Conserving quantitative genetic variation

Monitoring the potential for adaptation to environmental change might be well served by an explicit focus on monitoring changes in quantitative environmental variation, particularly the component that is available for evolution, namely, heritable genetic variation.

However, studies that measure overall levels of genetic variation within and among populations do not provide any information on genetic variation of fitness traits that are under selection in the field. The focus on maintaining variability among such populations raises the real threat that the adaptive potential of larger but less distinct populations is being minimised. A rapidly changing environment where the maximisation of genetic variation for traits associated with stress and disease resistance will be critical for long-term persistence.

Alternate approaches for maintaining appropriate genetic variation in populations have been advocated. These include shifting the focus to phenotypic variation of traits associated with stress and disease resistance that will have the effect of selection of favourable genes at many loci (Woods et al. 1999). Another approach is to concentrate on populations that persist at ecological margins as these will have been selected for environmental stress resistance and are likely to have higher frequencies of stress resistance genes than population inhabiting more favourable environments (Hoffman & Parsons 1997). Marginal populations may have evolved generalised stress resistance mechanisms that will allow them greater resistance in stress environments other than those in which they have been selected.

Management strategies should aim to conserve species across a broad range of climatic regions and to conserve all races, variants and subspecies. This will ensure that any genotypes fixed because of local adaptation will be conserved and available to counter future climatic changes.

Measuring genetic variation for State of the Environment reporting

Allelic richness: Allelic richness is defined as the number of alleles in a sample, standardised for sample size. Two statistics are relevant to allelic richness, the percentage of polymorphic loci (P), and the average number of alleles per locus (A). To compare allelic richness between species, or to compare samples from the same species at different times, it would be necessary to sample the same markers on each occasion, or to take a random sample of all possible markers. But methodological conventions are to survey available markers and to select informative ones for analysis and reporting. Selection is often haphazard but is rarely random (Glaubitz et al. 1999). In addition, loci are treated as polymorphic only if the frequency of the most frequent allele is less than 0.95, or sometimes less than 0.99 or 0.90. Typically, the number and identity of monomorphic markers are not published, sometimes by editorial policy, if not by convention.

It is also possible to estimate the number of alleles per polymorphic locus, which may be readily compared across studies. In addition, planned and coordinated monitoring strategies could overcome biases through appropriate marker choice.

Gene diversity: Gene diversity, H_e, is the expected level of heterozygosity in a randomly mating population, given observed allelic frequencies. In species with subdivisions (populations), the diversity may be partitioned into within-population (H_s) and between-population (G_ST) components. This information may be used to examine patterns of population structure and differentiation and non-random mating.

The level and type of information obtained from genetic studies depends on the markers chosen. Gene diversity is directly available in studies that use codominant markers (including restriction fragment length polymorphisms, RFLPs, isozymes and microsatellites). Studies using dominant markers (e.g. polymerase chain reaction-based markers including amplified fragment length polymorphisms and random amplified polymorphic DNA, RAPD) or haploid markers (e.g. mitochondrial DNA) usually report genetic distances. These studies do not provide any direct measure of gene frequencies. Mitochondrial DNA gives a measure of diversity equivalent to H_e (haplotype diversity) but no measure of inbreeding. This occurs because the various methods used to score genetic markers detect different proportions of the total variation in DNA base sequences. Moreover, the rates of mutation vary widely between different classes of marker. For example, isozymes and RFLPs have low mutation rates and microsatellites have very high mutation rates (Ennos 1996).

Observed heterozygosity: Observed heterozygosity, H_o, is the proportion of heterozygotes in a population, averaged over all loci, and is theoretically available in all studies that use codominant markers. However, this statistic is rarely reported in the literature. Most studies are concerned primarily with differentiation between populations in a single study, so that comparisons between times and between species are rarely considered. Microsatellite studies contrast with studies that use other codominant markers. Most microsatellite studies report both H_o and H_e, allowing computation of inbreeding coefficients (e.g. Baker et al. 2000; Lebas & Spencer 2000; Miwa et al. 2000).

Monitoring Round-leaf Honeysuckle (Lambertia orbifolia)

Coates and Hamley (1999) measured the genetic variation within and between populations of Lambertia orbifolia, a large woody shrub restricted to seven populations in Western Australia. They found 12 out of 19 isozyme loci were polymorphic. Levels of genetic variation were roughly comparable with other long-lived woody shrub species endemic to the region. Genetic divergence between population groups was very high, with 44% of variation in genetic diversity being composed of between population differences. They estimated rates of gene flow between the populations to be very low and found consistently low levels of outcrossing. Coates and Hamley (1999) recommended that one population be recognised as a separate conservation unit, on the basis of the degree of genetic differentiation. They reported A, P, H_eand H_o, together with standard errors for each statistic, for each of the populations, making it possible to resample these populations in future years, to measure trends in diversity.