USFWS
Genetic Monitoring for Managers
Alaska

 

Types of GEM

Category II Applications

Genetic variation (heterozygosity, allelic diversity, allele frequency):

Population genetic monitoring often evaluates changes in abundance inferred through changes in genetic diversity (e.g. expected heterozygosity, Ne; allelic diversity, A), allele frequencies (e.g. temporal changes in allele frequency, Ftemporal) or Ne. Empirical studies (Leberg 1992, Spencer et al. 2000) and simulations (Luikart et al. 1999, Ramakrishnan et al. 2005)both demonstrate that A is more sensitive than is Ne for detecting a reduction in population size, because Ne is insensitive to the loss of rare alleles. Consequently, A is often the target of monitoring efforts. Monitoring Ftemporal is an even more sensitive indicator of population decline than is loss of A (Spencer et al. 2000), because changes in allele frequency can be substantial without any loss of alleles (Luikart et al. 1999). For example, studies of brown trout (Salmo trutta) in Demark have monitored Ne, A and Ftemporal for five time periods from 1944 to 1997 [40]. It was found that Ne and A were high and stable over time, but that Ftemporal varied substantially among periods. The authors concluded that high genetic diversity was maintained by gene flow in the face of small, local Ne, leading to the moderately high Ftemporal. Ne monitoring usually is based on change in allele frequencies, and is similar to tests described above using Ftemporal. However, although the above tests are sensitive to relative changes in population size, they do not quantify it.

It is important to monitor genetic changes within captive populations, as the loss of genetic variation from the use of too few founders or faulty breeding protocols can compromise recovery efforts [74]. Dowling et al. [75] monitored gene frequencies for seven years in samples of repatriated (wild-produced larvae reared in captivity and re-released as juveniles) razorback sucker (Xyrauchen texanus) in Arizona to determine whether the captive-rearing program was transmitting sufficient genetic variation to the endangered wild population. Their results suggest that, to date, the program has been successful in avoiding use of progeny from only a small fraction of wild spawners. By contrast, monitoring of the hatchery population of the endangered Rio Grande silvery minnow Hybognathus amarus between 2001 and 2003 has consistently shown heterozygosity levels equal to, but allelic diversity much lower than, that of the wild population [76], a classic sign of a population bottleneck.

Genetic monitoring can be an efficient way of determining whether captive individuals are recruited into wild populations [77], and it is the most reliable method to determine whether they are making a reproductive contribution to subsequent generations. Hansen [78] examined long-term impacts of intense stocking on brown trout Salmo trutta in two Danish populations by comparing genetic profiles from contemporary samples with those from archived fish scales; one population changed dramatically over time owing to stocking, whereas the other showed little genetic contribution of stocked fish.

Example

Population mixtures (proportion of individuals originating from differentiated breeding groups):

Mixed-stock analysis estimates the proportions of individuals in a mixture that originate from each of two or more genetically differentiated breeding groups. Although mixed stock analysis (and detection of hybrids) can be based on diagnostic assays (Category Ib), it is often necessary to use probabilistic approaches based on allele or genotype frequencies. Genetic tools have been used for many years to assess the composition of mixed stock fisheries [55,56], but recent laboratory and statistical advances provide increased power for real-time genetic monitoring. Fraser River sockeye salmon Oncorhynchus nerka are the most valuable commercial salmon fishery in British Columbia, but efficient harvest of the summer run is complicated by conservation concerns for late-returning populations. In 2002, a large monitoring program (up to 600 individuals several times a week for two months) based on microsatellites and major histocompatibility (MHC) loci provided stock composition estimates within 9-30 h [57], enabling managers to avoid overharvest of the late run. Population admixtures also can be quantified and monitored with methods that estimate the degree of mixed ancestry of each individual. However, the power and resolution are generally much lower when the source populations are characterized by allele frequency differences, rather than by fixed differences in diagnostic markers ([56] as in Category Ib).

Example

Effective population size (Ne):

Estimating Ne enables direct tests for changes in population size. As two samples are needed for a single temporal estimate of Ne, monitoring changes in Ne via the temporal method requires samples from at least three time periods. Ne has been estimated for brown bears in Yellowstone National Park by analyzing samples from the 1910s, 1960s and 1990s [41]. Ne estimates were _85 for both time periods (1910-1960s and 1960s-1990s), providing no evidence of a recent population decline. Other studies have also used a combination of contemporary and historical samples to obtain multiple temporal estimates of Ne [42,43]. Palm et al. [44] monitored allele frequencies for 20 years in two Swedish brown trout populations and used a modified temporal method to obtain annual estimates of Ne. They found consistently small estimates, but no temporal trend in Ne. Ne monitoring can also be based on gametic disequilibrium, requiring only one sample for each Ne estimate. Unbiased gametic disequilibrium Ne estimators have been available only recently [45,46]. The precision and reliability of this method have not yet been thoroughly quantified, and further research is needed. However, the ability to estimate Ne based on a single sample should improve the power to estimate any trends.

Example

Population structure and migration (gene flow):

Effective conservation often depends on the identification of management units and timely information regarding the effects of natural and anthropogenic factors on movement and gene flow between these units [47]. Although several genetic methods yield point estimates of gene flow, a monitoring program that produces a temporal series of samples can provide richer insights [48,49]. Furthermore, monitoring changes in gene flow indices, such as FST, can detect changes in differentiation among populations [50]. For example, genetic monitoring of leopard frog Rana pipiens populations revealed that genetic structure was stable over 11-15 generations [51]. A study of cod Gadus morhua populations from the Baltic and North Sea using both historical and contemporary samples also found high temporal stability in FST over 47 and 89 years, respectively [35]. However, other studies have found the opposite result. Genetic monitoring of Scottish red deer Cervus elaphus demonstrated that fine-scale genetic structure in females declined at a steady rate over a 24-year period [52]. Based on demographic data collected over the same time period, the authors concluded that this decline was due to a combination of increasing population size and decreases in polygyny (multiple females breeding with a single male).

One of the most fruitful uses of genetic monitoring is likely to be the quantification of changes to movement patterns in response to events that disrupt historical patterns of connectivity, such as habitat fragmentation. Genetic monitoring is already a useful tool for evaluating the cumulative effects of habitat fragmentation. An allozyme and microsatellite-based study of California valley oak Quercus lobata pollen movement suggests that there was a decline in the effective number of fathers contributing pollen to the next generation between 1944 and 1999 [53]. The authors propose that this was the result of progressive stand thinning and that it might lead to future reproductive failure owing to genetic isolation. In a study of Pyne's ground plum Astragalus bibullatus, the genetic diversity of allozymes in multiple stratigraphic deposits was examined to discern the effects of cedar glade fragmentation [54]. Seeds from the top soil layer had higher levels of differentiation among sites than did the two lower layers, a result consistent with recent habitat loss.

Example