserved markers for large-scale and/or long-term studies of dispersal. Also, mitochondrial genes frequently exhibit greater population differentiation than nuclear genes [ Birky et al., 1989], because male dispersal does not cause exchange of mitochondrial genes.
The traits mentioned above are only a few of a great many genetic traits that may be assayed to measure genetic diversity. Since each trait and its variants has a technical name, genetics is fraught with special terminology. The variants discussed below include: alleles (alternate forms of a gene, characterized by electrophoretic mobility or DNA sequence variation), genotypes (the assortment of alleles in an individual), and haplotypes (the genotypes of single-copy genomes such as mtDNA).
The next step in the measurement of genetic diversity is to calculate one of
numerous statistics that take into account the numbers and/or types of variants
of a genetic trait. There are two fundamental types of genetic diversity
statistics: one is based on the numbers of differences between variants and one
is based on the number of variants. An example of the first type is nucleotide
diversity (
), which is calculated by the formula:
where 
is the proportion of different nucleotides between the ith and jth variant
and 
is the total number of sequence comparisons 
[ Nei, 1987]. An
example of a statistic based on the number of variants is the diversity measure,
h, which is calculated by the formula:
where 
is the frequency of the ith variant [ Nei, 1975].