Seattle, Washington This panel was convened to discuss applications of genetic technology to problems in salmon biology and management. The panelists included:. In the research presentations given earlier in the workshop, the speakers described a number of ways in which DNA technology can be used in salmon research.
In the session on management applications, the speakers raised several genetic issues of concern to salmon managers. For this discussion, the panelists were asked to consider these different issues, using one question as a guiding framework: How can DNA and other genetic technologies be applied to problems in salmon management? Three general categories of genetic concerns can be identified: genetic variability among populations, or population identity; genetic variability within populations; and genetic change through selection.
These areas are not always distinct in practice, but they are nevertheless useful as organizing constructs for summarizing the discussion. Although the workshop's focus was on DNA technology, the panelists also considered protein electrophoresis in the discussion, because it has played and continues to play such a prominent role in the identification of salmon gene pools for management and conservation. In addition, the discussion included applications of quantitative genetic analyses to problems in salmon management.
Molecular genetic analyses are well suited to issues related to population structure, migration, and gene flow. In evaluating the capability of DNA or protein analyses to address these issues, the advantages and disadvantages of the different analyses should be weighed. A key question is whether the potential for increased genetic resolution through DNA analysis will be realized, and whether this approach will enhance the ability to effectively manage natural salmon populations.
Underlying both protein and DNA analyses is a general assumption that the molecular genetic variation they detect is, on the whole, selectively neutral. The two approaches differ in other notable respects, however. Protein electrophoresis detects portions of the genome that code for functional biochemical products. Protein analysis therefore surveys a small, perhaps unrepresentative, fraction of the genome, in spite of recurrent assertions to the contrary in the literature see Lewontin , Watt Furthermore, a large contribution to genetic variation can result from silent nucleotide substitutions that are not detectable through protein analysis.
The fact that only functional genes can be detected may or may not be an advantage, depending on the purpose of the survey. By contrast, analysis of DNA detects genetic variation at its most fundamental level, the nucleotide sequence. Thus, DNA analysis potentially allows examination of all nucleotide sequences in the genome, including those that are not translated into protein products or that have no known function.
By virtue of providing more comprehensive coverage of the genome, DNA analysis potentially affords greater ability to resolve slight genetic differences between populations and, in some cases, to distinguish between individuals. Unlike the situation with allozymes, however, it may be unclear in DNA analysis how genetic variation corresponds to genotypic variation for coding portions of the genome, since a functional product is often not detectable.
Molecular Genetic Approaches in Conservation by Thomas B. Smith, Robert K. Wayne
Thus, although analyses of protein and DNA variation share some similarities, their differences suggest that, to some extent at least, these analyses are complementary. The need for more extensive allelic and genotypic data provided by protein electrophoresis to resolve current questions in fisheries management and the potential for DNA technologies to contribute to this resolution suggest that the procedures should be coordinated in a manner similar to the joint application of light and electron microscopy to problems of cellular and subcellular structure F.
Utter, Univ. Washington, pers. It is worth remembering when evaluating different molecular analyses that protein electrophoresis is a proven and generally effective technique for identifying gene pools in Pacific salmon, and extensive data on allozyme variation are available from a wide range of populations.
In whatever capacity DNA methods are applied as genetic tools to address salmon management problems, it would be imprudent to abandon allozyme analyses at this stage. It is nevertheless important to explore the potential that DNA analyses offer for salmon management. Most population-level DNA analyses can be organized into two broad classes that are defined by the type of DNA they detect. Variation in mtDNA has been widely analyzed in evolutionary and conservation biology, in part because mtDNA is capable of rapid evolution.
Unlike nDNA, mtDNA is a small, haploid molecule that is inherited maternally, is composed almost entirely of coding sequence, and is free from recombination; it is also easy to isolate. These features make mtDNA useful for phylogenetic reconstruction and for detection of genetic bottlenecks. The main disadvantage of mtDNA analysis is that this molecule represents a single genetic unit for which the markers are all absolutely linked.
Analytical techniques for mtDNA include indirect methods such as the detection of genetic variation via the analysis of restriction fragment length polymorphism RFLP , as well as the direct analysis of mtDNA sequences. The analysis of nDNA provides some advantages over the analysis of mtDNA for purposes that require greater discrimination among individuals, primarily because of nDNA's larger size, greater variability in particular classes of sequences, and recombination permitting variants at different sites to assort independently.
Microsatellites are short, highly polymorphic DNA sequences in which variation is manifested by differences in the number of adjacent sequence copies of a simple-sequence repeat see Bentzen et al. The analysis of microsatellites is a recent development in fisheries that shows considerable promise for genetic discrimination of closely related groups, such as salmon subpopulations with different run timing.
Microsatellites are also sometimes referred to as VNTRs. The development of the polymerase chain reaction PCR to amplify DNA sequences up to several million times has been a major breakthrough because it eliminates the need to clone each sequence separately. These promising techniques have proved useful in human genetics and for research involving many commercially important species, but are only now being widely used to survey genetic variation in salmonids. The relative capabilities of these and other molecular approaches to genetic analysis are summarized in Table 1.
Table 1. Comparison of some features of allozyme and DNA analyses.
It is important to recognize that the ratings in the cells of the matrix may vary widely in specific cases; this assessment is intended only to provide a general sense of the relative capabilities of these techniques. Additional research and development are necessary to develop sets of reliable DNA markers before these techniques can be widely applied to problems in salmon management. DNA analyses nevertheless merit consideration as tools for the genetic characterization of salmon populations.
One advantage of many DNA analyses over protein analysis is that, once appropriate markers or primers have been developed, nonlethal sampling is possible. Small amounts of tissue e. This can be an important feature when identifying populations for conservation or while evaluating genetic change in protected or declining populations. DNA analyses are already being utilized for the genetic monitoring of captive populations, such as endangered Snake River sockeye salmon Oncorhynchus nerka , because sampling does not require sacrifice of individuals.
A useful aspect of some DNA analyses that use PCR is that preserved specimens can provide genetic material amenable to DNA analysis again assuming that appropriate markers are available. The potential access to the DNA of museum specimens is appealing, because it may provide genetic information over extensive temporal as well as spatial scales of variation, especially from populations that no longer exist. Although this capability cannot reconstruct the ecological and behavioral diversity that accompanied these preserved specimens, it could add information to temporal genetic surveys that may be valuable in conserving existing lineages.
Taken together, these characteristics suggest valuable monitoring capabilities that incur fewer costs to populations, particularly those that are small, exploited, or declining. The transition from genotype to phenotype is an complex one and, to date, molecular genetic analyses have shed relatively little light on the mechanisms of adaptive evolution. The field of quantitative genetics, which has existed in essentially its current form for half a century, provides several indirect genetic methods to explain how evolution operates to produce phenotypic change, at least over the short term Lande These methods use the statistical properties of metric characters, which are typically polygenic, to describe the composite effects of the underlying genes.
Although the quantitative genetic approach does not observe the genotype directly, and although the effects of single genes are generally invisible to quantitative genetic analyses, these methods help to bridge the gap between molecule and phenotype and provide the evolutionary context for molecular genetic variation Barker and Thomas ; Barton and Turelli The main limitations of the quantitative genetic approach are that it typically requires extensive manipulations of individuals of known relationship, and that primary genetic processes, such as gene substitutions, are generally undetected Falconer Modern molecular analyses have the potential to reveal these processes by identifying changes at specific sites in the genome.
Unfortunately, many specific applications of these analyses to management problems remain far from clear, and rapid progress must be made on the integration of quantitative and molecular approaches to fill this void. Given the potential of DNA analyses and the recognition of their limitations in addressing problems in salmon management, what are the most profitable lines of inquiry to pursue?
Among the most prominent problems in genetic management are the identification and conservation of distinct populations. The often extensive anadromous migrations of salmon, coupled with the fidelity of adults to their freshwater spawning and rearing grounds, mean that fish from different populations may commingle for extensive periods, during which time harvest usually occurs. Determination of the genetic composition of such population mixtures is a major focus of salmon management, and the analysis of allozyme variation has several advantages for this purpose. First, it is relatively inexpensive and easy to perform, and well-established protocols are available for salmon Aebersold et al.
Second, the extensive allozyme surveys that have already been made form a broad basis for comparison of populations on both temporal and spatial scales. Third, genes with homologous functions from diverse groups can be compared. Consequently, allozyme electrophoresis is likely to remain an essential feature of genetic management in fisheries for the foreseeable future. Despite their sensitivity to environmental variation, other techniques used for stock identification and discrimination, such as scale pattern and microelement analysis, parasites, and freshwater age, are also valuable in this regard and should not automatically be discarded in favor of new and more sophisticated technologies.
Nevertheless, analyses of allozyme variation and morphometric variation do not always provide the desired resolution; DNA analyses may. One important application of molecular genetic technology to the problem of population identity is in detecting introgression of hatchery salmon genes into wild salmon stocks. The success of protein electrophoresis for this purpose has been mixed, especially in comparisons of populations exhibiting little allozyme variability.
DNA analysis has not yet been widely applied to this problem, but analysis of highly variable DNA sequences may provide greater power to resolve low levels of introgression, especially between similar stocks. Another major problem in the genetic management of salmonids is the estimation of genetic variability within populations. Investigation of this topic has a long history, and interest in it has been enhanced by growing awareness of the importance of conserving genetic variability in natural populations.
For 25 years, protein electrophoresis has been widely applied to the estimation of genetic diversity in salmon populations. Table 1. Statistics of genetic diversity across 15 SSR loci in the white lupin landrace accessions and the out-group genotypes. Table 2. Detail of allele types identified and landraces containing private alleles as revealed by 15 SSR loci among the white lupin landrace accessions. Table 3. Summary of the population diversity indices averaged over 15 loci.
Table 4. Fig 2.
- Polymorphism of Mhc genes: implications for conservation genetics of vertebrates | Edwards Lab.
- References cited?
- Our Car as Power Plant.
- Flying Guns of World War I.
- Entrepreneurship the New Rules.
- Policing Futures: The Police, Law Enforcement and the Twenty-First Century?
- Conservation and molecular methods | SpringerLink!
UPGMA dendrogram showing the genetic relationships among the Ethiopian white lupin landrace accessions and the out-group genotypes. Table 5. Clustering patterns of Ethiopian white lupin landraces from different origins over the clusters. Fig 3. UPGMA dendrogram showing the genetic relationships among the Ethiopian white lupin landrace populations collection areas.
Table 6. Description of Ethiopian white lupin landrace accessions proposed to constituent national core collection for Ethiopian white lupin accessions. Table 7. Fig 5. Model based population structure of Ethiopian white lupin landraces based on 15 SSR panels as obtained from Structure2. Table 8. Proportion of membership of each predefined population, in each of the clusters obtained at the best k, i.
Discussion 4. Conclusion The study showed that genotyping combined with clustering and population structure analysis is a powerful method for characterizing germplasm. Supporting information. S1 Table. Description of the Ethiopian white lupin landrace accessions and the out-groups used for the study. S2 Table. List of SSR markers used for the study. S3 Table. The Evanno table output. References 1. Kurlovich BS. Publishing House "Intan", St.
Petersburg, Russia. Dunn DB, Gillett J. Lupines of Canada and Alaska. Hondelmann W. The lupin—ancient and modern crop plant. Theor Appl Genet. Lupinus IN: Kole C, editor. Wild crop relatives: genomic and breeding resources. Springer Berlin Heidelberg. Lupin kernel fiber foods improve bowel function and beneficially modify some putative faecal risk factors for colon cancer in men. Arnoldi A. Hall RS. Australian sweet lupin flour addition reduced the glycaemic index of a white bread breakfast without affecting palatability in healthy human volunteers.
Asia Pac J Clin Nutr. Production of Lupinus angustifolius protein hydrolysates with improved functional properties. View Article Google Scholar 9. White lupin Lupinus albus L. The Importance of legumes in the Ethiopian farming system and overall economy: An overview. Am J Exp Agric. View Article Google Scholar Annl Bot.
Developing an appropriate strategy to assess genetic variability in plant germplasm collections.
Genetic diversity and core collection evaluations in common wheat germplasm from the northwestern spring wheat region in China. Mol Breed. Diversity in white Lupin Lupinus albus L. Genet Resour Crop Ev. Field Crops Res. Plant Breed. Genetic diversity of wild soybean Glycine soja and Japanese cultivated soybeans [ G. Genet Resour Crop Evol. Genetic diversity of the mung bean Vigna radiata L. Aust J Bot. Genetic diversity and population structure in lentil Lens culinaris Medik. Acta Agron Sin. Genetic diversity and core collection of alien Pisum sativum L.
Genetic diversity analysis of faba bean Vicia faba L. Agric Sci China. Characterization and genetic diversity analysis of selected chickpea cultivars of nine countries using simple sequence repeat SSR markers. Crop Pasture Sci. Open J Genet. Selection of a representative core collection from the Chilean common bean germplasm. Chilean J Agric Res. Gepts P.
A phylogenetic and genomic analysis of crop germplasm: necessary condition for its rational conservation and use. IN: Gustafson JP, editor. Diversity and origin of Andean landraces of common bean. Crop Sci. A core collection of common bean from the Iberian Peninsula. Identification of anthracnose resistance in Lupinus albus L. Metabolic changes associated with cluster root development in white lupin Lupinus albus L. Extent and pattern of genetic diversity in Ethiopian white lupin landraces for agronomical and phenological traits. Afr Crop Sci J.
Afr J Biotechnol. Genotype by trait biplot analysis to study associations and profiles of Ethiopian white lupin Lupinus albus L. The first genetic and comparative map of white lupin Lupinus albus L. DNA Res. PowerMarker: an integrated analysis environment for genetic marker analysis. Peakall R, Smouse PE. GenAlEx 6. Population genetic software for teaching and research—an update. Accuracy of estimated phylogenetic trees from molecular data II. Gene frequency data. J Mol Evol.
Development of an algorithm identifying maximally diverse core collections. Quality of core collections for effective utilization of genetic resources review, discussion and interpretation. DARwin software. Wright S. The genetical structure of populations. Ann Eugen. Previous work has demonstrated that sage-grouse chick DNA can be isolated from eggshells [ 12 ] and that avian predator DNA can be isolated from sea bird eggshells and carcasses [ 13 ].
Thus, we expanded on the single locus method employed by Steffens et al. We expect to further substantiate the use of molecular forensics for avian conservation and management as it allows the identification of depredating species and can provide managers with information on cause-specific mortality during important reproductive time periods. Sage-grouse currently occupy a significant portion of the sagebrush steppe throughout much of the state of Wyoming. The three factors with the greatest influence on sage-grouse population growth rates are hen survival, nest success, and recruitment; each can be heavily impacted by predation [ 17 — 19 ].
This study was conducted in the northwest portion of Bighorn Basin, Wyoming. We monitored hens to confirm survival and their location on nests. When hen mortality was detected we collected carcasses the day of detection with a mean detection of 3. We used cotton swabs Fisher Scientific, USA wetted with a few drops 1—3 of ultra-pure water to swab carcasses and eggshells. For each individual egg, we used one swab on the inside and one swab on the outside of the egg. We used a single swab for each carcass and targeted regions on feathers that appeared matted from saliva and or had bite marks.
All DNA extractions were completed in a room and biosafety cabinet with reagents and laboratory supplies dedicated to non-invasive extractions and included extraction blanks reagents only for each extraction to monitor contamination. However, in some cases this gene fragment failed to amplify and lacked variation to differentiate Canis species. We subjected each extraction blank to PCR and included negative controls for each reaction to monitor contamination. When contamination was detected, we reran the PCR once with fresh aliquots of reagents to determine if contamination occurred in the PCR, extraction, or field collection.
If the predator species was determined to be coyote Canis latrans we then amplified a second set of 10 microsatellite loci [Set B; 25 ] to identify individual animals and compare genotypes obtained from nests and carcasses and to those collected from feces and coyotes removed from sage-grouse leks during the course of this study [ 20 ]. The reason we employed two sets of microsatellite loci is that our laboratory already had developed a canid identification database based on set A but Orning [ 20 ] used set B for individual genotypes. Thus, to determine both species and individual identification we needed to use both sets.
All microsatellite reverse primers were PIG-tailed to facilitate accurate genotyping [ 26 ]. Each PCR was run three times to estimate allelic dropout or false alleles [ 27 ]. We only scored an allele if it was present in at least two of the PCR replicates [ 28 ]. False alleles and allelic dropout were estimated with gimlet v1.
Species identification of canid genotypes Set A was determined using the Bayesian clustering algorithm in structure v2. Piaggio; unpublished data. The genotypes from the eggs were compared to this database using structure with the admixture and allele frequencies correlated models, burn-in of 70,, and MCMC length of , The analysis was run with k set at six and replicated five times.
Orning [ 20 ] estimated the number of coyotes around sage-grouse leks and genotyped lethally removed coyotes and fecal samples collected along transects. We attempted to obtain genotypes from eggs and carcasses to identify individual coyotes from the area using microsatellite Set B.
We used the genotype match function in genalex v6. A total of 14 depredated nests were discovered and we recovered an average depredated egg count of 4. Using the genetic methods described herein, we successfully amplified mammalian mtDNA from 11 of 14 nests The average number of eggs per nest from which we obtained predator identification was 2.
We also opportunistically collected 7 hen carcasses. We did encounter problems with human contamination that seemed to primarily source from field collection but there were two instances where it originated in the lab. The set A allelic dropout rate across loci was 0. Two of the five nest genotypes N4 and N5 confirmed mtDNA results and identified the predator as coyote. Microsatellite set A also helped us refine the species identification for one nest N8 , which was a coyote, where the mtDNA could not distinguish between wolf or coyote. Neither the microsatellites nor the mtDNA could resolve whether the nest predator for N7 was a coyote or wolf, thus we classified it as a wild canid.
This was the same nest from where we also extracted and amplified deer mouse mtDNA from one egg. The single canid genotype we obtained from N11 further identified that a coyote was also in contact with this nest. Overall, we identified coyotes from six nests, dogs from three nests, and one we could not distinguish whether wolf or coyote, thus we called it a wild canid likely coyote as wolves are not common in this area [ 33 ]. The set B allelic dropout rate across loci was 0.
Coyotes which were removed and for which genotypes were obtained were on average within Coyotes in Wyoming have large annual home ranges These were compared to 27 tissue genotypes and 28 fecal genotypes from Orning [ 20 ]. We were unable to match any genotypes from eggs and carcasses to captured coyotes or fecal samples collected in the study area.
None of the nest or carcass predator genotypes matched each other. The goal of this study was to test the concept that non-invasive genetic sampling can be used as a forensic tool to identify predators of ground-nesting birds. In the majority of the cases the species was a canid, but we encountered two events, one nest and one carcass, where we were unable to resolve the canid species identification e. Both cases were most likely coyote because wolves have not been documented in the study area [ 33 ].
We could not distinguish the canid species because some mtDNA haplotypes found in wolves in eastern North America and coyotes are nearly identical, which is thought to be a result of historic hybridization or incomplete lineage sorting [ 36 ]. The molecular method we employed was successful in supplementing field and camera identifications of nest predators by either confirming or providing identification when other methods proved inconclusive.
There were only two disagreements between physical evidence and our results, in one case the field identification suggested bird or bobcat Lynx rufus as the predator but we identified cow, and in the second case the field identification was cow but we identified coyote. The camera data and the molecular identifications unequivocally agreed in four of the cases and all of these were coyote.
In one case the camera captured a close-up of the face of a potential predator but from the photo we could not differentiate whether it was a striped skunk or an American badger Taxidea taxus. In this case the molecular data clarified the camera data by determining that it was a striped skunk. There was a disagreement in only one case where the molecular data identified domestic dog DNA but the camera captured a photo of a weasel Mustela sp.
Possible reasons for disagreements between datasets are likely due to difficulties with predator species identification in the field from nest remains, non-mammalian predators, camera failure, or DNA isolated from a scavenger instead of the nest predator. Therefore, we recommend combining molecular data and cameras to increase the success in identifying mammalian predators of ground-nesting birds, similar to the approach by Steffens et al. Combining these two techniques provides valuable insight for management decisions to facilitate protection of threatened and endangered species.
All of the species we detected have been previously documented as nest predators [ 37 , 38 ]. The literature contains multiple reports of both coyotes and striped skunks feeding upon eggs of ground-nesting birds [ 39 , 40 ]. The high rate of coyote depredation detected in this study was not surprising given that this species is known to be abundant in the study areas and they have been documented eating sage-grouse eggs [ 8 , 20 ]. Even though dogs are not considered a common predator of sage-grouse nests [ 8 , 37 , Orning and Young, in review], we detected DNA evidence of domestic dogs from three depredated nests.
Domestic dogs have been documented disturbing nests of both ground-nesting birds and sea turtles [ 41 , 42 ], but not sage-grouse eggs. These nests were not close to human dwellings, the area is 30—40 miles from a populated area, but human activities that included the presence of dogs, such as oil and gas development, recreation, and livestock, were regularly observed in the area of the nesting sage-grouse E.
Orning; personal observation. However, the small sample size of this study may have artificially amplified the apparent impacts that dogs have on sage-grouse nests. Thus the dog results obtained herein should be interpreted with caution when considering management strategies. One approach that could easily reduce the potential for human-caused losses of this nature would be to increase public awareness of sage-grouse nesting by limiting human and pet access to areas during critical nesting periods. Deer mice, in particular native Peromyscus spp.
We also documented cow DNA from one nest. Cattle are possible ground-nesting bird egg predators [ 44 ]. However, cattle activity and feces were in close proximity to some of the nests and Orning [ 20 ] collected video of cattle investigating sage-grouse nests so whether this event was predation or contamination is unclear.
Finally, we detected human DNA from a few of the nests. The sensitivity of the general mammalian primers used for this study must be taken into account when conducting a molecular forensic study.
Conservation and molecular methods
Wildlife genetics laboratories should also apply the same stringent protocols of sample collection and processing required in the human forensics field [ 45 ]. Further, approaches can be applied in the laboratory to prevent amplification of human DNA such as species-specific PCR primers [ 5 ], human-blocking PCR primers [ 46 ], and metabarcoding with high-throughput sequencing technology. The most likely explanation is that the DNA had degraded beyond our detection ability. If the eggs were collected too long after the depredation event, and depending on weather conditions, the DNA could degrade quickly [ 47 ].
Increasing the number of visitations to nests to shorten the time intervals between predation and discovery is unlikely to be a feasible approach to limit DNA degradation. Intensified human disturbance can increase the chances of nest abandonment or predation which would be counterproductive to conservation goals [ 7 , 48 ]. One way to increase the success rate of detecting degraded DNA would be to target a smaller fragment. However, this approach has drawbacks because resolution can be lost for differentiating recently derived species when using shorter fragments of DNA but this shortcoming could be addressed by using genes with higher mutation rates.
For example, we increased the number of PCR cycles to 52 for microsatellite Set B which allowed us to obtain a higher percentage of full genotypes. However this high number of cycles could increase the chance of false alleles. In fact, any optimization strategy could increase non-specific amplification which could decrease the accuracy of species identification. Another explanation for failure to detect the predator species might be that the nest predator was avian. The primers we used were mammal-specific, thus we were not able to amplify avian DNA.
We did have one case where field reports suggested the nest predator was a raven Corvus corvax or a snake and another where the camera identified a raven around the nest, but could we did not obtain molecular species identification from either nest. Using this method to determine if the predator was a reptile i. We foresee multiple continuations of this study to increase the thoroughness and robustness of predator species identification.
The first and most obvious is to increase the sample size. We acknowledge that the sample size in this study is quite small, but our goal was to prove the concept rather than thoroughly quantify the diversity of predators on sage-grouse eggs and adults. To rigorously estimate nest predation rates, and provide a control for identifying predators of sage-grouse adults, which was lacking from this study, there would need to be a larger study with more cameras placed around leks and on more leks throughout the range of the species.
Another follow-up study would be to test DNA degradation rates by having captive predators deposit saliva on eggshells and carcasses. This would allow one to establish a reasonable time since depredation and evaluate the accuracy of predator species identification from adult carcasses as these are usually opportunistic samples and filming these depredation events is unlikely. To increase the breadth of predator taxa identified from nest remains and carcasses one could apply primers that amplify avian DNA. Avian predator DNA has been successfully sampled from black-fronted tern Chlidonias albostriatus eggshells [ 13 ], so this approach may also be feasible for sage-grouse.
The challenge of this approach is that DNA from the prey species could also be amplified which would obscure predator identification as seen in Steffens et al.