I have been asked to present examples from species that have been studied with both protein and DNA markers to illustrate the types of results that one might expect from various genetic approaches. I was also asked to identify any consistent patterns that have been obtained and to discuss what to do when the results from various methods are not in agreement. This is not an easy challenge to fulfill in a brief report. The relative utility or value of different genetic approaches depends upon the question that is being asked. That is, there is no single best technique; rather, different techniques are better suited to approach different questions.

Genetic techniques are being used to answer a diverse array of problems in the management of Pacific salmon. However, a fundamental characteristic in many of these problems is the distribution of genetic variation within a species of interest. The total amount of genetic variation (H_T) can be partitioned into genetic differences among individuals within a single population (H_S) and genetic differences among different populations (D_ST; Nei 1977):

(1)

The proportion of total gene diversity that is due to variation among populations can then be represented as G_ST:

I will focus my discussion on the relative capability of different techniques to estimate the parameters in these two expressions.

Protein markers have been the workhorse for describing the genetic structure of natural populations over the last 25 years (Lewontin 1991). The strengths of protein electrophoresis are many. First, genetic variation at a large number of nuclear loci can be studied with relative ease, speed, and low cost. Moreover, the gene duplication in the salmonid genome resulting from ancestral tetraploidy (Allendorf and Thorgaard 1984) further increases the number of loci that can be examined with this technique in salmonids compared to most other vertebrates. In addition, the genetic basis for variation of protein loci can often be inferred directly from electrophoretic patterns because of the codominant expression of isozyme loci, the constant number of subunits for the same enzyme in different species, and consistent patterns of tissue-specific expression of different loci. Third, it is relatively easy for different laboratories to examine the same loci and use identical allelic designations so that data sets from different laboratories can be combined.

Protein electrophoresis also has several weaknesses. First, it can examine only a specific set of genes within the total genome, those that code for water-soluble enzymes. In addition, this technique cannot detect genetic changes that do not affect the amino acid sequence of a protein subunit. Thus, silent substitutions within codons or genetic changes in noncoding regions within genes cannot be detected with protein electrophoresis. Finally, this technique usually requires that multiple tissues be taken for analysis and stored in ultra-cold freezers. Thus, fish to be analyzed generally must be sacrificed and the samples must be treated with care and stored under proper refrigeration. Techniques that examine DNA directly using the polymerase chain reaction (PCR) do not require lethal sampling and can often use old (or even ancient) specimens that have not been carefully stored; for example, PCR analysis can be done on scale samples that have been stored for years at room temperature (Beckenbach 1991).

There are a multitude of techniques available for examining genetic variation directly in nuclear DNA (nDNA; Hoelzel and Dover 1991). These techniques have the advantage over protein electrophoresis of virtually unlimited power and sensitivity to examine all nuclear loci. There are, however, some disadvantages of DNA techniques relative to protein electrophoresis. The first is that DNA techniques are generally more difficult, costly, and time consuming than protein electrophoresis. However, once a specific DNA polymorphism has been detected, it may be possible to screen many individuals rapidly and at low cost (Park 1993). The second disadvantage is the difficulty of generating comparable data sets in different labs using DNA techniques. A perhaps surprising disadvantage of nDNA polymorphisms is that it is often more difficult to determine the allelic relationships of observed variability and to identify homologous loci in different species than it is with protein electrophoresis.

The analysis of mitochondrial DNA (mtDNA) has become a well-established and valuable tool for many applications in population biology (Avise 1989). Nevertheless, there are a number of serious drawbacks for its use in describing the amount of genetic variation within and among natural populations. First, the entire mtDNA genome (approximately 17,000 base pairs in salmonids) is inherited as a single genetic locus because of the absence of genetic recombination. Second, an extremely small proportion of the total genetic information in an individual is encoded in mtDNA. There are approximately 37 genes encoded in the mtDNA molecule of vertebrates, compared to an estimated 50,000-100,000 genes in the nuclear genome (Wallace 1986). Thus, in salmonids, which have a duplicated nuclear genome, less than 0.05% of all genes are encoded in mtDNA (37/100,000 = 0.00037).

Perhaps most importantly, genetic variation in mtDNA is not representative of evolutionary forces acting on nuclear DNA (nDNA) because individuals are haploid and mtDNA genotypes are transmitted through maternal lines. The effective population size for mtDNA is considerably smaller than for a nuclear gene because each individual has only one copy and because of uniparental inheritance (Birky et al. 1989). The effective population size for a mtDNA gene is equal to the number of females in an ideal population. Thus, with a 1:1 sex ratio, there are four times as many nuclear genes as mitochondrial genes. In general, therefore, the effects of genetic drift are much greater for mtDNA than for a nuclear gene in the same population. For example, only 25% of the gene diversity at nuclear genes within a population (H_S) is expected to be lost in a bottleneck of two individuals for a single generation. However, 100% of the gene diversity for the mtDNA genome will be lost in such a bottleneck.

The expected amount of allele frequency differentiation with a given amount of gene flow is also different for mitochondrial and nuclear genes because of haploidy and uniparental inheritance of mtDNA. We expect relatively more differentiation at mitochondrial genes because of the generally smaller effective number of genes and because male migration will not affect allele frequencies. With a 1:1 sex ratio and equal migration rates in males and females, we expect four times as much allele frequency differentiation (G_ST) at mitochondrial genes as at nuclear genes at equilibrium with the island model of migration (Birky et al. 1989). This difference is expected to be even greater for species in which migration rates of males are greater than that of females.

Ovenden and White (1990) have presented a rich data set of genetic variation in mtDNA and protein loci in the Australian fish Galaxias truttaceus. This is a diadromous species in Tasmania that spawns in estuaries and the larvae spend 3 months at sea before returning to fresh water. There also are landlocked populations of this species that have no marine larval stage. Ovenden and White found a tremendous amount of mtDNA variability in diadromous populations, but almost no mtDNA variation within or between four landlocked populations. They concluded that over 95% of the diversity in the mitochondrial genome was lost during a bottleneck associated with becoming landlocked. In contrast, the landlocked populations had only an 11% reduction in the amount diversity in nuclear genes.

Grant and Leslie (1993) have found that many species of vertebrates in South Africa have high amounts of genetic variation at protein loci but almost no variation in mtDNA. They argue that the patterns of gene diversity in mtDNA are likely to be misleading in species in which extinction and recolonization are important. It is likely that extinction and recolonization have played an important historical role in salmonid populations because of repeated Pleistocene glaciation events.

For example, Ferguson et al. (1991) found a large amount of genetic variation at enzyme loci both within and among four adjacent river populations of brook trout (Salvelinus fontinalis) in Newfoundland. However, a single mtDNA genotype was found in all four rivers, except one in which nearly half of the fish had a second mtDNA genotype. The relative paucity of variation in mtDNA in these fish was attributed to historical bottlenecks associated with colonization of these rivers following glaciation.

A recent paper by Karl and Avise (1992) has provided an important challenge to the utility of allozyme markers for describing historical patterns and amounts of gene flow between populations. They reported similar patterns of genetic differences for mtDNA and four nuclear DNA loci in the American oyster (Crassostrea virginica) along the east coast of North America. In contrast, allozyme studies had not detected these genetic differences among these populations. Karl and Avise concluded that the pattern observed for the mtDNA and nDNA markers reflected the historical patterns of isolation and gene flow among these populations, while this pattern is obscured in the allozymes because of "balancing selection" at the allozyme loci in this species.

Genetic techniques have also been used to determine the reproductive success of hatchery fish that are released into the wild. There is some evidence that mtDNA frequencies consistently underestimate the genetic contribution of introduced hatchery fish into native populations. Forbes and Allendorf (1991) found that the frequency of mtDNA haplotypes from introduced cutthroat trout (Oncorhynchus clarki) was significantly less than the frequency of alleles at protein loci from the introduced hatchery fish in all three populations examined. Similar results have been found by Dowling and Childs (1992) and Richard N. Williams (Clear Creek Genetics, 510 Clear Creek Dr., Meridian, ID 83642. Pers. commun., March 1993). The most likely explanation for this observation is that introduced females are relatively less successful in spawning in the wild than are introduced males.

The utility of mtDNA for describing population structure in salmonids is limited because it is a single locus that is likely not to reflect accurately the evolutionary processes (genetic drift, gene flow, and natural selection) affecting nuclear genes. Nevertheless, the same characteristics of mtDNA that make it less useful for some applications make it more useful for others. For example, the lack of recombination in the mitochondrial genome makes it extremely valuable for reconstructing phylogenies (Avise 1989). In addition, the greater amount of divergence expected among populations at mtDNA and the greater sensitivity of DNA techniques make mtDNA potentially valuable for identifying the population of origin of fish caught in a mixed-stock fishery.

Molecular genetic techniques that allow direct examination of variation in DNA present new opportunities to increase our understanding of genetic variation in natural populations. An important use of DNA techniques in the near future should be to test the generality of the results and conclusions of Karl and Avise (1992). That is, does natural selection on allozyme loci limit their usefulness to describe the historical patterns of population size and gene flow among subpopulations? In addition, the capability to detect genetic variation at a virtually unlimited number of loci with much greater sensitivity than with allozymes provides exciting opportunities to approach questions that were not possible to answer with protein electrophoresis.

There has been a tendency to consider protein electrophoresis less valuable simply because DNA techniques are newer. John Avise has put an interesting twist on this perspective (Avise 1994):

In scientific advance, timing and context are all important. Imagine for sake of argument that DNA sequencing methods had been widely employed for the past thirty years, and that only recently had protein electrophoretic approaches been introduced. No doubt a headlong rush into allozyme techniques would ensue, on justifiable grounds that: (a) the methods are cost effective and technically simple; (b) the molecular variants represent independent Mendelian polymorphisms at numerous loci scattered around the genome (rather than tightly linked variants in a single sequenced region of DNA); and (c) the amino acid replacement substitutions revealed in the protein assays might bring molecular evolutionists closer to the real "stuff" of adaptive evolution.

Protein electrophoresis is likely to remain the primary tool for many applications of genetics to the management of fish populations because of its power to examine many nuclear loci rapidly at low cost. The biggest limitation of protein electrophoresis is that sampling usually requires sacrificing the fish. Thus, PCR-based DNA techniques are extremely useful for conservation problems with small populations in which lethal sampling is not possible. Moreover, examination of nDNA and mtDNA sequence variation will continue to provide valuable and exciting additional insight into many important problems.

Allendorf, F. W., and G. Thorgaard. 1984. Polyploidy and the evolution of salmonid fishes. In B. J. Turner (editor), The evolutionary genetics of fishes, p. 1-53. Plenum Press, New York.

Avise, J. C. 1989. Gene trees and organismal histories: a phylogenetic approach to population biology. Evolution 43:1192-1208.

Avise, J. C. 1994. Molecular markers, natural history, and evolution. Plenum Press, New York, 511p.

Beckenbach, A. T. 1991. Rapid mtDNA sequence analysis of fish populations using the polymerase chain reaction (PCR). Can. J. Fish. Aquat. Sci. 48(Suppl. 1):95-98.

Birky, C. W., P. Fuerst, and T. Maruyama. 1989. Organelle gene diversity under migration, mutation and drift: equilibrium expectations, approach to equilibrium, effects of heteroplasmic cells, and comparisons to nuclear genes. Genetics 121:613-627.

Dowling, T. E., and M. R. Childs. 1992. Impact of hybridization on a threatened trout of the southwestern United States. Conserv. Biol. 6:355-364.

Ferguson, M. M., R. G. Danzmann, and J. A. Hutchings. 1991. Incongruent estimates of population differentiation among brook char, Salvelinus fontinalis, from Cape Race, Newfoundland, Canada, based upon allozyme and mitochondrial DNA variation. J. Fish Biol. 39:79-85.

Forbes, S. H., and F. W. Allendorf. 1991. Associations between mitochondrial and nuclear genotypes in cutthroat trout hybrid swarms. Evolution 45:1332-1349.

Grant, W. S., and R. W. Leslie. 1993. Effect of metapopulation structure on nuclear and organellar DNA variability in semi-arid environments of southern Africa. S. African J. Sci. 89(6):287-293.

Hoelzel, A. R., and G. A. Dover. 1991. Molecular genetic ecology. Oxford Univ. Press, U.K. 75 p.

Karl, S. A., and J. C. Avise. 1992. Balancing selection at allozyme loci in oysters: implications from nuclear RFLPs. Science 256:100-102.

Lewontin, R. C. 1991. Twenty-five years ago in genetics: electrophoresis in the development of evolutionary genetics: milestone or millstone? Genetics 128:657-662.

Nei, M. 1977. F-statistics and analysis of gene diversity in subdivided populations. Ann. Hum. Genet., Lond. 41:225-233.

Ovenden, J. R., and R. W. G. White. 1990. Mitochondrial and allozyme genetics of incipient speciation in a landlocked population of Galaxias truttaceus. Genetics 124:701-716.

Park, L. K., P. Moran, and D. A. Nickerson. 1994. Application of the oligonucleotide ligation assay to the study of chinook salmon populations from the Snake River. In L. K. Park, P. Moran, and R. S. Waples (editors), Application of DNA technology to the management of Pacific salmon: Proceedings of the Workshop, p. 91-97. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-NWFSC-17.

Wallace, D. C. 1986. Mitochondrial genes and disease. Hosp. Pract. 21:77-92.

Table of Contents