GENETIC HISTORY OF CHINOOK AND SOCKEYE SALMON
ANALYZED USING ANCIENT AND CONTEMPORARY
MITOCHONDRIAL DNA
By
BOBBI MAY JOHNSON
A dissertation submitted in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
WASHINGTON STATE UNIVERSITY School of Biological Sciences
DECEMBER 2016
© Copyright by BOBBI MAY JOHNSON, 2016 All Rights Reserved
© Copyright by BOBBI MAY JOHNSON, 2016 All Rights Reserved
To the Faculty of Washington State University:
The members of the Committee appointed to examine the dissertation of BOBBI MAY
JOHNSON find it satisfactory and recommend that it be accepted.
______Patrick A. Carter, Ph.D., Chair
______Gary H. Thorgaard, Ph.D., Co-Chair
______Brian M. Kemp, Ph.D.
______Richard Gomulkiewicz, Ph.D.
______Colin Grier, Ph.D.
ii ACKNOWLEDGEMENTS
This dissertation could not have been completed without the support of many people over the years. I offer my most heartfelt gratitude to the people listed here. To anyone I have forgotten to list, I sincerely apologize and thank you as well.
I would first like to thank my advisor, Dr. Gary Thorgaard. It would be a great challenge to find a more enthusiast advisor. Gary provided valuable insight and much needed encouragement along the way (along with some great fishing tips) and I am honored to have been his student. I would also like to thank all those who served on my committee with Gary:
Brian Kemp, Colin Grier, Dick Gomulkiewicz, Mary Collins, and Pat Carter. Special thanks to
Brian, who taught me to think about laboratory work in a new way and greatly expanded my musical repertoire. Thanks also to Dick who provided valuable feedback over the many years of this project and consistently challenged me to think critically about the research questions.
I am especially grateful for the work of the archeologists and curators who meticulously gathered, cataloged, and provided context for the ancient samples. Also, to the many people who assisted directly in our efforts to secure and use these collections including: Diane
Curewitz, Guy Moura, John Matt, Mary Collins, Sarah Walker, Shannon Tushingham, and Stan
Gough. Special thanks to Mary Collins and Sarah Walker who also provided the ages for the ancient samples. I thank the Confederated Tribes of the Colville Reservation as well as the
Spokane Tribe of Indians for allowing us to utilize the sample collections. I also thank the agencies who provided non-bone and contemporary samples (listed in the dissertation). Thanks
iii to Kelly Cassidy for suggesting lid paper as a potential source of contamination and for providing thoughtful answers to my inquiries related to contamination of museum specimens.
I thank the many undergraduates who worked on this project, all as volunteers. Special thanks to John Monda, Lauryn Watkins, and Matt Day who contributed greatly. Thanks to Cara
Monroe for assistance in the ancient DNA laboratory and to Joe Brunelli for teaching me how to stretch a research budget to maximum capacity. Thank you to Lindsay Hilldorfer who worked on the very difficult Spokane samples with me. I am also grateful for those who provided editorial comments to the many drafts that eventually became this dissertation. A special thank you to Mike Garvin for his helpful review of the sockeye study. I am grateful for the assistance of Fernando Villanea in the development of the extended Bayesian skyline plots and Kris
Christensen who transformed my thoughts into a working script for the drift simulation.
On a personal note, I would like to recognize Dave Yost, who helped me find my way many years ago as well as Dave Willis and Katie Bertrand who introduced me to research and encouraged me to go to graduate school. Special thanks to my family; no one will ever mistake us for conventional, but I would not have it any other way. I would also like to thank my dear friends Anna Heink, Kris Christensen, and Lindsay Hilldorfer. I hate to imagine what it would have been like without you.
Finally, I would like to thank my husband McLain who provided feedback on dozens of grant proposals and manuscript drafts and who patiently spent many years living apart while I completed my degree. Thank you.
iv GENETIC HISTORY OF CHINOOK AND SOCKEYE SALMON
ANALYZED USING ANCIENT AND CONTEMPORARY
MITOCHONDRIAL DNA
Abstract
by Bobbi May Johnson, Ph.D. Washington State University December 2016
Chair: Patrick A. Carter Co-Chair: Gary H. Thorgaard
Pacific salmon (Oncorhynchus spp.) serve an important social and economic role in western North America. Despite historical abundance, native salmonids are now at risk of extinction throughout much of their native range. The accurate characterization of historic range, population size, and gene flow is essential for the development of successful conservation strategies. Therefore, conservation disciplines may look to the past to inform the future. One framework for such investigations is phylogeography, which examines geographical and genealogical connections in an effort to understand the evolutionary history of organisms.
Another avenue is the study of genetic data from temporally diverse samples, which facilitates the direct observation of a genetic history.
We utilize both methods to examine the history of two salmonid species, Chinook salmon (Oncorhynchus tshawytscha) and sockeye salmon (Oncorhynchus nerka). Additionally, two studies specific to challenges associated with degraded and low copy number (LCN) DNA sources were pursued. Chapter one characterizes the genetic diversity of Chinook salmon in the
v Columbia River basin using ancient and contemporary samples. The results indicate less genetic diversity in contemporary samples relative to ancient counterparts for both the Snake River and mid/upper-Columbia River subbasins studied. However, there appear to be higher losses of diversity in the mid/upper-Columbia than in the Snake subbasin. Chapters two and three explore specific challenges related to the study of ancient DNA: inhibition and contamination.
The investigation of inhibition describes the development of a modified PCR protocol, rescue
PCR, which successfully amplifies DNA of low-copy number in the presence of inhibitors.
Chapter three, focusing on contamination, summarizes attempts to obtain DNA from formalin- preserved specimens and the resulting identification of non-target DNA from those samples.
Finally, chapter four examines the history of sockeye salmon using DNA from contemporary samples in a phylogenetic framework. The genetic marker utilized was directly comparable to previous studies of Chinook salmon and rainbow trout (Oncorhynchus mykiss), allowing for comparisons of the phylogeographic history of three salmonid species with varying life histories. The results revealed limited genetic diversity and a more recent within-species divergence for sockeye salmon.
vi TABLE OF CONTENTS
Page ACKNOWLEDGEMENTS ...... iii
ABSTRACT ...... v
LIST OF FIGURES ...... x
LIST OF TABLES ...... xii
PERSPECTIVES ...... 1
RESEARCH INTRODUCTION AND SUMMARY ...... 4
CHAPTER ONE: Seven millennia of change: Comparison of ancient and modern Columbia basin Chinook salmon using mitochondrial DNA
1. INTRODUCTION ...... 13
2. METHODS ...... 19
Samples ...... 19
DNA extraction and amplification ...... 25
Data analysis ...... 32
3. RESULTS...... 43
4. DISCUSSION ...... 49
Morphometric vs. genetic species identification ...... 50
Size estimation ...... 52
Tests for correlation of haplotypes with climatic periods ...... 53
Genetic diversity and demography ...... 55
Notes on reintroduction ...... 69
Summary ...... 71
vii
5. LITERATURE CITED ...... 74
6. FIGURES AND TABLES ...... 87
CHAPTER TWO: Rescue PCR: Reagent-rich PCR recipe improves amplification of degraded DNA extracts
1. INTRODUCTION ...... 118
2. METHODS ...... 121
3. RESULTS AND DISCUSSION ...... 130
4. LITERATURE CITED ...... 140
5. FIGURES AND TABLES ...... 145
CHAPTER THREE: Archival archrival: Evidence of contamination via non-human DNA in formalin preserved specimens
1. INTRODUCTION ...... 157
2. METHODS ...... 160
3. RESULTS...... 167
4. DISCUSSION ...... 168
5. LITERATURE CITED ...... 173
6. FIGURES AND TABLES ...... 179
CHAPTER FOUR: Mitochondrial survey of sockeye salmon reveals limited genetic diversity and recent within-species divergence
1. INTRODUCTION ...... 189
2. METHODS ...... 192
viii 3. RESULTS...... 197
4. DISCUSSION ...... 201
5. LITERATURE CITED ...... 208
6. FIGURES AND TABLES ...... 214
APPENDICES
A. Supplemental tables ...... 226
B. Annotated script used to model genetic drift ...... 259
ix LIST OF FIGURES
CHAPTER ONE
1. Figure 1: Map of sample localities and locations noted in the study ...... 87
2. Figure 2: Salmon vertebrae ...... 88
3. Figure 3: Conceptual schematic of general ancient DNA processing procedure ...... 89
4. Figure 4: Primer targets and positions for DLoop haplotype determination ...... 90
5. Figure 5: Three-dimensional haplotype network for Columbia R. sample group ...... 91
6. Figure 6: Conceptual outline for simulation of genetic drift ...... 92
7. Figure 7: Regression plots for size estimation ...... 93
8. Figure 8: Haplotype networks ...... 94
9. Figure 9: Rarefaction curves for ancient and contemporary samples ...... 96
10. Figure 10: Bayesian cladogram ...... 97
11. Figure 11: Outcome distributions for simulation of genetic drift ...... 98
12. Figure 12: Extended Bayesian skyline plots ...... 99
13. Figure 13: Spatial and temporal comparison of haplotypes...... 101
CHAPTER TWO
14. Figure 1: Sample location and age data ...... 145
15. Figure 2: Experimental set-up to investigate decreasing DNA concentrations ...... 146
16. Figure 3: Experimental set-up to investigate changing DNA-inhibitor ratios ...... 147
17. Figure 4: Amplification results for decreasing DNA concentrations ...... 148
18. Figure 5. Amplification results for changing DNA-inhibitor ratios ...... 149
x
CHAPTER THREE
19. Figure 1: Summary of methods applied to fixed specimens ...... 179
20. Figure 2: Photographs of specimen jars and lid with paper ...... 181
CHAPTER FOUR
21. Figure 1: Bayesian analysis and distribution of haplotypes for sockeye salmon ...... 214
22. Figure 2: Nested haplotype network for sockeye salmon ...... 215
23. Figure 3: Mismatch distributions for significantly associate clades of sockeye salmon ..... 216
24. Figure 4: Comparative Bayesian phylogenetic analysis ...... 217
25. Figure 5: Nested haplotype network for rainbow trout ...... 218
xi LIST OF TABLES
CHAPTER ONE
1. Table 1: Haplotype and diversity summary ...... 102
2. Table 2: DLoop primer sequences and annealing temperatures ...... 104
3. Table 3: φST comparisons ...... 105
4. Table 4: Success rate comparisons for ancient and contemporary samples ...... 106
5. Table 5: Species identification by morphometric and genetic methods ...... 108
6. Table 6: Predicted fork length based on vertebra size ...... 110
7. Table 7: SNP table for 12S species and DLoop haplotypes ...... 111
8. Table 8: Probability summary for drift simulation ...... 114
9. Table 9: Probability summary for abbreviated drift simulation ...... 115
10. Table 10: ESS and Bayes factor comparison values ...... 116
11. Table 11: Comparison of spatial and temporal haplotype frequencies ...... 117
CHAPTER TWO
12. Table 1: Results of PCR amplification test of standard and rescue PCR protocols...... 151
13. Table 2: Results of increasing individual PCR reagents and combinations of reagents ...... 153
14. Table 3: Sequence quality comparison for rescue and standard PCR products ...... 154
15. Table 4: Application of method and efficiency for standard and rescue PCR ...... 156
xii CHAPTER THREE
16. Table 1: Sample summary with extraction, PCR, and sequencing results ...... 182
17. Table 2: Extraction variables tested on single samples ...... 183
18. Table 3: Specimen and processing data for investigation of contamination ...... 187
19. Table 4: DNA sequence results ...... 188
CHAPTER FOUR
20. Table 1: Locality and grouping information for sockeye salmon samples ...... 219
21. Table 2: Sample localities and genetic diversity for Chinook salmon ...... 220
22. Table 3: Sample localities and genetic diversity for rainbow trout ...... 221
23. Table 4: Haplotype composition and genetic diversty for sockeye salmon ...... 222
24. Table 5: Comparative statistics for diversity and differentiation ...... 224
SUPPLEMENTAL TABLES
25. Table S1: CHAPTER ONE; Sample and processing summary ...... 226
26. Table S2: CHAPTER ONE; Summary statistics for drift simulation ...... 245
27. Table S3: CHAPTER ONE; Summary of samples for which haplotype determined ...... 247
28. Table S4: CHAPTER ONE; Details of contemporary samples described in the study ...... 248
29. Table S5: CHAPTER TWO; Sample and processing summary ...... 251
30. Table S6: CHAPTER FOUR; Exact tests of population differentiation for sockeye salmon ... 255
31. Table S7: CHAPTER FOUR; Exact tests of population differentiation for Chinook salmon .. 257
32. Table S8: CHAPTER FOUR; Exact tests of population differentiation for rainbow trout ...... 258
xiii
DEDICATION
For my father, who encouraged me to always wonder why.
xiv PERSPECTIVES
For geneticists who confine their research to contemporary samples (as I once did), it is difficult to appreciate the efforts required to incorporate ancient samples. In our case, these efforts started with sample acquisition. The ancient salmon vertebrae incorporated in this study originated as midden, (i.e. the trash of ancient peoples) in archeological collections. It is no small feat to identify collections with such samples and then subsequently obtain the necessary permissions to analyze those samples. Understandably, there was often hesitation to allow for the destructive analysis of artifacts. In some cases, our requests were denied or delayed past the point where incorporation of the samples was possible.
For those samples that were obtained, the subsequent extractions and analyses were demanding and time-consuming. Genetic analysis could not be initiated until the careful process of cataloging (photographing, measuring, and weighing the fish vertebrae) was completed. When the laboratory analysis finally began, it too moved slowly. The risk of contaminating DNA is ever present, but is particularly pronounced during the initial stages.
Should contaminating DNA be co-extracted with that from the sample the resulting DNA template is useless. To minimize the risks, samples are processed in small batches (seven at a time). Unlike contemporary samples where an extraction protocol generally occupies a single afternoon, the extraction of ancient DNA (by our protocols) spans multiple days. Consider that
579 extracts were generated for this study, processing seven at a time required 83 separate batches. Assuming each batch took two days the result is 165 full eight hour days of lab work just to generate the DNA extracts for study. This represents only a small fraction of the total
1 process. There is still the testing for and removal of inhibitors, genetic species confirmation,
PCR and sequencing multiple DNA targets for the haplotype characterization, assembling and cleaning the genetic sequence, confirmation protocols, and of course data analysis. Further, ancient DNA extracts behave of their own accord and amplification, even when DNA is present, is sporadic. Indeed, most of the days spent in the lab generated little more than a paper trail.
The difficulty of ancient samples can also complicate efforts related to study design. A powerful aspect of incorporating ancient samples is that it permits direct observations of the past. For salmonids, there has been much speculation about the impacts of the early commercial fishery versus those that came after. We originally hoped to empirically evaluate this question using samples dated between 1800 and 1950. The samples consisted of preserved scales, vertebrae, dried skin, and formalin-preserved specimens. Despite extensive efforts, only a single 1800-1950 sample was successful, an 1850 vertebra from Fort Colvile, a Hudson’s Bay
Company fort. In this case, although we were able to locate, successfully acquire, and dedicate a great deal of time to processing these samples, we were ultimately unable to answer one of the key objectives originally pursued. The efforts were not entirely fruitless; in addition to the single sample characterized it was through these efforts that we identified contaminating non- human DNA in formalin-preserved museum specimens (Chapter Three).
Another unintended product related to ancient DNA is the work described in Chapter
Four. The ancient vertebrae we acquisitioned were from geographic locations where several salmonid species were historically present. We expected that both Chinook salmon and sockeye salmon might be abundant in the sample collections. As no comparable dataset existed for sockeye salmon we generated one. A total of 468 samples were obtained and sequenced for
2 the genetic target pursued in the studies here. However, the final complement of ancient samples from which genetic data could be obtained was 80% Chinook salmon and 18% non- salmonid species (e.g. suckers and minnows), with two coho salmon (O. kisutch). Not a single sockeye salmon was observed in the ancient datasets. The contemporary data was used instead to develop a phylogeographic study that was comparable to that of Chinook salmon and rainbow trout. In addition to the perspectives gained from the phylogeographic analysis, we expect that future studies of ancient DNA from sockeye salmon may find the characterization valuable.
However difficult, ancient DNA work is also rewarding. Difficult extracts provide the opportunity for creative thinking and experimentation. This is evidenced in the development of the rescue PCR method described in Chapter Two. The method developed is simple, perhaps going against the convention that ideas must be complex to be valuable. Further, it was developed in defiance of a long-standing, but perhaps unwritten, assumption about inhibition.
Specifically, that increasing the concentration of any reagents other than MgCl2 or Taq polymerase would be moot in efforts to overcome inhibition. The fleeting nature of success also delivers moments of great triumph. Perhaps most rewarding is the knowledge that even a single successful result is the opportunity to travel back in time; a chance to know something that could otherwise not be known.
3 RESEARCH INTRODUCTION AND SUMMARY
Environmental variation over space and time can have profound impacts on species success. Conservation disciplines seek to understand and predict responses to changing conditions to develop appropriate management strategies. To this end, comparisons of ecological and genetic data are widespread. The characterization of genetic patterns across both time and space, permits assessment of specific changes in population histories as they relate to changing environments. Many methodologies have been developed for such characterizations, we will explore two of these: phylogeographic analysis and the use of temporally spaced samples.
Phylogeography seeks to describe historical processes that shaped the contemporary distributions of genetic diversity (Avise 2000). In its original form, the field was limited to somewhat qualitative post hoc descriptions of historical processes (Knowles and Maddison
2002). However, subsequent work developed the framework to test specific hypotheses within the phylogeographic context [see Knowles and Maddison (2002) Nielsen and Beaumont (2009), and Templeton (2009)]. Phylogeographic methods have proven to be a robust means to infer ancestral source populations [e.g. Takahata (1993), Knowles (2001), Knowles and Maddison
(2002)], the effective population size of ancestral populations [e.g. Takahata et al. (1995),
Wakeley (2003)], speciation events [e.g. Hewitt (1996), Stone (2000)], and specific glacial refugia and dispersal history [e.g. Taberlet et al. (1998), Waltari et al. (2007)]. However, these inferences are generally limited to deep-time evolutionary histories and cannot always
4 disentangle the complexities of a species history which can include population contraction and expansion, migration, and vicariance (Knowles and Maddison 2002).
In contrast, the retrieval of genetic information from the remains of organisms that died long ago permits direct observations of the past. Empirically sampled phylochronological approaches permit more accurate assessment of specific changes in population histories than do studies of contemporary gene pools alone (Hadly et al. 2004; Johnson et al. 2007;
Ramakrishnan and Hadly 2009). The pursuit of population-level questions is dependent on the availability of such samples, accurate dating of samples and events to be considered, as well as the successful retrieval of genetic data. The information from ancient samples is particularly powerful for expounding population dynamics through time when multiple individuals from a single locality are available (Rizzi et al. 2012). Such data sets have provided valuable information on the genetic history of species including humans (Fu et al. 2016), birds (Lambert et al. 2002), and mammals (Orlando et al. 2002). However, in many cases the number of ancient samples is limited and valuable insights have been gained despite small sample sizes
[e.g. N=12 in Hofreiter et al. (2002), N=10 in Loreille et al. (2001), N=1 in Ovchinnikov et al.
(2000)].
We utilized both the phylogeographic approach and the inclusion of ancient samples to examine the history of two salmonid species, Chinook salmon (Oncorhynchus tshawytscha) and sockeye salmon (Oncorhynchus nerka). Pacific salmon (Oncorhynchus spp.) serve an important social and economic role in North America. Despite historical abundance, native salmonids are now at risk of extinction throughout much of their native range (Waples and Hendry 2008). As of 2007, it was estimated that 30% of historic populations of Pacific salmon in the United Sates
5 had been lost. Currently, 50% of the remaining populations are protected under the
Endangered Species Act (Gustafson et al. 2007).
The dissertation focuses on mitochondrial DNA (mtDNA) variation. Mitochondrial DNA is a single marker representing a uniparental (maternal) history. While representing only a single locus, mtDNA can be an effective marker for examining genetic variation, population structure and effective population size. Multiple measures of genetic diversity can be drawn from mtDNA; this marker is frequently employed to address population-level questions (Avise 1995;
Bernatchez and Martin 1996; Castro et al. 2010; Excoffier et al. 1992; Glenn et al. 1999; Li et al.
2010). The clonal inheritance of mtDNA frees it from the analytical complications associated with recombination. The mitochondria also lack enzymes that function to repair damage and errors during replication which causes the mitochondrial genome to evolve at a rate five to ten times that of the nuclear genome (Castro et al. 2010). Mitochondrial DNA has unique characteristics that make it useful for studies of ancient DNA (aDNA). This form of DNA persists more reliably in archived samples than does nuclear DNA, as each cell can contain hundreds to thousands of mitochondria, thereby increasing the available copy number of their genomes and mitigating some common problems encountered in aDNA studies. Further, the use of the mtDNA marker employed here allowed for comparisons of the data generated to two previous data sets. Genetic surveys of Chinook salmon (Martin et al. 2010) and rainbow trout (Brunelli et al. 2010) were conducting using the same region of mtDNA characterized here and provided valuable data for comparative analysis.
In the first chapter we use ancient DNA to examine the genetic history of Chinook salmon from two regions of the Columbia River basin. The Columbia River and its tributaries provide
6 essential spawning and rearing habitat for many salmonid species including Chinook salmon.
Chinook salmon were historically abundant throughout the basin. Native Americans in the region relied heavily on these fish for thousands of years. Following the arrival of Europeans in the 1800s, salmon in the basin experienced broad declines linked to overfishing, water diversion projects, habitat destruction, connectivity reduction, introgression with hatchery- origin fish, and hydropower development. Despite historical abundance, many native salmonids are now at risk of extinction. Research and management related to Chinook salmon is usually explored under what are termed “the four H’s”: habitat, harvest, hatcheries, and hydropower; here we explore a fifth H, history. We analyze patterns of ancient and contemporary mitochondrial DNA variation from Chinook salmon to characterize population genetic structure prior to recent alterations and, thus, elucidate a deeper history for this species. Our analysis provides the first direct measure of reduced genetic diversity for Chinook salmon from the ancient to the contemporary period, as measured both in direct loss of mitochondrial haplotypes as well as reductions in haplotype and nucleotide diversity. However, these losses do not appear equal across the basin, with higher losses of diversity in the mid-Columbia than in the Snake subbasin.
In the second chapter we describe a novel method for the amplification of low-copy- number (LCN) DNA extracts. Minimizing the inadvertent co-extraction of polymerase chain reaction (PCR) inhibitors and/or subduing their influence are two of the most pervasive challenges in the study of ancient DNA. Some commonly employed methods to circumvent inhibition include dilution of DNA extracts and/or removal of inhibitors via silica-based treatments. While these methods have been shown to be effective, they may not be useful for
7 all aDNA extracts. Samples with very low copy number, for instance, may not benefit from such methods as dilutions to lower DNA concentration in tandem with the inhibitors and some DNA loss is expected to follow silica-based treatments. Therefore, the development of additional options to overcome PCR inhibition is at a premium. In this study, we present evidence that a reagent-rich PCR protocol, where all reagents are increased in equal relative proportion can increase amplification success when DNA concentration is reduced relative to inhibitors. The reagent-rich PCR recipe, termed rescue PCR, increased amplification success by 51% for the 112 extracts used in the study. Rescue PCR represents a simple and robust addition to the suite of options currently available to work with DNA in the presence of inhibition, especially ancient and degraded DNA extracts.
In chapter three we explore another challenge of low-copy number samples, contamination. Museum collections, which contain specimens collected and curated from spatially and temporally diverse contexts, represent potential treasuries of molecular data.
However, access to such data has proven challenging, particularly in the case of formalin fixed specimens from wet collections. Despite a number of studies that have developed methods specific to obtaining molecular data from these types of specimens, few have successfully utilized such methods in applied studies. We present a summary of our attempts to apply seven variations on published methods to a series of fixed samples and describe an additional challenge we encountered: contaminating DNA. Although the presence of contaminating human DNA has previously been demonstrated for fixed specimens, we identified DNA initiating from non-human origins. These observations have critical implications for archive curation practices, as well as the authentication of data retrieved from fixed specimens.
8 Finally, in chapter four we examine the deep-history of sockeye salmon in a phylogeographic framework. Pacific salmonids commonly exhibit anadromous life histories, hatching in freshwater before migrating to the ocean to mature and then returning to their natal locations to spawn and perish. However, great variation exists in this model both across and within the genus. Successful postglacial dispersal from refugia was inextricably linked to these life history requirements and flexibility in regards to those requirements. We conducted a survey of 468 sockeye salmon from thirty localities ranging from the lower Columbia Basin in
North America to Hokkaido Island, Japan that is directly comparable to existing data for
Chinook salmon and rainbow trout. These three species represent a spectrum of life history requirements. Chinook salmon are highly dependent on the ocean environment while rainbow trout can complete their full life cycle in freshwater. Sockeye require access a lentic spawning environment to complete their life cycle causing the contemporary genetic profile to structure around lakes. Our data indicates that this lake requirement may have also influenced the ability of this species to disperse from refugia, resulting in limited genetic diversity, shallow population structure, and a recent within-species divergence compared to Chinook salmon and rainbow trout. Our results also indicate that genetically distinct and diverse populations are present in the Columbia River, which may represent important conservation targets for this species.
9 LITERATURE CITED
Avise, J. C. 1995. Mitochondrial DNA polymorphism and a connection between genetics and demography of relevance to conservation. Conservation Biology 9(3):686-690.
Avise, J. C. 2000. Phylogeography: The history and formation of species. Harvard university press.
Bernatchez, L., and S. Martin. 1996. Mitochondrial DNA diversity in anadromous rainbow smelt, Osmerus mordax Mitchill: a genetic assessment of the member-vagrant hypothesis. Canadian Journal of Fisheries and Aquatic Sciences 53(2):424-433.
Brunelli, J. P., C. A. Steele, and G. H. Thorgaard. 2010. Deep divergence and apparent sex-biased dispersal revealed by a Y-linked marker in rainbow trout. Molecular Phylogenetics and Evolution 56(3):983-990.
Castro, J. A., A. Picornell, and M. Ramon. 2010. Mitochondrial DNA: A tool for populational genetics studies. International Microbiology 1(4):327-332.
Excoffier, L., P. E. Smouse, and J. M. Quattro. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131(2):479-491.
Fu, Q., C. Posth, M. Hajdinjak, M. Petr, S. Mallick, D. Fernandes, A. Furtwängler, W. Haak, M. Meyer, and A. Mittnik. 2016. The genetic history of Ice Age Europe. Nature.
Glenn, T. C., W. Stephan, and M. J. Braun. 1999. Effects of a population bottleneck on whooping crane mitochondrial DNA variation. Conservation Biology 13(5):1097-1107.
Gustafson, R. G., R. S. Waples, J. M. Myers, L. A. Weitkamp, G. J. Bryant, O. W. Johnson, and J. J. Hard. 2007. Pacific salmon extinctions: Quantifying lost and remaining diversity. Conservation Biology 21(4):1009-1020.
Hadly, E. A., U. Ramakrishnan, Y. L. Chan, M. Van Tuinen, K. O'Keefe, P. A. Spaeth, and C. J. Conroy. 2004. Genetic response to climatic change: insights from ancient DNA and phylochronology. PLoS Biology 2(10):e290.
Hewitt, G. M. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biological journal of the Linnean Society 58(3):247-276.
Hofreiter, M., C. Capelli, M. Krings, L. Waits, N. Conard, S. Münzel, G. Rabeder, D. Nagel, M. Paunovic, and G. Jambresić. 2002. Ancient DNA analyses reveal high mitochondrial DNA sequence diversity and parallel morphological evolution of late Pleistocene cave bears. Molecular Biology and Evolution 19(8):1244-1250.
10 Johnson, J. A., P. O. Dunn, and J. L. Bouzat. 2007. Effects of recent population bottlenecks on reconstructing the demographic history of prairie‐chickens. Molecular Ecology 16(11):2203-2222.
Knowles, L. L. 2001. Did the Pleistocene glaciations promote divergence? Tests of explicit refugial models in montane grasshopprers. Molecular Ecology 10(3):691-701.
Knowles, L. L., and W. P. Maddison. 2002. Statistical phylogeography. Molecular Ecology 11(12):2623-2635.
Lambert, D. M., P. A. Ritchie, C. D. Millar, B. Holland, A. Drummond, and C. Baroni. 2002. Rates of evolution in ancient DNA from Adélie penguins. Science 295(5563):2270-2273.
Li, S.-F., Q.-L. Yang, J.-W. Xu, C.-H. Wang, D. C. Chapman, and G. Lu. 2010. Genetic diversity and variation of mitochondrial DNA in native and introduced bighead carp. Transactions of the American Fisheries Society 139(4):937-946.
Loreille, O., L. Orlando, M. Patou-Mathis, M. Philippe, P. Taberlet, and C. Hänni. 2001. Ancient DNA analysis reveals divergence of the cave bear, Ursus spelaeus, and brown bear, Ursus arctos, lineages. Current Biology 11(3):200-203.
Martin, K. E., C. A. Steele, J. P. Brunelli, and G. H. Thorgaard. 2010. Mitochondrial variation and biogeographic history of Chinook salmon. Transactions of the American Fisheries Society 139(3):792-802.
Nielsen, R., and M. A. Beaumont. 2009. Statistical inferences in phylogeography. Molecular Ecology 18(6):1034-1047.
Orlando, L., D. Bonjean, H. Bocherens, A. Thenot, A. Argant, M. Otte, and C. Hänni. 2002. Ancient DNA and the population genetics of cave bears (Ursus spelaeus) through space and time. Molecular Biology and Evolution 19(11):1920-1933.
Ovchinnikov, I. V., A. Götherström, G. P. Romanova, V. M. Kharitonov, K. Liden, and W. Goodwin. 2000. Molecular analysis of Neanderthal DNA from the northern Caucasus. Nature 404(6777):490-493.
Ramakrishnan, U., and E. A. Hadly. 2009. Using phylochronology to reveal cryptic population histories: review and synthesis of 29 ancient DNA studies. Molecular Ecology 18(7):1310-1330.
Rizzi, E., M. Lari, E. Gigli, G. De Bellis, and D. Caramelli. 2012. Ancient DNA studies: New perspectives on old samples. Genetics Selection Evolution 44(1):1.
Stone, G. 2000. Phylogeography, hybridization and speciation. Trends in Ecology & Evolution 15(9):354-355.
11 Taberlet, P., L. Fumagalli, A.-G. Wust-Saucy, and J.-F. Cosson. 1998. Comparative phylogeography and postglacial colonization routes in Europe. Molecular Ecology 7(4):453-464.
Takahata, N. 1993. Allelic genealogy and human evolution. Molecular Biology and Evolution 10(1):2-22.
Takahata, N., Y. Satta, and J. Klein. 1995. Divergence time and population size in the lineage leading to modern humans. Theoretical Population Biology 48(2):198-221.
Templeton, A. R. 2009. Statistical hypothesis testing in intraspecific phylogeography: nested clade phylogeographical analysis vs. approximate Bayesian computation. Molecular Ecology 18(2):319-331.
Wakeley, J. 2003. Inferences about the structure and history of populations: coalescents and intraspecific phylogeography. The Evolution of Population Biology:193-215.
Waltari, E., R. J. Hijmans, A. T. Peterson, Á. S. Nyári, S. L. Perkins, and R. P. Guralnick. 2007. Locating Pleistocene refugia: comparing phylogeographic and ecological niche model predictions. PLoS One 2(7):e563.
Waples, R. S., and A. P. Hendry. 2008. Special issue: Evolutionary perspectives on salmonid conservation and management. Evolutionary Applications 1(2):183-188.
12 CHAPTER ONE
Seven millennia of change: Comparison of ancient and modern Columbia basin Chinook
salmon using mitochondrial DNA
1. INTRODUCTION
1.1. Background
Chinook salmon are the largest, and perhaps most well known, of the Pacific salmon.
Hatched and reared in freshwater, young fish migrate to the ocean to mature and grow for two to six years before returning to their natal location for spawning and subsequent mortality.
Chinook salmon exhibit a wide variety of life histories that are intimately linked to environmental variables such as temperature, photoperiod, and stream discharge (Crozier et al.
2008; Utter et al. 1995). Two generalized forms have evolved, often termed races (Healey et al.
1991). The ocean-type race limits its time in freshwater, migrating to the ocean shortly after hatching and delaying their return until shortly before spawning. The stream-type race spends a greater proportion of their lives in freshwater during both the juvenile and spawning portions of the life cycle. Chinook salmon also display seasonal variability in migrations, with runs occurring in all four seasons. The runs are named for the season of freshwater reentry which initiates the spawning migration. Mainstem and lower tributary streams of both the Columbia and Snake Rivers produce predominately ocean-type races while the upper tributaries produce stream-type races (Waknitz et al. 1995). In the central basin, tributaries produce both races, with spring runs producing stream-type offspring and fall runs producing ocean-type offspring.
13 Summer runs differ by river, producing ocean-type offspring in the Columbia River and stream- type offspring in the Snake River (Waknitz et al. 1995).
The Pacific Northwest provides essential habitat for Chinook salmon. Conditions in the region varied dramatically throughout the Holocene. Chatters et al. (1995) described three distinct climate periods for the basin in relation to their quality for salmonids. The first, prior to
4000 YBP were ‘poor’, followed by ‘optimum’ conditions from 2200 to 3500 YBP, and then
‘good’ conditions from 1000 YBP to present day (Chatters et al. 1995). Geological and terrestrial characteristics in the Pacific Northwest have been relatively stable for the past 5000 years.
However, periodic disturbances have been documented. Multiple eruptions from regional volcanos over the past 2500 years likely resulted in largescale sedimentation of nearby rivers
(O'connor et al. 2003). The Bonneville Landslide or Bridge of the Gods, which occurred around
1450 AD, temporally dammed the mainstem Columbia River and formed the Cascade Rapids
(O'connor et al. 2003; Waples et al. 2008). It is likely that these events, among others, altered migration and survival of Chinook salmon in the basin.
Native American utilization of Chinook salmon, along with other salmonids, has been documented in the region for over 9000 years (Butler and O'connor 2004). Historically, the region supported large populations of Native Americans. Diseases introduced during European contact rapidly reduced population sizes. By 1851, population estimates were less than 9000, a reduction of as much as 85% (Dart 1851). Salmon were, and are today, of universal importance to Native Americans in the region (Boas 1894). In addition to having cultural and religious significance, salmon represent an important source of nutrition. Reports from early settlers note both an impressive number of salmon and an impressive number of means by which they
14 were captured. In fact, most mass capture methods implemented by European settlers were based on those already in use by Native Americans. These methods included baskets, dip nets, hook and line, seines, traps, and weirs (Hewes 1947). Aboriginal fishing tended to center around natural barriers, most commonly waterfalls, which concentrated runs as fish attempted to navigate upriver (Netboy 1980). At these sites, fish could be speared or harpooned individually as well as taken en masse with dipnets or baskets. Baskets were capable of capturing several hundred fish at one time (Wilkes 1845). Scaffolding was often constructed over the falls to aid in the taking of fish. These fishing stands carried equity, passed down paternally or even occasionally used as payment for wives (Netboy 1980). Prominent falls fishing sites such as Celilo Falls and Kettle Falls were so productive that several thousand Native
Americans may have been present at one location over the course of a fish run (Ross 1986).
Fishing was not exclusively focused on the falls. On the sandy shoals of the lower river, large seines were used to capture fish and in smaller tributaries weirs were commonly constructed to guide fish into small baskets or bins for collection (Smith 1979).
European settlers were quick to recognize the economic opportunity presented by the extensive salmon runs. Lichatowich and Mobrand (1995) broadly segregate commercial exploitation of Chinook salmon in the Columbia River system into four major phases: (1) 1866 to 1888: initial development of the fishery, (2) 1889 to 1922: the productive phase, (3) 1923 to
1958: period of notable decline, (4) 1958 to current: maintenance of reduced productivity.
During the productive phase (1889 to 1922) as many as 25 million pounds of Chinook salmon were harvested annually; this number was reduced to 15 million during the decline (1923 to
1958), and is now maintained at around 5 million pounds (Lichatowich and Mobrand 1995).
15 Hatcheries were viewed as a means to mitigate the losses and increase production. By the
1905, 62 million eggs and fry were released by hatcheries in the Pacific Northwest (Cobb 1921).
Despite these efforts, fisheries continued to decline and annual harvest losses of 14 million pounds of fish were reported from 1905 to 1909 (Smith 1979). Hatchery mitigation in the
Columbia River intensified in the 1960s, and by 1995 as much as 80% of the Columbia River
Chinook salmon were hatchery-origin fish (Lichatowich and Mobrand 1995).
Further threats to salmon in the region came in 1933 when the first mainstem dam was constructed on the Columbia River at Rock Island, near Wenatchee, Washington. Construction of Bonneville Dam began the same year, with completion in 1938. This was followed by the construction of Grand Coulee Dam in 1941. Grand Coulee Dam had the greatest impact, blocking anadromous salmonids from 1770 river kilometers of the upper-Columbia, approximately 40% of their historically available habitat (Smith 1979; Waknitz et al. 1995).
Hydroelectric development in the Snake River followed a decade later. Dams were constructed on the lower portion of the river (i.e. Ice Harbor, Lower Monumental, Little Goose, and Lower
Granite Dams) from 1956 to 1975. These lower river dams included at least some fish passage for Chinook salmon. However, the Hells Canyon dam, located on the upper Snake and completed in 1967, did not. The Columbia River Basin system is now the most hydroelectrically- developed in world (Lang 2011). Over four hundred dams can be found in the Columbia River
Basin, 56 of which were constructed exclusively for hydropower. Today there are nine dams between the furthest inland salmon spawning tributaries in the mid-Columbia and the ocean, and eight dams between the furthest inland salmon spawning tributaries in Snake River
16 tributaries and the ocean. In total, more than 55% of the historically available spawning habitat in the Columbia River Basin is now blocked by dams (Harrison 2008a).
1.2. Objectives and study overview
As just summarized, largescale declines of Chinook salmon are well-documented following the early 19th century arrival of Europeans in the Pacific Northwest. Although it is often hypothesized that losses in genetic diversity should be coincident with losses in census sizes
(Waknitz et al. 1995), no direct quantification of diversity has been made of Chinook salmon from the pre-contact era to make such a comparison. We sought to directly test the hypothesis that genetic diversity was higher in the pre-contact period relative to the post-contact
(contemporary period). To this end, our objectives were to (1) characterize the genetic diversity in ancient Chinook salmon using mitochondrial DNA, (2) to compare those characterizations to analogous contemporary groups, (3) to quantify changes in diversity (if present), and (4) to broadly explore demographic scenarios that may have occurred in the past.
There are various definitions for subdividing the regions of the Columbia River basin, especially notable are definitions of the mid-Columbia and upper-Columbia subbasins. For the purposes of this study we use lower to refer to the area downstream of the Columbia-Snake confluence, mid for area from the Snake-Columbia confluence to the Grand Coulee Dam and upper for the area above the Grand Coulee Dam (Figure 1). Although the upper-Columbia historically supported many anadromous fish populations, the area we refer to as the mid-
Columbia now represents the uppermost spawning habitat for these life histories.
17 The research presented here focuses on three complementary groups referred to as the
Columbia River group, Snake River group, and the Spokane River group. The Columbia River and
Snake River sample groups have broad commonality. They shared a lower river connection to ocean and would have experienced broadly similar climate and geological events. However, the groups are notably distinct. The Snake River subbasin experienced weather and fine-scale climate patterns that differ from the mid- and upper-Columbia River (Davis et al. 1986). Pre- contact utilization may not have been equivalent as historically Native Americans living in the basin were concentrated in the shared lower river or around mid- and upper-Columbia falls sites (Hewes 1947). Post-European settlement patterns of exploitation, hydroelectric development, and fisheries management also varied in time and scale. The Spokane River is part of the upper-Columbia subbasin. The Columbia and Snake river samples represent fish caught by Native Americans as the salmon migrated up the mainstem portion of the respective rivers to a number of terminal spawning locations. These groups can be thought of as “mixed stock” sample collections. In contrast, samples in the Spokane group were collected at a terminal fishing location a short distance below an impassable falls rather than a site where multiple local stream populations would be passing (Scholz et al. 2014). It is likely that the remains found at this site represent fish that were returning to spawn very near the location and this group can be thought of as a “single stock”.
18 2. METHODS
2.1. Samples
A total of 757 ancient and contemporary samples were processed for this study. Ancient samples consisted of a total of 365 vertebrae (Figure 2) previously identified in their respective collections as “Oncorhynchus spp.” or “likely Oncorhynchus spp.” based on morphometric analysis. Twenty-six non-bone samples (dried skin, preserved scales, and formalin-fixed tissues) were also attempted (Table S1). None of the non-bone samples resulted in successful amplification of target DNA and are not further discussed in this chapter. However, a subset of these samples are discussed in Chapter 3. Ancient sample dates (Table 1 and Table S1) were derived from the materials referenced below and are based on radiocarbon dating and well understood cultural and geological chronologies. To facilitate interpretation, all dates were converted to years before present (YBP). These samples represent a unique research resource, one which has been made available through extensive archeological work carried out over many decades, most often related to dam building and reservoir development in the Columbia
River basin.
Contemporary samples consisted of 366 fin clips, stored in ethanol, taken from fish identified as Chinook salmon (Table 1 and Table S2). The data directly extracted and sequenced for this dissertation is supplemented with that of Martin et al. (2010) in a survey of Chinook salmon throughout the species range.
19 2.1.1. Columbia River group: Ancient
Ancient samples for the Columbia River group are comprised of vertebrae from three locations near or above the current location of the Grand Coulee Dam (locations 1-3 depicted in
Figure 1). These samples were provided courtesy of the Bureau of Reclamation and the Colville
Confederated Tribes Repository, History/Archaeology Program.
Fort Colvile and Kettle Falls
Fort Colvile (45ST97) (location 2 depicted in Figure 1), excavated in 1976, was a trading post that operated from 1825 to 1871 under the Hudson’s Bay Company (Barman and Watson
1999). The one sample successfully sequenced from this site represents an intermediate time point: ~1850 AD. This temporal period occurs post-European contact but prior to that considered ‘contemporary’. The sample is omitted from selected analyses and such instances are noted.
Kettle Falls (location 3 depicted in Figure 1) was an important Native American fishing site with evidence of extended and diverse occupation over 9000 years (Galm 1994). Samples were obtained from two archeological sites at Kettle Falls: Ksunku (45FE45) and Shonitkwu (45FE44).
Shonitkwu (45FE44) was excavated in 1971 and 1974, however a large portion of records from the 1971 excavations were lost in a fire (Chance and Chance 1982) so the materials and associated data are from the 1974 excavation (Pouley 2008). Ksunku (45FE45) was excavated in
1974 and 1978 (Galm 1994). Both the Fort Colvile site and Kettle Falls were inundated by Lake
Roosevelt congruent with construction of Grand Coulee Dam but were excavated during periods of reservoir drawdown in the 1970s.
20 Grand Coulee Dam Project Area
Excavated in 1986, the Grand Coulee Dam Project Area (45DO189) (location 1 depicted in
Figure 1), is a collection excavated from a single housepit and two yard blocks. The site experienced multiple occupations over the past 7000 years with up to 28 people in residence at any given time (Galm and Lyman 1988).
2.1.2. Columbia River group: Contemporary
The Columbia River Group study focuses on spawning aggregates that historically migrated upstream of Grand Coulee Dam (locations 1-3 depicted in Figure 1). No comparable contemporary samples are available from these upper-Columbia sites due to the construction
Grand Coulee Dam in the 1930s, which abolished salmon passage upstream of its location
(Waknitz et al. 1995). However, during construction of the dam, biologists implemented the
Grand Coulee Fish Maintenance Project (GCFMP), which attempted to redirect spawning efforts of fish that would naturally pass the dam into downstream tributaries. From 1939 to 1943, fish were captured at Rock Island Dam (see Figure 1) and transported to release points in tributaries
(the Wenatchee, Entiat and Methow Rivers) below Grand Coulee or were propagated in hatcheries (Fish and Hanavan 1948). The rivers selected for transplant had previously supported large runs of salmonids but had highly depressed populations by the time the transplant operation was initiated (Fish and Hanavan 1948). As a result of the effort, subsequent generations of Chinook salmon in the redirection tributaries became a mix of the progeny of the relocated stocks and any fish autochthonous to the tributaries (Waknitz et al. 1995).
Contemporary samples in this group are from fish collected from the transplant rivers utilized
21 for the Grand Coulee Fish Maintenance project (location 4 depicted in Figure 1). The contemporary portion of the Columbia River group is comprised of 240 samples collected from the redirection tributaries between 1995 and 2011 (Table S2). These samples were provided by the National Oceanic and Atmospheric Administration (NOAA), U.S. Fish and Wildlife Service
(USFWS), and Washington Department of Fish and Wildlife (WDFW).
2.1.3. Snake river group: Ancient
Ancient samples for the Snake River group are comprised of vertebrae from seven excavation locations coinciding with three contemporary dams along the Snake River (locations
6 - 8 depicted in Figure 1). Sample dating for the Snake River group proved difficult. Many of the collections indicated high levels of disturbance attributed to construction of a nearby railroad line and/or looting of the archaeological sites. In these cases, the dates are conservatively provided a range of possible ages (Table 1 and Table S1). These samples were provided by the Department of Anthropology at Washington State University who holds these collections in trust for the Walla Walla District United States Army Corps of Engineers under the provisions of 36CFR79 (Title 36, Chapter I of the Code of Federal Regulations, part 79).
Ice Harbor Dam
The Windust Caves (45FR46) at Ice Harbor Dam (location 6 depicted in Figure 1) are a series of nine caves distributed along approximately 2000 feet of cliff face on the Snake River
(Rice 1965). The caves were excavated between 1959 and 1961 prior to inundation by Lake
Sacajawea in 1962. The caves were likely used for shelter, as well as for storage, with artifacts
22 dating as old dating to 9000 years ago. Radiocarbon dates were not generated for this site and the Mazama ash layer, a calibration point for Columbia Plateau sites, is not present (Rice 1965).
Therefore, samples from this site were dated exclusively via comparison to cultural sequences.
Time points indicated in the sequence were correlated with the geological stratigraphy as well as with other Columbia Plateau sites (Rice 1965).
Lower Granite Dam
Two sites, Granite Point (45WT41) and Wexpusnime (45GA61) are associated with Lower
Granite Dam (location 8 depicted in Figure 1). Granite Point, was excavated from 1967 to 1968 and Wexpusnime from 1969 to 1970. Granite Point appears to have been occupied, with intermittent hiatus, as a camp site over 10,000 years (Leonhardy 1970). Evidence of both camp and house village occupations are noted in the Wexpusnime collection (Nakonechny 1998).
Lower Monumental Dam
Three ancient house pit villages are located near Lower Monumental Dam (location 7 depicted in Figure 1): (1) Harder (45FR40, excavated 1957), (2) Hatiuhpuh (45WT134, excavated
1987 and 1989), and (3) Three Springs Bar (45FR39, excavated in 1961) are located near Lower
Monumental Dam (Brauner 1990; Daugherty et al. 1967; Kenaston 1966). A fourth site, Marmes
Rockshelter (45FR50) is also associated with Lower Monumental Dam. However, none of the samples processed from this site yielded amplifiable DNA and therefore it is not summarized here. If desired, information about the site is available in Hicks (2004) and Lyman (2012).
23 2.1.4. Snake River group: Contemporary
Contemporary samples in the Snake River group consisted of wild and hatchery origin fish with fall, spring, and summer run timing life histories (location 9 depicted in Figure 1). Locations were targeted to include all reporting groups used to determine stock, giving the best chance of capturing the genetic diversity currently present in the Snake subbasin (Ackerman et al. 2012).
These samples were all collected and provided by the Idaho Department of Fish and Game
(IDFG). Two of the sample subgroups, the Lyons Ferry Hatchery and Tucannon River samples, represent data collected by Martin et al. (2010). These samples were provided by the
Washington Department of Fish and Wildlife.
2.1.5. Spokane River group: Ancient only
The third sample grouping originated from a collection of materials excavated near the
Spokane River (45SP266) (location 5 depicted in Figure 1). The location was an ancient fishing site with evidence of use spanning over 7000 years (Butler 2006; Lyman 2006). The collection had three components, approximated at 2500, 3250, and 7200 YBP (Butler 2006; Galm 1994). A portion of samples in this collection were discriminated from the 2500 YBP component but could not be placed definitively into either the 3250 or 7200 YBP. units due to rodent disturbance (S. Walker, Department Archaeological and Historical Services EWU; personal communication) (Table 1 and Table S1).
Anadromous fish have been extirpated from the Spokane River, leaving no directly comparable contemporary counterpart for this group. Throughout this study, comparisons are made between the ancient Spokane River group and the contemporary Columbia River
24 subgroups. While not directly connected, our intent is to compare them as single-stock components of Chinook salmon, ones sampled from proximate geographic locations in the
Columbia Basin. The Spokane River samples were obtained courtesy of the Spokane Tribe of
Indians, the Eastern Washington University Department Archaeological and Historical Services, and the City of Spokane.
2.2. DNA extraction and amplification
2.2.1. Methodology specific to ancient DNA
Ancient samples were processed using a multi-step processing procedure (Figure 3). Prior to destructive analysis, samples were measured, weighed, and photographed. Measurement data consisted of length, height, and transverse diameter and was collected for all intact vertebrae (i.e. not partial or fragmented, see Figure 1) in the Columbia and Snake River sample groups. Unfortunately, measurements of length and height were not obtained for the Spokane
River sample group. All measurements were taken three times with a digital micrometer and the average of these was used for analysis. Sample weight was collected for all vertebrae, including partial samples.
Research incorporating aDNA cannot be performed in a standard genetic laboratory.
Typically, ancient samples have been subject to a varying, but high degree of post-mortem degradation. The taphonomy (decay) of each sample is unique. However, in general, the DNA that remains in these specimens is typically degraded with regards to strand length and carries numerous forms of mis-coding lesions. Because of these characteristics, the recovery of genetic material and authentication of results from ancient samples remains a formidable challenge,
25 given the presence of contaminating DNA molecules and co-extracted impurities that inhibit the polymerase chain reaction (PCR) (i.e., PCR inhibitors). The ancient work described here was performed in the dedicated ancient DNA lab at Washington State University. This lab has a well- documented history of overcoming such challenges and producing authentic, publishable studies utilizing ancient DNA (Barta et al. 2014a; Barta et al. 2014b; Kemp et al. 2007; Kemp et al. 2014; Kemp et al. 2006; Kemp and Smith 2005; Kemp and Smith 2010; Kemp et al. 2009;
Monroe et al. 2013; Winters et al. 2011). A brief summary of the major challenges as well as the applied protocols to overcome such challenges follows below.
Contamination
Ancient DNA is generally present in a low copy number while modern DNA tends to be ubiquitous, particularly near any area where PCR amplifications are performed. To prevent potential contamination, samples were decontaminated prior to extraction using bleach, negative controls were utilized at both the extraction and PCR reaction level, and separated pre- and post- PCR facilities were utilized (Barta et al. 2013; Kemp and Smith 2005; Kemp and
Smith 2010). Additionally, all pre-PCR work for ancient samples was conducted where no modern samples (including pre-PCR reactions) are stored or processed and appropriate behaviors and lab attire were enforced (Kemp and Smith 2010).
Inhibition
Prevention of PCR amplification by non-target molecules is one of the most pervasive challenges in aDNA research. Some classes of inhibitors compete with reagents in PCR, causing
26 them to become limited too soon for adequate amplification of DNA while others may encapsulate DNA, making it inaccessible during the reaction (Cooper 1994; Handt et al. 1994).
We systematically tested ancient extracts for inhibition and took steps to remove potential inhibitors via repeat silica extractions prior to attempting sequencing reactions (Kemp et al.
2014; Kemp et al. 2006). We also utilized polymerase empirically tested for success in overcoming inhibition cofactors (Monroe et al. 2013).
Authentication
PCR reactions using template aDNA often start from small numbers of molecules relative to modern templates. When only a few molecules initiate the PCR, a single damaged molecule or error incorporated by Taq polymerase during the reaction can lead to artefactual
“mutations”. To control for this possibility, we utilized a two-fold process. First, only short
(<200bp) targets were directly sequenced in an effort to incorporate as many molecules into the reaction as possible. Second, all new or rare haplotype results, any sequences with multiple peaks, and a random sample of all results confirmed via repeat PCR amplification and sequencing of the template (Gilbert et al. 2005).
2.2.2. Ancient samples: DNA extractions
Samples were extracted in batches of seven with one accompanying extraction negative control. First, approximately 2 – 530 mg of bone was submerged in 6% (w/v) sodium hypochlorite (bleach) for 4 min (Barta et al. 2013) and the bleach decanted. The samples were then twice submerged in DNA-free water, with the water decanted following each submersion.
27 Following these decontamination efforts, extractions utilized either a silica or phenol:chloroform (p:c) based procedure (Table S1). Samples extracted via silica method were transferred to 1.5 mL tubes, to which aliquots of 500 L of EDTA (ethylenediaminetetraacetic acid) were added, and gently rocked at room temperature for >48 hours. DNA was then extracted following the WSU method described by Cui et al. (2013). Samples extracted via p:c methods were transferred to 15 mL tubes, to which aliquots of 2mL of EDTA were added, and gently rocked at room temperature for >48 hours. DNA was extracted following a modified protocol of Kemp et al. (2007) described in Moss et al. (2014).
2.2.3. Ancient samples: Tests for amplification and inhibition
All extracts were initially tested for amplification and inhibition. Amplification was pursued using primers OST12S-F 5’-GCTTAAAACCCAAAGGACTTG-3’ and OST12S-R 5’-
CTACACCTCGACCTGACGTT-3’ that target a 189 basepair (bp) portion of the 12S mitochondrial gene (Jordan et al. 2010). Note that Jordan et al. (2010) described the OST12S-R primer in the incorrect orientation. It has been corrected here. These primers have been demonstrated to be especially effective in amplifying salmonid mtDNA, however the sequences produced can be used to differentiate a variety of other fish to the species level (Grier et al. 2013; Halffman et al.
2015; Jordan et al. 2010; Kemp et al. 2014). Specific PCR protocols are described in section 2.2.4 below.
Inhibition tests followed Kemp et al. (2014) (see schematic illustration in their Figure 1). In brief, PCRs were set-up with an aDNA control, one comprised of pooled DNA extracted from
~3500 year old northern fur seal (Callorhinus ursinus) remains (Barta et al. 2014a; Barta et al.
28 2013; Winters et al. 2011). This pool was created using individual DNA extracts previously verified to yield 181 base pair (bp) amplicons of northern fur seal mitochondrial cytochrome B gene using the following primers: CytB-F 5’-CCAACATTCGAAAAGTTCATCC-3’ and CytB-R 5’-
GCTGTGGTGGTGTCTGAGGT-3’ (with an annealing temperature of 60°C) (Moss et al. 2006). This control PCR mix is then “spiked” with the DNA recovered from the salmonid vertebrae, that is
DNA to be tested for the presence of sufficient inhibition to prevent the northern fur seal mitochondrial DNA (mtDNA) from amplifying (and possibly that for fish mtDNA, which could explain a false negative). Note that the northern fur seal primers are incapable of amplifying salmonid mtDNA. One advantage of this approach to monitoring for the presence of PCR inhibitors is that the control is aDNA and exhibits characteristics common in ancient extracts
(i.e., signatures of post-mortem chemical degradation, high levels of DNA fragmentation, and low concentrations) (Barta et al. 2014a; Barta et al. 2013; Winters et al. 2011). Another advantage, given that the degree of PCR inhibition is directly related to the size of DNA to be amplified (Mccord et al. 2015), is that the northern fur seal mtDNA fragment size targeted by these reactions is similar to that targeted in salmonids (189 bp). Samples that had inhibition indicated were subjected to additional silica treatments as described in Kemp et al. (2006) until either (1) amplification was possible and inhibition was not indicated or (2) amplification was not possible but inhibition was not indicated. In the case of outcome one, amplifications were submitted for sequencing (described in section 2.2.5 below). In the case of outcome two, modified PCR methods were pursued (described in section 2.2.4 below).
29 2.2.4. Ancient samples: PCR protocols
Polymerase selection was based on results from Monroe et al (2013) indicating Klentaq LA was the least susceptible of nine polymerase or polymerase blends to inhibition associated with
DNA obtained from prehistoric salmonid vertebrae recovered from two archaeological sites in the Pacific Northwest (DgRv-003 and DgRv-006). Unmodified PCRs contained: 1X Omni Klentaq
Reaction Buffer mix (containing a final concentration of MgCl2 at 3.5 mM), 0.32 mM dNTPs,
0.24 µM each of forward and reverse primer, 0.3 U of Omni Klentaq LA polymerase, and 1.5 µL of template DNA (for 15 µL PCR reaction volume). Reaction conditions consisted of an initial three-minute denaturation at 94°C, followed by sixty 15 s cycles of 94°C (denaturation), primer specific temperature (annealing), and 68°C (extension). This was followed with a final extension at 68°C for 3 minutes. Negative PCR controls and positive PCR controls (utilizing DNA extracted from contemporary Chinook salmon, added in the post-PCR lab prior to initiating PCR) accompanied all sets of standard PCRs and modified PCRs. Modified PCR protocols consisted of increasing DNA concentration 2X or 3X as well as the use of rescue PCR, which is described in
Chapter 2. In brief, rescue PCR consists of increasing reagents in equal proportion while decreasing the volume of water to keep the reaction volume the same; rescue PCR treatments at +25% and +50% were utilized for the data presented here (Table S1).
2.2.5. Ancient samples: Sequencing and species identification
Amplification following all PCRs was confirmed via separation on a 4% agarose gel with approximate size determined against a 20 bp ladder (Bayou BioLabs). Successful amplifications were submitted for sequencing in both forward and reverse directions using the same primers
30 utilized for amplification. Product clean-up and sequencing were performed by Molecular
Cloning Laboratories (South San Francisco, CA). Sequences were aligned and analyzed using
Sequencher v 4.8 (Gene Codes; Ann Arbor, MI). Sequences generated from the amplification test were compared to those provided by Jordan et al. (2010) and to the NCBI nucleotide database using the Basic Local Alignment Search Tool (BLAST) to determine species.
2.2.6. Ancient samples: DLoop haplotype determination
Haplotypes were based on a 563 bp sequence of the mitochondrial genome, a region homologous to that previously sequenced from Chinook salmon throughout their range (Martin et al. 2010). The region spans 414 bp of the 3’ portion of the displacement loop (or DLoop), through the full 68 bp of the phenylalanine tRNA gene, and 81 bp of the 5’ portion of the 12S ribosomal RNA gene. Haplotype characterization was completed using overlapping targets of less-than 200 bp in length (Figure 4). Target position, primers, and annealing temperatures are given in Table 2.
2.2.7. Contemporary samples: DNA Extractions and DLoop haplotype determination
Contemporary samples were rinsed in TE for 1 minute followed by brief submersion in
DNA-free water. DNA digested via proteinase K was extracted following the p:c procedure of
Sambrook et al. (1989). PCR protocol followed that described in Martin et al. (2010) utilizing primers Forward: 5’-CCCGCCCCTGAAAGCCGAAT-3’, Reverse: 5’-CGTGCCCCCAGGTGCGTATG-3’ with an annealing temperature of 62°C. Amplification was confirmed via separation on a 1% agarose gel and approximate size was determined against a 1kb ladder (New England BioLabs).
31 PCR was repeated one additional time for any failures. Successful amplifications were submitted for sequencing using a nested primer 5’-ATAACCGCGGTGGCTGGCAC-3’. Product clean-up and sequencing were performed by Molecular Cloning Laboratories (South San
Francisco, CA). Sequences were aligned and analyzed using Sequencher v 4.8 (Gene Codes; Ann
Arbor, MI).
2.3. Data analysis
2.3.1. Success rates
A total of 391 ancient and 366 contemporary samples were processed during this study.
From the 391 ancient samples, species was determined for 131 and haplotypes for 78 of these.
From the 366 contemporary samples haplotype was determined for 336 (no species identification was necessary as species was determined antemortem). Success rates at the individual PCR, DNA extract, and sample/vertebrae level were calculated for both ancient and contemporary samples. In ancient samples, the rate follows the procedural flow described in section 2.2.1 and Figure 3. The rate assumes one PCR was completed for those that amplified using standard PCR, three total for those that required rescue PCR (two amplification attempts using standard PCR + one attempt using rescue PCR), and six total for those that failed to amplify (two using standard PCR + two using rescue PCR + two with increased DNA template).
The procedure represents the minimum number of PCR attempts for each sample and is therefore the maximum success rate realized. For contemporary samples, the rate is based on one PCR attempt assumed for those that amplified and two for those that amplified only after a second attempt (N=8) or that failed to amplify.
32 2.3.2. Morphometric vs. genetic species identification
Huber et al. (2011) proposed a method to distinguish Oncorhynchus spp. using morphometric measurements for both type II and type III vertebrae. The method was developed using contemporary samples and then tested on a single archeological assemblage.
In that assemblage, the morphometric method correctly identified Chinook salmon 87% to 90% of the time, depending on vertebrae type. Application of the morphometric method with DNA proofing to multiple assemblages may be useful in two ways. First, independent application to multiple assemblages provides the opportunity for replicate testing of the method. Second, given that the method is robust, multiple success rates may be used to calculate a confidence parameter for the morphometric method when genetic identification is not possible.
Morphometric species identification using the methods described in Huber et al. (2011) was compared to genetic species identification for 49 samples from the Columbia and Snake River sample groups. These samples represented the complement of vertebrae that were fully intact and that were successfully sequenced for genetic species identification (omitting 23 that were genetically determined to be species other than Oncorhynchus spp.). The morphometric identification model was implemented in program R v.3.3.1 (R Core Team 2016).
A total of four unique 12S sequences were identified for Chinook salmon (Table 3 and
Table S1). One matched that indicated in Jordan et al. (2010), and three were confirmed via
GenBank comparisons. The 12S type matching that described by Jordan et al. (2010) was the most common in all samples, comprising 94% of that determined in the Columbia, 92% for the
Snake, and 91% for the Spokane.
33 2.3.3. Size estimation
Historically a phenotypically distinct form of Chinook salmon of notable size, termed the
“June Hog”, existed in the Columbia River (Seufert and Vaughan 1980; Waknitz et al. 1995).
Historical accounts indicate that these fish continued to run in the upper-Columbia and
Spokane Rivers into the early 1900s, but were likely extirpated before 1939. No evidence of fish large enough to be classified as June hogs was present in the data collected as part of trapping and transport for the Grand Coulee Fish Maintenance Project (Fish and Hanavan 1948; Waknitz et al. 1995). If any fish of such notable size remained by 1939 they would likely have been sampled in the five years of fish collection.
We correlated vertebrae size to fork length (the distance from the tip of the snout to the tail fork) for 50 ancient vertebrae to investigate if any June Hogs were present in the sampling.
The 50 vertebrate represented all intact vertebrae genetically confirmed as Chinook salmon, omitting five for which transverse diameter was not recorded. Hofkamp (2015) measured diameter for both type 2 and type 3 vertebrae as well as fork length for contemporary fish carcasses found near Hanford Reach in the Columbia River. These measurements (N=50) were used to develop a regression equation that could estimate fork length from vertebrae diameter for the samples in this study.
2.3.4. Genetic diversity and differentiation
Genetic diversity
Haplotype richness (the raw number of haplotypes sampled) was drawn directly from the data. However, the final number of haplotyped individuals varied between the contemporary
34 and ancient sample components (7:1 for the Columbia River group and 6:1 for the Snake River group). To normalize for comparison, we generated rarefaction curves for the Columbia, Snake, and Spokane samples. Rarefaction is a technique commonly used to investigate the effect of sampling on species richness (Gotelli and Colwell 2001). The method uses random subsampling of the data to fit a curve (i.e. a rarefaction curve) to the number of species sampled as a function of the number of samples collected (Heck et al. 1975; Hurlbert 1971). During the initial sampling period estimates of richness are generally higher (thus the curve steeper) as common species are sampled. However, as sample sizes increase only rare types remain to be sampled and the slope of the curve trends towards zero (Gotelli and Colwell 2001). For our data, haplotypes were used in place of species allowing application of a supported method to compare haplotype richness (in lieu of species richness). Rarefaction curve data was generated using Vegan (Oksanen et al. 2007) in program R v.3.3.1 (R Core Team 2016). An adjusted value of haplotype richness (HRADJ), based on the smallest sample size, was taken from the rarefaction data. The data was also plotted with SigmaPlot v.12 to allow graphical comparison of ancient and contemporary data sets.
Haplotype (h) and nucleotide (π) diversity were calculated using DNAsp (Librado and
Rozas 2009). Haplotype diversity is the probability that two randomly chosen haplotypes from a sample will be different (Nei 1987). Nucleotide diversity is the probability that the nucleotides at two randomly chosen homologous sites are different (Nei 1987). Nucleotide diversity can quantify how distinct haplotypes are from each other, which is expected to increase when long- standing genetic variation is present. All values of nucleotide diversity are given as π x100.
Samples were grouped both temporally and spatially as well as in combinations of these (Table
35 1). The intermediate Fort Colvile sample was excluded when calculating genetic diversity for pooled ancient and contemporary groupings.
Haplotype networks
Haplotype networks were constructed in program Network version 4.613 (Bandelt et al.
1999). Character weights were set to 20 for indels and 30 for transversions to account for the rarity of such events. Transitions and other characters were left at default weight ten. Three- dimensional haplotype networks were created as described here, followed by modification by hand using graphical software The Fort Colvile sample was excluded from the ancient and contemporary networks (Figure 8) but was included in the three-dimensional network (Figure
5).
Genetic differentiation
The extent of genetic differentiation between ancient and contemporary subgroups was investigated with pairwise comparisons of sequence data (φST) (Excoffier et al. 1992). Analogous to FST, φST indexes the genetic variation in subpopulations relative to that in the total population
(Excoffier et al. 1992; Nei 1982). Higher values of φST indicate the presence of structure between subgroups. The index of φST can be connected to migration, specifically that migration erodes φST (Wright 1949). In comparisons of the ancient and contemporary subgroups, migrants are actually temporal migrants, genetic variants persisting through time, instead of traveling through space.
36 Values of pairwise φST were calculated in Arlequin (Excoffier and Lischer 2010).
Permutation of haplotypes between populations was used to test null hypothesis of no difference between the defined populations where the P-value is the proportion of the permutations (of 5000) that generated a φST statistic equal to or larger than that observed
(Excoffier et al. 1992). Pairwise differences were computed in two ways: (1) pooled ancient samples in each group were compared to pooled contemporary in each group as well as subgroups for both the Columbia and Snake as well as the full species range and (2) the 28 samples from the Columbia River group dated to 3127 YBP. were compared to the contemporary groups and subgroups previously indicated. Quartiles of low, moderate, high, and very high differentiation were defined directly from the data. The intermediate Fort Colvile sample was excluded from the analysis.
2.3.5. Phylogenetic analysis
The survey of Chinook salmon mtDNA haplotypes provided by Martin et al. (2010) indicated distinct haplotypes are present in the northern and southern portions of the range and a southern-to-northern phylogenetic pattern for the species. To explore if any of the novel types sampled fit with the northern or southern clades, we generated a Bayesian cladogram using the haplotypes sampled by Martin et al. (2010) with the novel types sampled for this study. The phylogeny was generated using MrBayes 3.2 (Ronquist and Huelsenbeck 2003). By default, this program ignores positions that include gaps (indels) so these were coded as binary characters and included in the analysis. The three gene regions were set as independent, unlinked partitions. We used the General Time Reversible model (GTR) with a proportion of
37 invariable sites (I), and gamma-shaped distribution of rates across sites (Γ) as determined as the best fitting model using jModelTest (Posada and Crandall 1998). The GTR model is a parameter rich model that allows for varying nucleotide frequencies and substitution rates for transitions and transversions (Tavaré 1986). The program was run for 10E6 generations and sampled every
100th generation. To ensure that the sampling was taken from a stationary posterior distribution, two independent, simultaneous runs were used and the standard deviation of the runs compared. The first 25% of runs were discarded as burn-in and the sampling taken from the remaining runs.
2.3.6. Demographic history
Genetic drift
The genetic information present in any given generation is a mere sample of the parental generation and therefore, subject to sampling effect (i.e. chance). The change in genetic variation in a population due to such chance is known as genetic drift. Drift can lead to complete losses of genetic variants and reduce genetic diversity. To investigate the probability that differences in genetic diversity between the ancient and contemporary groups were a function of genetic drift we applied a simulated model of the Wright-Fisher process.
The drift simulation produced an expectation of haplotype frequencies a contemporary group of samples given the haplotype frequencies in an ancient group of samples, with the effect of sampling over time. An annotated version of the script, written using Perl programming language, is provided in Appendix B and a detailed conceptual outline is provided in Figure 6. In short, an initial “population” is generated with user defined haplotype
38 frequencies and effective population size (Ne), which in this case is the effective female population size. A random sample of size Ne is drawn (with replacement) from this population and becomes the new population, completing one simulated generation. The process is repeated, keeping Ne constant for the number of generations specified and the final output produces the haplotype frequencies expected under a model of genetic drift given the parameters defined.
Independent simulations were conducted for the Columbia, Snake, and Spokane groups using the largest temporal samples for each group and a generation time of four years. The simulation for the Columbia group was initialized using the haplotype frequencies of 28 samples estimated at 3127 YBP. (3127.5 +/- 132.5 YBP) and simulated 795 generations. The
Snake group simulation was based on nine samples aged 700 - 1000 YBP, and simulated 225 generations. The Spokane simulation started with the haplotype frequencies present in the eleven samples dated at 7200 YBP, and simulated 1813 generations. As previously noted, the
Spokane samples lack a direct demographic connection to any contemporary subgroup but are compared to the Columbia subgroups in an effort to compare levels of diversity in a single stock of Chinook salmon sampled in a proximate geographic location. Simulations were repeated
5000 times for each group. Effective population sizes of 100, 500, 1000, 2500, 5000, 10,000,
25,000, 35,000, 45,000, and 55,000 were simulated using 5000 replicates for each Ne.
The model was also used to explore the effect the four smallest Ne values over abbreviated time intervals. Effective population sizes of 100, 500, 1000, and 2500 were modeled for 50, 100, and 200 year intervals for each of the three groups. The abbreviated
39 simulation was implemented to gain an approximation of how quickly diversity could be lost due to drift.
In all cases, haplotype frequencies were converted to diversity (h) using manual application of the equation for haplotype diversity (Nei and Tajima 1981):
푘 푛 ℎ = (1 − ∑ 푥2) 푛 − 1 푖 푖 where n is the sample size, k is the number of haplotypes in the sample, and xi is the frequency of haplotype (i). Probability estimates were developed and used to directly evaluate the null hypothesis under the following arguments:
. H0: The reduction in change in genetic diversity from the ancient to the contemporary
time period is the result of genetic drift.
. The test statistic desired is the probability of obtaining a result equal to or "more
extreme" than that actually observed, given the null hypothesis is true. This can be
rewritten to: the probability of obtaining diversity as low or lower than that actually
observed, given the null hypothesis is true.
. The model of genetic drift utilized here is a simulation in which the null hypothesis is
always true.
. The test statistic can be drawn directly from the simulation as P(D≤d), where D denotes
the estimate of haplotype diversity generated in the random simulation and d denotes the
haplotype diversity calculated for the contemporary group.
. We rejected the null hypothesis of genetic drift when P < 0.05.
40 Extended Bayesian skyline plots
In addition to the forward-in-time drift model, a coalescent model was applied to the data. The coalescent, described by Kingman (1982) is a process of lineages merging (i.e. coalescing) backwards in time until they reach a common ancestor. Coalescent models can be quite powerful, even with limited sample sizes, as long as samples have a common history
(Drummond and Rambaut 2007; Drummond et al. 2012). Bayesian skyline plots (BSPs) use coalescent theory, combined with Markov Chain Monte Carlo (MCMC) algorithms, to quantify the relationship between demographic history and genealogy for a set of sequences
(Drummond and Rambaut 2007). In a BSP model, the effective population size is estimated by fitting a demographic function to a distribution of simulated genealogies (Drummond and
Rambaut 2007; Heled and Drummond 2008).
Coalescent analysis was implemented in BEAST v. 1.8.3 (Drummond and Rambaut 2007).
Two demographic models (1) constant population size (Kingman 1982) and (2) extended
Bayesian skyline plot (EBSP) (Heled and Drummond 2008) were applied to three data sets (1) the Columbia River Group, (2) the contemporary Columbia River group, and (3) the Snake River group. Based on the recorded history for Chinook salmon, we hypothesized that the EBSP models would have greater support than those of constant population size and that one or more of the following scenarios would be evident: (1) a large bottleneck present within the past
200 years for both the Columbia and Snake River sample groups due to post-European development and industrialized fishing (2) a bottleneck event for both the Snake and Columbia
River samples groups approximately 600 years ago in relation to the Bonneville Landslide
(O'connor et al. 2003), and/or (3) a reduction in effective population size around 2000 years ago
41 as a response to changes in Native American fishing methods from single-take to mass capture methods (Galm and Lyman 1988) evident for either the Columbia only due to prevalence of falls capture sites, or for both the Columbia and Snake River sample groups.
Final analyses were performed with the GTR + I + Γ model of sequence evolution (see section 2.3.5) under a strict molecular clock with a mutation rate of 7.5E-9 to 1.0E-8. This rate was based on that proposed for the mitochondrial control region of mtDNA in salmonids by
Shedlock et al. (1992) and Thomas et al. (1986). Population size was modeled with a lognormal distribution with initial value of 10,000, mean 5.0, and standard deviation 2.0. Ages for ancient samples were entered as the radiocarbon date and associated error. In the case of the Snake
River samples where only a range of data is available, the full range was included in the model.
Markov chains were run for 5 X 108 (10% discarded as burn-in) generations, sampled every 5 X
104 generations which generated a final posterior of 10,000 samples. Quality of the posterior distribution was evaluated using the effective sample size (ESS) as determined in Tracer v1.6
(Drummond and Rambaut 2007). ESS quantifies the number of independent data points in the posterior distribution of the genealogies (Drummond et al. 2012). ESS values under 100 are generally considered to be undesirable (Drummond and Rambaut 2007). However Kuhner
(2009) suggests this may be liberal and recommends excluding models with ESS values under
200. Comparisons of models utilized Bayes factor (BF) as determined by differences in marginal likelihood estimates (MLE) (Newton and Raftery 1994; Suchard et al. 2001) calculated in Tracer v.1.6 (Drummond and Rambaut 2007), and described qualitatively using the framework provided in Kass and Raftery (1995):
BF < 2 Not worth more than a bare mention 2 < BF < 6 Positive evidence
42 6 < BF < 10 Strong support BF > 10 Decisive
The final model parameters were based comparison of independent exploratory trials. For the final models, two independent runs of each were performed and posterior estimates taken from the combined Markov chains. EBSPs were plotted to ~10,000 YBP using SigmaPlot v.12.0
(Systat Software, San Jose, CA).
3. RESULTS
3.1. Success rates
Success rates were considerably lower for ancient samples than for contemporary (Table
4). For the ancient samples, the Snake River group had the lowest success rate; only 29% of the samples were successfully sequenced for species. The Columbia group was the most successful, with 63% of the samples successfully sequenced for species. The Spokane samples were intermediate, with 38% of the samples successfully sequenced for species. Success rates for contemporary samples were higher, 85% for the Snake River group and 75% for the Columbia
River group. The rates for the contemporary group are based on DLoop sequencing alone, as no genetic species identification was necessary for these samples.
3.2. Morphometric vs. genetic species identification
Of the 50 ancient Snake River samples successfully sequenced for 12S species identification, more than half (N=26) were determined to be species other than Chinook salmon. These included two coho (Oncorhynchus kisutch), sixteen suckers (Catostomus spp.),
43 and eight minnows (Ptychocheilus spp.). (Table S1). It is of interest that the native Snake River coho salmon, represented by the two Oncorhynchus kisutch vertebrae identified, went extinct in the 1980s and contemporary populations were reintroduced from lower Columbia River populations. Only one other case of a misidentified sample was indicated, a scale sample from
Golden, BC which sequenced as O. mykiss and was likely either misidentified when collected or mislabeled when stored.
Omitting samples genetically identified as non-salmonid, type II vertebrae were correctly identified for 12 of 15 cases (80%) with morphometric methods for both the Columbia and
Snake groups (Table 5). Type III vertebrae were correctly identified for 14 of 14 samples (100%) and for 1 of 5 samples (20%) for the Columbia and Snake groups, respectively. Overall rates were 24 of 30 (80%) for type II vertebrae and 15 of 19 (79%) for type III vertebrae. In total 26 of
29 (90%) of the Columbia samples identified correctly and 13 of 20 (65%) of the Snake samples identified correctly.
As all but two of the samples analyzed in this comparison were genetically determined to be Chinook salmon the success rates for Chinook salmon are similar to those described previously. The Columbia River sample group were all Chinook salmon so the results remain unchanged. For the Snake River group, the modified rates are 10 of 13 (77%) for type II vertebrae, 1 of 5 (20%) for type III vertebrae and 13 of 20 (65%) overall.
3.3. Size estimation
Regression equations developed from data in Hofkamp (2015) were y = 0.1402x + 0.1739
(type II vertebrae) and y = 0.1515x - 0.0669 (type III vertebrae) (Figure 7). Values of R2 were
44 0.90 (type II vertebrae) and 0.91 (type III vertebrae). P values for both type II and type III vertebrae were < 0.000. Using these equations, we estimated fork length for the Columbia
River group to range from 39.8 to 85.4 cm, with an average of 68.4 cm (Table 6). The Snake
River group ranged from 49.1 to 109.5 cm with an average of 77.3 cm. The Spokane River group ranged from 29.5 to 137.8 cm with an average of 95.7 cm.
3.4. Genetic diversity and differentiation
A total of fifteen haplotypes were sampled in the study (Table 1). All haplotypes were separated by five or fewer mutations (mean = 2.8) (Table 3 and Figure 8). Seven novel types were sampled, one exclusive in contemporary samples, five exclusive in ancient samples, and one sampled in both ancient and contemporary samples (Table 1). Novel types were designated using the established nomenclature for Chinook salmon (Martin et al. 2010; Nielsen et al. 1998) and will be deposited in GenBank with upon publication of the study. Comparisons of haplotype networks indicated a reduction in number of haplotypes and shift towards domination by a single haplotype in the Columbia River group (Figure 8). The Spokane (compared to the contemporary Columbia subgroups) had a pattern similar to that in the Columbia with less haplotypes in the contemporary samples than the ancient and a prevalence of haplotype
TSA17. In contrast, the Snake group had more haplotypes present in the contemporary group than in the ancient. However, after rarefaction adjustment to the smallest sample size (N=24) the ancient groups had the highest richness values: 6 (Spokane and Columbia), and 5 (Snake, unadjusted) (Table 1 and Figure 9).
45 All estimates of haplotype and nucleotide (x100) diversity are given in Table 1. In the
Columbia River group, greater diversity is indicated for the ancient samples (h=0.73 and
π=0.196) than for the contemporary samples (h=0.24 and π=0.080, respectively) as well as for all of the five subgroups (ranges of h=0.17 to 0.34 and π=0.055 to 0.119). In the Snake River group, the diversity in ancient samples (h=0.64 and π=0.216) is greater than the contemporary
(h=0.42 and π=0.141) as well as seven of the nine contemporary subgroups. Of the remaining two subgroups, one (Imnaha River) has an equal haplotype diversity (0.64) and a lower nucleotide diversity (0.189). The other contemporary subgroup, Lyons Ferry Hatchery, has greater haplotype and nucleotide diversity than the ancient samples. The Spokane group, representing a single stock of Chinook salmon, had ancient diversity of h=0.52 and π=0.186 which is greater than that present in any of the contemporary Columbia subgroups.
When spatially pooled samples were grouped temporally (Columbia, Snake, and Spokane combined) basin-wide diversity increased from the 4001 – 8000 period to the 151 – 2550 period then declined to the contemporary era (Table 1). Three sets of samples were grouped both spatially and temporally: Snake 700 – 1000 YBP (N=9), Columbia 3127 YBP (N=28), and
Spokane 7200 YBP (N=11). Although the diversity estimates decreased slightly in these subgroups compared to the pooled ancient samples from the same spatial group the comparative results between these subgroups are identical to those described above for each spatial grouping, with higher diversity indicated in the ancient set than in the contemporary counterpart.
46 Genetic differentiation
The majority (68%) of pairwise comparisons for the ancient Columbia to the contemporary subgroups indicated moderate to high differentiation (Table 7). The ancient
Columbia samples were highly differentiated from the contemporary Columbia samples. The ancient Snake samples were less differentiated, with the majority (also 68%) of ancient-to- contemporary comparisons indicating low to moderate differentiation. The ancient Snake River samples were also more similar to their contemporary counterpart, with non-significant and low (φST = 0.04) differentiation indicated. The Spokane group was the most differentiated with
76% of the comparisons to contemporary samples indicating high to very high differentiation.
When the single Columbia River temporal subgroup (3127 YBP) was compared the φST values differed slightly but ultimately fell into the same data quartiles except in two cases. When compared to the upper Salmon (Snake River subgroup) the ancient subgroup was less differentiated (moderate vs. high) than the pooled ancient group and when compared to
Gulkana (species range subgroup) the ancient subgroup was more differentiated (high vs. moderate) than the pooled ancient group. All the ancient samples had low levels of differentiation from each other except for the Spokane and Snake, which were moderately differentiated.
3.5. Phylogenetic analysis of haplotypes
Seven novel haplotypes were identified in this study (Table 1 and Table 3). When these were combined with published haplotypes for Chinook salmon (Martin et al. 2010), two fell into
47 the clade of shared northern and centrally sampled haplotypes and the remaining five into the clade of shared central and southerly sampled haplotypes (Figure 10).
3.6. Demographic history
Genetic drift
All probability estimates for the model of genetic drift are given in Table 8 and outcome distributions of the simulations for each scenario are given in Figure 11. Summary statistics related to the simulations are listed in Table S3. Drift could not be rejected for any of the scenarios with effective population sizes of 100, 500, 1000, or 2500. In the Columbia the model of drift was rejected for effective population sizes of 5000 and higher. For the Snake scenario, drift was rejected for effective population sizes of 35,000 and higher. When the Spokane was compared to the contemporary Columbia subgroups, drift was rejected in all cases of effective population sizes above 25,000 and in three of the five for effective population size 10,000.
When we investigated how quickly drift could reduce diversity given the ancient diversity, short-term effects were not well supported for the Columbia River scenarios (Table 9). The abbreviated model was only supported when NeF was limited to 100 for 200 years. In contrast, for the Snake scenarios, the abbreviated model of drift could not be rejected in all but one case
(NeF of 2500 for 50 years). For the comparisons of the Spokane to the contemporary subgroups, drift was rejected in all of the cases where NeF was greater or equal to 500. For the
NeF of 100, drift was rejected for three of five comparisons over 50 years, but not for any of those over 100 or 200 years.
48 Bayesian inference
Six total models were implemented in BEAST. A model of constant population size was compared to a demographic EBSP model for Columbia and Snake River groups with all temporal samples as well as with ancient samples excluded for the Columbia group. All models described here had ESS values over 200 (Table 10). When only contemporary Columbia River sample data was provided to the model, constant population size had a better fit than the demographic
EBSP model at the “positive” support level (Kass and Raftery 1995). When data from both ancient and contemporary samples was included, the demographic (EBSP) model was more supported and the support level was higher, at the “strong” level. For the Snake River samples
(ancient and contemporary) the constant and EBSP models had equal support. The EBSP plot for the Snake River samples does not indicate a historic population decline and the estimated mean appears to increase over time (Figure 12). The plot of the contemporary Columbia River samples also has a slight increase in the estimated mean population size. However, the plot of the combined ancient and contemporary Columbia River samples does indicate a population decrease starting around 6000 YBP and increasing again after 400 YBP.
4. DISCUSSION
The data presented here is the result of considerable effort to generate a substantial ancient dataset to compare with a contemporary counterpart. However, success rates were mixed. In the case of the Spokane group, success rates were initially so low that a novel PCR method was developed (see Chapter Two). Even with the improved method, species identification was possible for only 38% of the samples in this group. Of the samples genetically
49 confirmed to be from Chinook salmon, only 58% resulted in successful determination of DLoop haplotype. The ancient Snake River group also proved to be challenging, with species determination possible for only 29% of the samples. Of those, just over half were determined to be Chinook salmon and thus able to provide the data sought for the study. As is common in ancient data sets, the final product is a set of relatively small samples sizes with mixed temporal placement and context. Some spatial and/or temporal groups are better represented than others and in several samples, the age of the sample is not well characterized. While we acknowledge limitations exist for both our sampling and mtDNA, we do not feel these compromise our ability to effectively pursue the objectives. Future studies that increase sampling, utilize additional markers, or apply more advanced computational models may expand or modify the conclusions presented here and we look forward to these developments.
4.1. Morphometric vs. genetic species identification
While genetic methods are available to distinguish species from archeological collections, these methods are costly and time-consuming compared to direct morphometric methods.
Further, they require specialized lab space and protocols which may not be available in all cases. Morphometric identification of fish species traditionally relied on cranial elements
(Casteel 1974; Gorschkov et al. 1979). While these methods tend to be accurate, the skeletal elements they rely on are rarely preserved over the long term. In contrast, vertebrae are the most commonly sampled skeletal element in zooarcheological assemblages (Casteel 1977;
Huber et al. 2011). Thus the species identification method developed by Huber et al. (2011), which uses size and shape of vertebrae, is particularly desirable.
50 Under empirical testing, Huber et al. (2011) noted difficulty distinguishing intermediately- sized salmonids steelhead/coho/chum and sockeye/pink. Chinook salmon were usually highly distinguishable, with success rates of 90% for type II vertebrae and 87% for type III vertebrae.
Results from Moss et al. (2014) indicated similar difficulty for intermediately-sized salmonids.
The method correctly identified coho, chum, and sockeye only 57% of the time, a rate statistically indistinguishable from random assignment. However, the assemblage tested did not include Chinook salmon. Applied to our dataset (omitting all samples determined to be non-Oncorhynchus spp.), the model correctly identified 80.9% of the Chinook salmon vertebrae.
This rate supports that indicated for the assemblage tested by Huber et al. (2011). However, the rate varied considerably between groups, with successful morphometric identification in the Columbia group being 23% higher than in the Snake group. Perhaps of note, the assemblage of ancient samples identified in the Huber et al. (2011) study were from a site on the
Wenatchee River, located very near the Columbia River sites described in our study (see map in their Figure 3 and our Figure 1). Most of the misidentifications in the Snake River group were in type III vertebrae. Poor model performance for this type of vertebrae was noted in the previous studies, with the success rates between the two vertebrae types significantly different (Huber et al. 2011; Moss et al. 2014). The results here are interesting because although four of the five type III vertebrae were incorrectly identified in the Snake group, all 14 type III vertebrae in the
Columbia group were correctly identified by the model. It appears that both assemblage and vertebrae type (II vs. III) can impact the success rate of morphometric analysis.
The identification of many non-salmonid samples listed in their collection as salmonid or
“likely salmonid” as well as the ~22% rate of mischaracterization of salmonid species following
51 morphometric methods described by Huber et al. (2011) highlights the importance of pursuing genetic species identification whenever feasible. However, such pursuits are not without risks.
In the case of the Spokane River sample group, amplification of DNA proved to be very difficult with maximum PCR success rates of 6% (Table 4). Of the 45 samples that were successfully sequenced for 12S, 19 were exhausted before DLoop haplotypes could be determined. A limited amount of material may be available in ancient samples. When multiple PCR attempts are necessary to successfully identify species, the amount of DNA available for the remaining genetic target(s) is reduced. In our case, many overlapping fragments were necessary to determine haplotype and thus the risk of sample exhaustion was increased. However, without genetic confirmation, genetic differences between species may be confused as novel haplotypes. We argue that the risk of erroneous results presents a more serious threat than that posed by small sample sizes, and thus, for the inclusion of genetic species identification when possible.
4.2. Size estimation
Historical evidence indicates that a large form of Chinook salmon historically ran in the
Columbia River. These fish, termed “June Hogs” have been reported to weigh as much as 85 pounds (35 kg) or more, with lengths reaching four or five feet (122 to 152 cm) (Harrison 2016;
Seufert and Vaughan 1980; Waknitz et al. 1995). Waknitz et al. (1995) deemed in unlikely that a specific and distant population comprised of large fish existed in the Columbia River. Instead, they considered it more likely that the fish noted were simply very large members of a number of historical populations. Accounts of June Hogs are limited to the Columbia River and we did
52 not expect to see evidence for large fish in the Snake River samples. Further, given the prediction of Waknitz et al. (1995), we expected relative similarity between the upper-Columbia and Spokane River samples. Our prediction for the Snake River was supported, where the largest estimate was 109.5 cm (~3 feet 7 inches) (Table 6). However, of 29 samples from the upper-Columbia the largest estimated length was 85.4 cm (about 2 feet 10 inches), a value surpassed by a majority (5 of 8) of the Spokane samples. Three of these five Spokane samples were estimated to be from fish over 122 cm (four feet) long, therefor qualifying as June Hogs.
Based on our limited data set, it appears that larger fish may have been more common in the
Spokane River than in the upper-Columbia as a whole.
4.3. Tests for correlation of haplotypes with climatic periods
Climate-mediated range shifts have been demonstrated for many species including birds
(Melles et al. 2011), freshwater fish (Hickling et al. 2006), insects, (Karban and Strauss 2004), rodents (Betancourt et al. 1990) and plants (Davis and Shaw 2001). Contemporary data suggests that at least some mitochondrial haplotypes in Chinook salmon are geographically- associated. Of the seventeen published haplotypes for Chinook salmon through their range, two are found only in northern populations, seven in central populations, and two in only southern populations (Martin et al. 2010). Three other haplotypes are shared between northern and central populations and another three between southern and central populations
(Table 11 and Figure 13). This phylogeographic distribution is likely a reflection of post-glacial colonization and subsequent genetic drift. However, it is possible that the distribution also
53 reflects selection on functional regions of mitochondrial DNA; evidence for such a phenomenon has recently been described for Pacific salmon (Garvin et al. 2011).
Climate and stream conditions in the Columbia River Basin have fluctuated dynamically over several millennia and these conditions have been characterized for the past 11,500 years.
Published data confirm three macro climates that overlap our samples: drought conditions with maximum summer warmth (prior to 8000 YBP), transitioning toward moist and cool (8000-4000
YBP), and cool temperatures while slightly drier with late period warming (4000 YBP to contemporary) (Chatters et al. 1995; Galm et al. 2010; Walker and Pellatt 2008). Two specific variables impacting salmon life history (water temperature and stream discharge/flow) have also been described. Prior to 5500 YBP flows were ~30% lower than current and freshet ended in June while from 2300 – 4500 YBP flows were ~30% higher with freshet ending in August
(Chatters et al. 1995). Within the macro climates, four micro periods exist (9000-8000, 7000-
6500, 4400-3900, and 2400-1800 YBP) (Chatters and Hoover 1992). The conditions during these microclimates are characterized by moderate precipitation occurring primarily in winter and warmer than normal winter temperatures. These conditions are similar to those that are expected for the future Columbia Basin under the CO2 doubling scenarios (Chatters and Hoover
1992).
We compared haplotype frequency and phylogeographic haplotype characterizations with the available climate data to test the hypothesis that haplotypes currently associated with southern (warmer) regions of the species range coincide with warmer climate periods, and those that associate with northern (cooler) regions with cooler periods. Such a distribution might be anticipated if the mitochondrial haplotypes are correlated with cold or warm-adapted
54 gene complexes in Chinook salmon or if the haplotypes served as indicators of southern or northern-adapted nuclear gene complexes. No indication of haplotype movement was indicated in either our ancient or contemporary sampling. The phylogeny of all haplotypes, including the novel types, indicated that all of the haplotypes novel to this study correlated with the central based types (Figure 10). Our data does not support any movement of Chinook salmon haplotypes during varying climate periods and instead, that the distribution of mtDNA lineages appears to have been stable over long periods of time.
4.4. Genetic diversity and demography
The current Holocene epoch represents a period of population decline for many species
(Barnosky et al. 2011). Reductions of census size can lead to lower survival, lower reproductive success, and increased inbreeding which erodes genetic diversity (Frankham et al. 2004). Until rather recently, no methods existed to empirically investigate or quantify historical losses in genetic diversity. Instead, genetic histories were extrapolated from data provided by contemporary samples. Methodological and technological advances now permit the study of genetic material extracted from ancient specimens, allowing for more direct characterizations of past dynamics (Butler and Bowers 1998; Leonard 2008; Pääbo et al. 2004). The inclusion of genetic information from ancient samples has confirmed large losses in diversity for many species including musk ox (Campos et al. 2010b), rodents (Conroy and Cook 2000; Hadly et al.
2003), antelope (Campos et al. 2010a) and sea otter (Larson et al. 2002), among others.
However, other investigations have revealed unexpected results. In a study of the genetic history of Scandinavian bears, Bray et al. (2013) found that low genetic diversity in
55 contemporary populations was similar to that in ancient populations, and not the result of demographic declines as previously hypothesized.
Our results reveal contemporary Chinook salmon in parts of the Columbia River basin are genetically depauperate relative to their ancient counterparts. However, the comparisons for genetic diversity were not uniformly distributed through the basin and distinct patterns were present for the groups examined here. Here we briefly summarize the data for each of the three groups included in this study and then explore scenarios that may account for the patterns observed.
4.4.1. Summary by sample group
Columbia River group
Differences between the ancient and contemporary samples in the Columbia River group were stark. Rarified haplotype richness is 50% lower in the contemporary samples than the ancient and contemporary haplotype and nucleotide diversity values are less than half those indicated for the ancient samples. Three haplotypes present in the ancient sampling were not detected in contemporary samples, despite a sample size seven times greater than that in the ancient group. The rarefaction analysis indicated that sampling of the haplotypes was likely close to maximized for contemporary samples. Therefore, it is likely that these types have been lost or are now very rare. One of the unsampled haplotypes (TSA01A) is a component of what
Martin et al. (2010) describes as the evolutionary “backbone” for the species. This type appears to have been lost relatively recently as it was present in the sample dated to 100 YBP (Figure 5).
56 Few temporal migrants appear to be present between the ancient and contemporary components, as indicated by high levels of differentiation (Table 7).
Some amount of bias may exist when contemporaneous (single time period) samples are compared with heterochronous (serial) samples (Depaulis et al. 2009). A majority (N=28) of the ancient Columbia River samples were from a single time period (3127 YBP). Analysis of these samples as a temporal subgroup allowed for temporal comparisons without the associated bias, as well as providing a snapshot of genetic diversity at a specific historical time point. As these samples dominated the total ancient sample for this group, the data quantifications were similar for these as for the pooled ancient samples. Results indicated a loss of 66% of haplotype diversity, and 60% of nucleotide diversity has been realized in the past three millennia.
Drift was not well supported as an explanation for the loss in genetic diversity from 3127
YBP to the contemporary period. We rejected the explanation of drift when the probability estimate was under 5%. By this method, we rejected drift for all effective population sizes
≥5000. The abbreviated models of low effective population size (Table 9) indicated that the effective population size would have to be reduced as low as 100 for more than 100 years for the reduction in genetic diversity observed to be realized (based on the previously indicated 5% probability cutoff). The combined data indicates that it is unlikely that the losses of genetic diversity for this group are only a function of time (drift) and that the demographic influences for the reduction in genetic diversity were either extreme, sustained, or both.
Absolute interpretation of the drift simulation is contingent on choice of effective population size. Originally described by Wright (1931), effective population size is the size of an ideal population which would give the same value for a specified genetic parameter as would
57 the total population. Pre-contact census sizes for Chinook salmon in the basin were between two and four million (Ebel et al. 1989) allowing for the potential for very high effective population sizes. However, in the ‘ideal’ population described by Wright mating and variation in reproductive success is random, generations are discrete, and sex ratios are equal. Therefor accurate estimates of effective population size using only census size are rarely possible (Hare et al. 2011; Waples 1990). Moreover, in this case the measure of effective population size is specifically female effective population size due to the strict maternal inheritance of mtDNA in salmonids. Therefor the results of drift are interpreted as part of the comparative pattern and not as absolute tests of the demographic history.
The Bayesian skyline plot for the ancient Columbia River indicates a reduction in effective population size (NeF) that is initiated around 6000 YBP and continues until approximately 500
YBP (Figure 12). Due to the relatively low, sustained nature of the change indicated we also plotted the rate of this reduction (decrease in NeF per year) (Figure 12). The connection between this rate and the EBSP is discussed further below.
Snake River group
The Snake River sample group also indicated some reduction in genetic diversity from the ancient to the contemporary period. However, the losses in genetic diversity were of less magnitude than that observed in the Columbia group. In contrast to the Columbia River group, more haplotypes were observed in the contemporary samples (N=8) than in the ancient (N=5).
Two haplotypes sampled in the ancient were not sampled in the contemporary. However, the rarefaction analysis of the contemporary Snake did not approach asymptotic maximum,
58 indicating that contemporary sampling may not have been maximized. The rarified comparisons are very similar; with four haplotypes predicted for the contemporary compared to five for the ancient sample set. Diversity comparisons indicated approximately one-third less diversity in the contemporary samples than in the ancient. The dating of the Snake River samples was poor compared to the Columbia River group. The best temporal subgroup was nine samples estimated between 750 and 1050 years old. The diversity in this subgroup indicated a loss of less than 25% over the past ~900 years. A method to correct diversity estimates for pooled contemporaneous data has been described by Depaulis et al. (2009). However, this method requires specific age estimates for all samples and cannot be applied to the Snake River dataset.
Bias from pooling of ancient samples would increase the diversity estimate for the ancient samples. Therefore, adjustment for the bias may reduce the differences indicated for the ancient and contemporary Snake River samples and strengthen the case that the Columbia
River group has experienced a greater change in genetic diversity than the Snake River group.
The ancient and contemporary Snake samples indicated only low, non-significant differentiation
(Table 7), indicating that many temporal migrants are likely to be present in the dataset.
Demographic investigations for the Snake differed from those for the Columbia group as well. Drift was supported up to, and including, the effective population size of 25,000 (Table 8) and reductions in effective population size for only short intervals could reduce the genetic diversity to that observed in the contemporary (Table 9). For example, drift was a plausible explanation for the observed contemporary data when effective population size was reduced to
2500 for only 100 years or to 1000 for only 50 years. The Snake River scenario is based on a sample of nine vertebrae, and we thus must interpret the results with caution. However, the
59 simulation results are supported in the other analysis which utilized greater sampling of the ancient Snake River component. The EBSP is computationally superior to that of the drift simulation and included all the Snake River samples. In this analysis, the model of a constant population size and the EBSP have essentially equal support (BF = 0.65, not worth more than a bare mention) and the plotted EBSP does not indicate a reduction in effective population size at any point in the past 10,000 years.
Spokane River group
The Spokane River group provides insight into the historic genetic composition of a single stock. An equal number of haplotypes (N=6) were sampled from the single-stock Spokane group as were sampled from the mixed-stock ancient Columbia River group. The rarified haplotype richness was the highest of all the groups considered. The diversity in the samples from contemporary Columbia subgroups was, on average, more than 50% lower than that in the
Spokane. The group was more differentiated from the contemporary Columbia samples than the contemporary Snake samples.
4.4.2. Discussion of patterns observed
The Columbia and Snake River samples display contrasting patterns throughout the analyses pursued in this study. This result is somewhat unexpected as the two groups were predicted to share a common history as parts of the larger Columbia River Basin. However, each of the groups contains distinct populations of Chinook salmon which may have divergent demographic histories. The past 200 years represent a period of documented decline for
60 Chinook salmon. However, results from both the simulation of drift and the demographic reconstruction indicate that differences for the Columbia and Snake River samples extend into the pre-contact era. We explore both pre- and post-contact era differences in the history of
Chinook salmon spawning in the upper-Columbia and Snake Rivers in an effort to highlight some of the potential factors possibly contributing to the patterns observed.
Pre-contact
Historically, Native Americans were capturing Chinook salmon from both the Columbia and Snake rivers. However, the cumulative intensity of fishing may have been higher for mid-
Columbia runs than for Snake River runs. Prior to European contact an estimated 61,500 Native
Americans were living in the Columbia River Basin (Hewes 1973). More than half of these individuals lived near the lower or mid-Columbia River (Hewes 1947) with the remainder dispersed throughout the basin (Walker 1967). Celilo Falls and Kettle Falls drew large numbers of Native fisherman during spawning runs. Celilo Falls was likely the most productive single fishing location. As many as 5000 Native American fishermen are estimated to have been present at Celilo Falls annually (Nisbet 2005). Celilo is located downstream of the Snake-
Columbia confluence, therefore catch from this location would have impacted both Columbia and Snake River stocks. Upper-Columbia fish that escaped capture at Celilo Falls encountered the second most productive single fishing location, Kettle Falls. Annually, upwards of 2000 fishermen were present at Kettle Falls (Harrison 2008b; Nisbet 2005) and the take was substantial. For example, it is estimated that the Colville Tribe alone took as much as 600,000 pounds of salmon from Kettle Falls in a single year (Craig and Hacker 1940; Hewes 1947; Teit
61 1930). Additionally, a portion of the upper-Columbia fish were destined for the Spokane River, a location with a substantial fishery (Scholz et al. 2014). As many as 2000 Native Americans (1000 at the Little Spokane River and another 1000 at the Spokane Falls) historically gathered to fish at this location (Scholz et al. 1985). Celilo, Kettle, and Spokane Falls were all productive trading centers where non-local tribes from the Central Plains, Great Lakes, Northwest Coast, and
Southwest would gather to trade for salmon with Columbia basin tribes (Critfc 2016; Nisbet
2005; Scholz et al. 2014). The Snake had many quality fishing sites but lacked mainstem falls capture locations to match the scale of Kettle or Spokane Falls (Freeman and Martin 1954;
Hewes 1947). Although salmon constituted an important food source, for some Snake River tribes this catch had to be substantially supplemented with other hunting and gathering activities (Steward 1938). Non-salmonid fish alternatives may have also been more readily available in the Snake River. Pacific sturgeon (Acipenser transmontanus), a species commonly reaching 100 – 650 pounds (45 – 295 kg) likely formed a substantial component of the diet for
Snake River tribes such as the Nez Perce (Hewes 1947). Other smaller fish species of chub, suckers, and minnows were also heavily consumed (Hewes 1947).
Both the Snake and upper-Columbia River spawning aggregates investigated in this study would have been exploited at Celilo Falls as well as any additional fishing locations between the ocean and their spawning grounds. However, the upper-Columbia aggregates experienced the additional, formidable pressure of the Kettle Falls fishery and potentially the Spokane Falls fishery resulting in increased exploitation relative to those in the Snake sample group. The skyline plot for the ancient Columbia River correlates strongly with the pre-contact hypotheses for Native Americans in the mid- and upper-Columbia. Historical reconstructions for Native
62 Americans at Kettle Falls indicate a notable intensification in salmon exploitation around 4400
YBP (Nps 1980). This trend was then extended with the arrival of Salish speaking people about
2200 years ago (Chance 1986). Population sizes for the Native Americans in the area likely increased rapidly over the next millennium, with maximum levels obtained around 1200 YBP
(Chance 1986; Nps 1980). Around the same time that Salish peoples arrived in the region
(~2000 years ago), Native Americans appear to have transitioned from single-capture fishing methods to mass-capture methods (Galm and Lyman 1988). In the EBSP, a loss in effective population size is initiated around 6000 YBP (Figure 12). The rate of reduction is low until around 4500 YBP when it increases rapidly, reaching maximum at approximately 1000 YBP and then falling to zero (effective population size reaches minimum size and stabilizes) at 500 YBP.
Although the chronology of Kettle Falls fits well with the demographic estimate, however, it is not the only possible explanation. The period of influence for the data presented here stretches thousands of years into the past where, despite extensive efforts by anthropologists, much of the specific history is uncharacterized. The Native Americans living near Kettle Falls were but one component in a suite of potential factors that could have impacted the genetic structure of upper-Columbia Chinook salmon.
There has been some debate to whether Native American exploitation could have significantly depressed salmon in the region (Chapman 1986). Estimates of the total aboriginal catch vary widely and many are dependent on co-estimates of Native American census size
(Craig and Hacker 1940; Hewes 1947; Hewes 1973; Marshall 1977; Walker 1967). Early estimates of the total annual catch by Native Americans in the Columbia River watershed were between 18 (Craig and Hacker 1940) and 22 million pounds (Hewes 1973).The estimate of
63 Hewes (1973) represents a notable 14 to 31% of the 11 to 16-million-member annual salmon run, however it is likely conservative. The most plausible estimate was developed by Schalk
(1986), who included updated information on Native American demography as well as the biology of both salmon and humans. This estimate put the total harvest at almost 42 million salmon annually, as much as half the total run (Schalk 1986). This level of harvest is comparable to that of commercial fishing between 1883 and 1919 (Taylor Iii 2009). Although levels of Native
American exploitation may have rivaled those of commercial fishing, the pattern of implementation was different. While the commercial fishery concentrated all the pressure at the lower river, aboriginal fishing was more largely dispersed. This allowed harvest rates to track with trends in local run sizes (Scholz et al. 2014; Taylor Iii 2009). Such flexibility may have protected the basin-wide salmon resource despite heavy exploitation. In the case of large
Native American populations and expansive trade, it is possible that Native American fishing pressure was substantial enough to influence genetic structure, without initiating catastrophic declines analogous to those documented post-contact. Our data supports pre-contact genetic influence for the upper river fish stocks, represented in the Columbia River samples, where a relatively low, sustained reduction in effective population size is indicated.
Post-contact
Differences in the exploitation realized by upper-Columbia and Snake River stocks continued beyond the aboriginal fishing period. The commercial fishery that developed shortly post-contact was concentrated on the lower river, impacting both upper-Columbia and Snake
River stocks. However, there is evidence that the upper-Columbia stocks were more heavily exploited than Snake River stocks due to morphology and life history characteristics.
64 Commercial fisheries heavily targeted large fish. As runs collapsed, stocks containing large
Chinook salmon were reduced faster than those with smaller counterparts (Beiningen 1976;
Thompson 1951). Fish migrating to the upper-Columbia were commonly 30 to 80 pounds, whereas lower river stocks were considerably smaller (Scholz et al. 2014). The targeting of larger fish depleted upper-Columbia runs faster than lower regions of the Columbia River Basin
(Scholz et al. 2014). The removal of large fish happened quickly, by the 1920s commercial fishing methods had reduced the mean size of Chinook salmon by as much as 50% (Ricker
1981). Run timing may have also been a factor for differences in post-contact exploitation. Fall- run Chinook salmon arrive in freshwater in a more advanced spawning condition making spring and summer-run fish more commercially desirable. Accordingly, the early commercial Chinook salmon fishery focused on spring and summer runs. In the 1880s canneries closed in late July, effectively forgoing any opportunity to capture fall-run Chinook salmon (Smith 1979). As populations declined, commercial exploitation shifted to include fall runs by the early 1920s
(Rich 1942; Smith 1979). Fall runs were prevalent in the Snake River and lower reaches of the
Columbia River. However, longer migrations and differences in mainstem river characteristics likely limited historic upper-Columbia fall run sizes. Though commercial fishery operations were primarily conducted in the shared lower river, their impact on the upper subbasins was likely unequal, with increased pressure on mid and upper-Columbia River aggregates relative to those in the Snake River.
Although fall-run Chinook salmon represented a smaller portion of total stocks in the
Columbia River, these fish may represent an important component of diversity. In the interior
Columbia basin (which includes the sample groups described in this study) all fall-run and all
65 spring-run Chinook salmon populations fall into two distinct genetic lineages (Waples et al.
2008). The divergence of these lineages is believed to have occurred during the Pleistocene
(Mcphail and Lindsey 1986; Waples et al. 2004). The distinct spawning times of the life history is likely to maintain reproductive isolation between the groups, allowing for the accumulation of genetic differences via drift. Evidence for historical inclusion of fall-run Chinook salmon in native fisheries is strong (Seufert and Vaughan 1980). Fall run fish are likely to be represented in the ancient Columbia River sample group. The Grand Coulee Dam Project Area (site
45DO189), which provided 28 samples, was indicated to include fall and potentially winter occupations of Native Americans (Galm and Lyman 1988). Fall Chinook salmon spawn from
September to December and are likely to be present in the sample in the case of either occupation. In contrast, no fall-run fish are included in the contemporary Columbia samples.
When the Grand Coulee Fish Maintenance Project was implemented, an estimated 20 to 34% of
Chinook salmon spawning above Rock Island dam were fall-run (Fish and Hanavan 1948). Fall
Chinook salmon confine their spawning almost exclusively to mainstem rivers or large tributaries (Riggers et al. 2003) and the GCFMP redirection tributaries were not sufficiently large as to meet the spawning requirements of these life histories. As a result, contemporary runs of fall Chinook salmon in the Columbia River are almost exclusively confined to a 90 rkm stretch of lower river known as Hanford Reach.
Previous studies of contemporary Chinook salmon indicate genetic similarity between summer-run and fall-run Chinook salmon in the Columbia River (Brannon et al. 2004; Utter et al. 1995; Waknitz et al. 1995). Summer-run fish were included in the contemporary Columbia samples (N=19). All but one of the samples from this summer-run subset were monomorphic
66 for haplotype TSA17 and the resulting haplotype diversity was 0.11. In contrast, the fall-run subset of Snake River samples (N=24) contained seven haplotypes, with a haplotype diversity of
0.72. Based on our data, there is no evidence for summer runs in the upper-Columbia as stores of historic genetic diversity. However, a broader sampling of summer-run fish from the mid-
Columbia would be useful to further test this hypothesis. Notably, the ancient Columbia River samples display a number of similarities to contemporary fall-run Chinook salmon from both the Snake (Lyons Ferry Hatchery) and Columbia Rivers (Priest Rapids Hatchery) (Martin et al.
2010). Both of these groups represent stocks which are indicated to be genetically similar to summer-run Chinook salmon in the mid-Columbia (Brannon et al. 2004; Utter et al. 1995;
Waknitz et al. 1995). Comparisons of haplotype composition, overall genetic diversity, and differentiation (based on φST) indicate more similarity to the sample of contemporary fall
Chinook salmon than to those in the contemporary tributaries used for redirection. Genetic components have been indicated for functional life history differences, including run timing in
Chinook salmon (Bernier et al. 2008; Campbell and Narum 2008; Hess and Narum 2011; Narum et al. 2007). However, these connections were detected in more complex nuclear DNA markers.
Although untested, it is unlikely that any direct links to mtDNA haplotypes exist. While haplotypes are unlikely to be discretely linked, it is likely that some haplotypes are more common in fall or spring runs due to drift or potentially yet undetected selection as has been indicated for portions of the mtDNA genome in salmonids (Garvin et al. 2011).
Catastrophic census declines for Columbia basin Chinook salmon were documented during the mid-19th century. It is difficult to imagine that these declines came without genetic consequence. Yet, no evidence for a coincident decline in genetic diversity is indicated in the
67 EBSP for either the Columbia or the Snake. Coalescent analyses are powerful tools to elucidate historical demography. However, they are not without limits. Ancient samples can often improve demographic reconstructions. This is evidenced in the Columbia River samples. When all samples were included evidence of a sustained decline in diversity was evident. When the analysis was limited to the data from contemporary samples, patently different results were obtained. No reduction in effective population size was present and the model of constant population size was supported over that of the demographic EBSP (BF = -5.87, “positive support”) (Table 10). However even with heterochronous data, recent events are not always detectable especially when genetic losses are brief, extreme, or occur very near the sampling events (Frankham et al. 2002; Welch et al. 2012).
For Columbia Basin Chinook salmon, demographic inference may also be influenced by more recent management actions. The cumulative impact of 200 years of varying management strategies on basin-wide stocks is not entirely predictable. Potential impacts related to hatchery influence include a reduction in genetic diversity related to bottlenecks, inbreeding, and founder effects (Allendorf and Phelps 1980; Christie et al. 2012; Winans 1989). However, hatcheries may also increase genetic diversity when genetically diverse populations are incorporated (Ciborowski et al. 2007; Nielsen et al. 1994). For example, the samples from Lyons
Ferry Hatchery have the highest estimates of genetic diversity for any of the contemporary subgroups. The broodstock potentially represents a mix of multiple stocks with a common life history. Until 1990, broodstock was obtained from both returns to the hatchery fish ladder as well as adult Chinook salmon trapped at Ice Harbor Dam, which would represent a mix of upriver fall-run spawning aggregates (Bugert et al. 1995; Bugert and Hopley 1991). It is likely
68 that a combination of effects have been realized in a majority of Chinook salmon populations in the Columbia River basin. Many of the historical management practices may confound attempts to approximate the demographic history.
However, we cannot exclude the possibility that the demographic patterns indicated are true representations of the history. Examples of sustained diversity despite periods of largescale population declines have been demonstrated. For example, Hawaiian petrel
(Pterodroma sandwichensis) populations were so reduced in the 1900s that many believed the species to be extinct (Baldwin and Hubbard 1948; Richardson and Woodside 1954). However, comparisons of ancient and contemporary DNA revealed limited losses in genetic diversity and maintenance of effective population size through the period of population decline and recovery
(Welch et al. 2012). In this case, the maintenance was facilitated by high genetic exchange between groups, which could occur in Chinook salmon as a function of straying as well as long generation times (~15 years), a characteristic dissimilar to Chinook salmon (Welch et al. 2012).
The demographic reconstructions fit with the larger patterns for the Columbia and Snake River sample groups and provides evidence of contrasting patterns into the pre-contact period.
4.5. Notes on reintroduction
An inter-agency investigation into the feasibility of reintroducing anadromous salmon above the Chief Joseph and Grand Coulee Dams is currently being pursued (Hanrahan et al.
2001; Ucut 2015; Us Columbia Basin Tribes and Canadian First Nations 2014). One element of this evaluation is the identification of candidate fish stocks for use in reintroduction. We caution against extrapolating the data presented in this study to inferences about successful
69 spawning and recruitment. In the differentiation comparisons (φST), the populations that are the most similar are not necessarily those located most proximate to the potential reintroduction area (Table 7). However, the majority of contemporary management strategies recognize the importance of utilizing geographically proximate stocks. A large body of evidence exists for local adaptation in salmonids (Fraser et al. 2011; Grant 1995). Goals related to strictly to levels of genetic diversity ignore the relationship between genes and the environment demonstrated for salmonids. Goals related to differentiation may be inadvisable as well.
Studies of Atlantic salmon data indicate that sub-populations with minimal differentiation (FST
= 0.018) can still have largely varying life histories including differences in growth rate, maturation and migration timing (Aykanat et al. 2015). The comparisons in this study examine genetic similarity between ancient samples and contemporary stocks both in the Columbia
River Basin and in the broader species range. However, genetic similarity does not necessarily translate to genetic suitability for a particular environment.
Instead, reintroduction may present an interesting opportunity in relation to the historic upper-Columbia stocks. When novel habitat becomes available, colonizers into that habitat form new populations. This population expansion can buoy genetic diversity, even over short time scales (De Bruyn et al. 2009). This phenomenon has been directly indicated for Chinook salmon when a single population was introduced to New Zealand developed distinct life history differences in 10 years (Quinn et al. 2001) and discernable population structure in 30 years
(Kinnison et al. 2002; Quinn et al. 2001). It may be that the comparative value of this mtDNA survey is not in stock selection but in identifying a way in which reintroduction efforts can most benefit the descendants of the upper-Columbia stocks. Under a careful management strategy,
70 reintroduction of stocks from the GCFMP tributaries or fall-run Chinook salmon from the lower river may have the long-term result of increasing the overall genetic diversity in the combined mid- and upper-Columbia basin.
4.6. Summary
The data generated for this study was relevant to questions related to a better understanding of culture history, climate change, conservation, and history. Comparisons of genetic and morphometric methods to distinguish salmonid species from zooarcheological assemblages provided support for using the genetic method whenever possible. The Spokane samples correlated to the largest estimates of body size (fork length). Three of the vertebrae were estimated to come from fish over 122 cm in length, classifying them as June Hogs. No such lengths were estimated from the ancient Columbia or Snake samples. Haplotype distributions did not indicate evidence for climate-mediated migration. The majority of the analysis focused on comparisons of genetic diversity in ancient and contemporary sample sets.
Our data indicates that for the Columbia, Snake and Spokane Rivers ancient samples are more diverse than contemporary counterparts (with the Spokane compared to the Columbia subgroups). In all cases, the differences between ancient and contemporary samples was greater for the Columbia than for the Snake. We hypothesize that these differences are the result of cumulative effects of pre- and post-contact exploitation along with direct losses of stocks and life-history variants coincident to the construction of the Grand Coulee Dam. Our data provided no direct evidence that large-scale genetic changes are tied to any event in recent history, although empirical sampling of genetic data near these events was not possible.
71 Future investigations may be advanced in several ways. Potentially beneficial would strategies include: (1) increased sampling of ancient single-stocks that have directly comparable contemporary counterparts, (2) the application of more advanced demographic models, and (3) sampling very near the period of European arrival and development.
The sampling of additional ancient stocks could help elucidate the influence of population structure on genetic diversity and demographic reconstructions. In our dataset we have no means by which to separate the larger metapopulations represented by the Columbia and
Snake River data sets. However, evidence from the Spokane sample group does indicate that important components of diversity are contained in single stocks. Additional characterizations of single-stocks would allow for this hypothesis to be further evaluated.
The results presented here indicate that both pre- and post-contact events have influenced the genetic structure of Chinook salmon. However, the demographic reconstructions are limited in scope. Application of Approximate Bayesian Computation (ABC) using programs such as Serial SimCoal (Anderson et al. 2005) could be used to explore specific hypotheses and compare the likelihood of multiple scenarios given the genetic data. For example, one can test the hypothesis of a bottleneck of size X at time Y due to event Z and then compare the fit of that model to the data to the fit of other hypothesized scenarios. These models provide means to estimate effective population size and explore demographic events essential to understanding the genetic history of samples (Drummond and Rambaut 2007; Drummond et al.
2012). Previous studies have successfully implemented similar methods to identify factors that influenced genetic history and contemporary structure (Chan et al. 2006; Pinsky et al. 2010;
Valentine et al. 2008). For our dataset, ABC analysis may be a means to deal with the complex
72 history for Chinook salmon in the Columbia River Basin. Unlike Bayesian Skyline Plots implemented in BEAST, programs such as Serial SimCoal are not limited by assumptions of simple demographic processes or a panmixia (Anderson et al. 2005).
Perhaps most valuable for understanding the history of Chinook salmon in the Columbia
River Basin is the inclusion of temporally-distinguished samples targeted to the initial period of
European arrival and development (~AD 1800 – 1900) and the period of hydroelectric dam construction (~AD 1930 – 1950). Unlike the modelling methods discussed previously, these samples may allow for the empirical extrication of the impacts of events which occurred in the near past. We attempted to include samples such as these in our analysis. The samples, formalin preserved and dried specimens, proved technologically challenging and attempts at extraction of genetic information were unsuccessful. However, future technologies or the identification of additional sample sources may allow for the inclusion of samples from this time period.
73 LITERATURE CITED
Ackerman, M., J. McCane, C. Steele, M. Campbell, A. Matala, J. Hess, and S. Narum. 2012. Chinook and steelhead genotyping for genetic stock identification at Lower Granite Dam. Idaho Department of Fish and Game Report:12-15.
Allendorf, F. W., and S. R. Phelps. 1980. Loss of genetic variation in a hatchery stock of cutthroat trout. Transactions of the American Fisheries Society 109(5):537-543.
Anderson, C. N. K., U. Ramakrishnan, Y. L. Chan, and E. A. Hadly. 2005. Serial SimCoal: a population genetics model for data from multiple populations and points in time. Bioinformatics 21(8):1733-1734.
Aykanat, T., S. E. Johnston, P. Orell, E. Niemelä, J. Erkinaro, and C. R. Primmer. 2015. Low but significant genetic differentiation underlies biologically meaningful phenotypic divergence in a large Atlantic salmon population. Molecular ecology 24(20):5158-5174.
Baldwin, P., and D. Hubbard. 1948. The Hawaiian dark-rumped petrel reappears on Hawaii. Condor 51(5):231-232.
Bandelt, H.-J., P. Forster, and A. Röhl. 1999. Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16(1):37-48.
Barman, J., and B. M. Watson. 1999. Fort Colvile's Fur Trade Families and the Dynamics of Race in the Pacific Northwest. The Pacific Northwest Quarterly 90(3):140-153.
Barnosky, A. D., N. Matzke, S. Tomiya, G. O. Wogan, B. Swartz, T. B. Quental, C. Marshall, J. L. McGuire, E. L. Lindsey, and K. C. Maguire. 2011. Has the Earth/'s sixth mass extinction already arrived? Nature 471(7336):51-57.
Barta, J. L., C. Monroe, S. J. Crockford, and B. M. Kemp. 2014a. Mitochondrial DNA preservation across 3000 year old northern fur seal ribs is not related to bone density: Implications for forensic investigations. Forensic Science International 239:11-18.
Barta, J. L., C. Monroe, and B. M. Kemp. 2013. Further evaluation of the efficacy of contamination removal from bone surfaces. Forensic Science International 231(1):340- 348.
Barta, J. L., C. Monroe, J. E. Teisberg, M. Winters, K. Flanigan, and B. M. Kemp. 2014b. One of the key characteristics of ancient and forensic DNA, low copy number, may be a product of its extraction. Journal of Archaeological Science 46:281-289.
Beiningen, K. T. 1976. Fish runs. Investigative Reports of Columbia River Fisheries Project. Pacific Northwest Regional Commission, Vancouver, WA (Section E).
74 Bernier, J. C., S. R. Birkeland, M. J. Cipriano, A. G. McArthur, and M. A. Banks. 2008. Differential Gene Expression between Fall- and Spring-Run Chinook Salmon Assessed by Long Serial Analysis of Gene Expression. Transactions of the American Fisheries Society 137(5):1378-1388.
Betancourt, J. L., T. R. Van Devender, and P. S. Martin. 1990. Packrat middens: the last 40,000 years of biotic change. University of Arizona Press.
Boas, F. 1894. Chinook texts. US Government Printing Office.
Brannon, E. L., M. S. Powell, T. P. Quinn, and A. Talbot. 2004. Population structure of Columbia River Basin chinook salmon and steelhead trout. Reviews in Fisheries Science 12(2-3):99- 232.
Brauner, D. 1990. Archaeological data recovery at Hatiuhpuh, 45WT134, Whitman County, Washington. Department of Anthropology, Oregon State University.
Bray, S. C., J. J. Austin, J. L. Metcalf, K. Østbye, E. Østbye, S. E. Lauritzen, K. Aaris‐Sørensen, C. Valdiosera, C. J. Adler, and A. Cooper. 2013. Ancient DNA identifies post‐glacial recolonisation, not recent bottlenecks, as the primary driver of contemporary mtDNA phylogeography and diversity in Scandinavian brown bears. Diversity and Distributions 19(3):245-256.
Bugert, R., C. Hopley, C. Busack, and G. Mendel. 1995. Maintenance of stock integrity in Snake River fall Chinook salmon. American Fisheries Society Symposium. 1995.
Bugert, R., and W. Hopley. 1991. Fall chinook salmon trapping on the Snake River in 1990. Completion report, Cooperative Agreement:14-16.
Butler, V., and N. J. Bowers. 1998. Ancient DNA from salmon bone: a preliminary study. Ancient Biomolecules 2:17-26.
Butler, V. L. 2006. Fish Remains from the Spokane Site (45SP266). Eastern Washington University, Cheney, Washington.
Butler, V. L., and J. E. O'Connor. 2004. 9000 years of salmon fishing on the Columbia River, North America. Quaternary Research 62(1):1-8.
Campbell, N. R., and S. R. Narum. 2008. Identification of novel single-nucleotide polymorphisms in Chinook salmon and variation among life history types. Transactions of the American Fisheries Society 137(1):96-106.
Campos, P. F., T. Kristensen, L. Orlando, A. Sher, M. V. Kholodova, A. Götherström, M. Hofreiter, D. G. Drucker, P. Kosintsev, and A. Tikhonov. 2010a. Ancient DNA sequences point to a large loss of mitochondrial genetic diversity in the saiga antelope (Saiga tatarica) since the Pleistocene. Molecular Ecology 19(22):4863-4875.
75 Campos, P. F., E. Willerslev, A. Sher, L. Orlando, E. Axelsson, A. Tikhonov, K. Aaris-Sørensen, A. D. Greenwood, R.-D. Kahlke, and P. Kosintsev. 2010b. Ancient DNA analyses exclude humans as the driving force behind late Pleistocene musk ox (Ovibos moschatus) population dynamics. Proceedings of the National Academy of Sciences 107(12):5675- 5680.
Casteel, R. W. 1974. Identification of the species of Pacific salmon (genus Oncorhynchus) native to North America based upon otoliths. Copeia:305-311.
Casteel, R. W. 1977. Fish remains in archaeology and paleo-environmental studies.
Chan, Y. L., C. N. K. Anderson, and E. A. Hadly. 2006. Bayesian estimation of the timing and severity of a population bottleneck from ancient DNA. PLoS genetics 2(4).
Chance, D. H. 1986. People of the Falls. Kettle Falls Historical Center.
Chance, D. H., and J. V. Chance. 1982. Kettle Falls: 1971 and 1974 Salvage Archaeology in Lake Roosevelt.
Chapman, D. W. 1986. SALMON AND STEELHEAD ABUNDANCE IN THE COLUMBIA RIVER IN THE 19TH-CENTURY. Transactions of the American Fisheries Society 115(5):662-670.
Chatters, J. C., V. L. Butler, M. J. Scott, D. M. Anderson, and D. A. Neitzel. 1995. A paleoscience approach to estimating the effects of climatic warming on salmonid fisheries of the Columbia River basin. R. Beamish, editor. Climate change and northern fish populations. Canadian Special Publication of Fisheries and Aquatic Science.
Chatters, J. C., and K. A. Hoover. 1992. Response of the Columbia River fluvial system to Holocene climatic change. Quaternary Research 37(1):42-59.
Christie, M. R., M. L. Marine, R. A. French, and M. S. Blouin. 2012. Genetic adaptation to captivity can occur in a single generation. Proceedings of the National Academy of Sciences 109(1):238-242.
Ciborowski, K., S. Consuegra, C. G. De Leániz, J. Wang, M. Beaumont, and W. Jordan. 2007. Stocking may increase mitochondrial DNA diversity but fails to halt the decline of endangered Atlantic salmon populations. Conservation Genetics 8(6):1355-1367.
Cobb, J. N. 1921. Pacific salmon fisheries. US Government Printing Office.
Conroy, C. J., and J. A. Cook. 2000. Phylogeography of a post‐glacial colonizer: Microtus longicaudus (Rodentia: Muridae). Molecular Ecology 9(2):165-175.
Cooper, A. 1994. DNA from museum specimens. Pages 149-165 in Ancient DNA. Springer.
76 Craig, J. A., and R. L. Hacker. 1940. The history and development of the fisheries of the Columbia River. US Government Printing Office.
CRITFC. 2016. Celilo Falls. Columbia River Inter-Tribal Fish Commission, Portland, OR.
Crozier, L. G., A. P. Hendry, P. W. Lawson, T. P. Quinn, N. J. Mantua, J. Battin, R. G. Shaw, and R. B. Huey. 2008. Potential responses to climate change in organisms with complex life histories: evolution and plasticity in Pacific salmon. Evolutionary Applications 1(2):252- 270.
Cui, Y., J. Lindo, C. E. Hughes, J. W. Johnson, A. G. Hernandez, B. M. Kemp, J. Ma, R. Cunningham, B. Petzelt, and J. Mitchell. 2013. Ancient DNA analysis of mid-holocene individuals from the Northwest Coast of North America reveals different evolutionary paths for mitogenomes. PLoS One 8(7):e66948.
Dart, A. 1851. Report of the Commissioner of Indian Affairs. Department of the Interior, Washington D.C.
Daugherty, R. D., B. A. Purdy, and R. Fryxell. 1967. The descriptive archaeology and geochronology of the Three Springs Bar archaeological site, Washington. Washington State University Laboratory of Anthropology.
Davis, M. B., and R. G. Shaw. 2001. Range shifts and adaptive responses to Quaternary climate change. Science 292(5517):673-679.
Davis, O. K., J. C. Sheppard, and S. Robertson. 1986. Contrasting climatic histories for the Snake River Plain, Idaho, resulting from multiple thermal maxima. Quaternary Research 26(3):321-339.
De Bruyn, M., B. L. Hall, L. F. Chauke, C. Baroni, P. L. Koch, and A. R. Hoelzel. 2009. Rapid response of a marine mammal species to Holocene climate and habitat change. PLoS Genet 5(7):e1000554.
Depaulis, F., L. Orlando, and C. Hänni. 2009. Using classical population genetics tools with heterochroneous data: time matters! PloS one 4(5):e5541.
Drummond, A. J., and A. Rambaut. 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC evolutionary biology 7(1):1.
Drummond, A. J., M. A. Suchard, D. Xie, and A. Rambaut. 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29(8):1969-1973.
Ebel, W. J., C. D. Becker, J. W. Mullan, and H. L. Raymond. 1989. The Columbia River-- toward a holistic understanding. Canadian Special Publication of Fisheries and Aquatic Sciences.
77 Excoffier, L., and H. E. Lischer. 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10(3):564-567.
Excoffier, L., P. E. Smouse, and J. M. Quattro. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131(2):479-491.
Fish, F. F., and M. G. Hanavan. 1948. A report upon the Grand Coulee fish-maintenance project 1939-1947. US Fish and Wildlife Service.
Frankham, R., J. D. Ballou, and D. A. Briscoe. 2004. A primer of conservation genetics. Cambridge University Press.
Frankham, R., D. A. Briscoe, and J. D. Ballou. 2002. Introduction to conservation genetics. Cambridge University Press.
Fraser, D. J., L. K. Weir, L. Bernatchez, M. M. Hansen, and E. B. Taylor. 2011. Extent and scale of local adaptation in salmonid fishes: review and meta-analysis. Heredity 106(3):404-420.
Freeman, O. W., and H. H. Martin. 1954. The Pacific Northwest. Soil Science 78(5):407.
Galm, J., T. Fulkerson, Y. Zhupikov, M. Glantz, and R. Suleimanov. 2010. Late Holocene Climate Change: An Examination of Proxy Data Sets From the Pacific Northwest and Aral Sea Basin. The 63rd Annual Northwest Anthropological Conference.
Galm, J. R. 1994. A design for management of cultural resources in the Lake Roosevelt basin of northeastern Washington, volume 100. Archaeological and Historical Services, Eastern Washington University.
Galm, J. R., and R. L. Lyman. 1988. Archaeological Investigations at River Mile 590: The Excavations at 45DO189. Eastern Washington University Archaeological and Historical Services.
Garvin, M. R., J. P. Bielawski, and A. J. Gharrett. 2011. Positive Darwinian selection in the piston that powers proton pumps in complex I of the mitochondria of Pacific salmon. PLoS One 6(9):e24127.
Gilbert, M. T. P., H.-J. Bandelt, M. Hofreiter, and I. Barnes. 2005. Assessing ancient DNA studies. Trends in Ecology & Evolution 20(10):541-544.
Gorschkov, S., Y. Dorofeyeva, V. Klyukanov, and N. Kulikova. 1979. Osteological characteristics of the Pacific salmons in the genus Oncorhynchus. Journal of Ichthyology 19:1-19.
Gotelli, N. J., and R. K. Colwell. 2001. Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecology letters 4(4):379-391.
78 Grant, W. S. 1995. Genetic effects of straying of non-native hatchery fish into natural populations. Proceedings of the NOAA workshop.
Grier, C., K. Flanigan, M. Winters, L. G. Jordan, S. Lukowski, and B. M. Kemp. 2013. Using Ancient DNA Identification and Osteometric Measures of Archaeological Pacific Salmon Vertebrae for Reconstructing Salmon Fisheries and Site Seasonality at Dionisio Point, British Columbia. Journal of Archaeological Science 40:544-555.
Hadly, E. A., M. van Tuinen, Y. Chan, and K. Heiman. 2003. Ancient DNA evidence of prolonged population persistence with negligible genetic diversity in an endemic tuco-tuco (Ctenomys sociabilis). Journal of Mammalogy 84(2):403-417.
Halffman, C. M., B. A. Potter, H. J. McKinney, B. P. Finney, A. T. Rodrigues, D. Y. Yang, and B. M. Kemp. 2015. Early human use of anadromous salmon in North America at 11,500 y ago. Proc Natl Acad Sci U S A 112(40):12344-8.
Handt, O., M. Höss, M. Krings, and S. Pääbo. 1994. Ancient DNA: Methodological challenges. Experientia 50(6):524-529.
Hanrahan, T. P., D. D. Dauble, and D. R. Geist. 2001. An assessment of potential chinook salmon spawning habitat in the upper Columbia River: Chief Joseph Dam to Grand Coulee Dam. Final report to the Colville Confederated Tribes.
Hare, M. P., L. Nunney, M. K. Schwartz, D. E. Ruzzante, M. Burford, R. S. Waples, K. Ruegg, and F. Palstra. 2011. Understanding and estimating effective population size for practical application in marine species management. Conservation Biology 25(3):438-449.
Harrison, J. 2008a. Dams: impacts on salmon and steelhead. Northwest Power and Conservation Council Columbia River History Project.
Harrison, J. 2008b. Indian Fishing. Northwest Power and Conservation Council Columbia River History Project.
Harrison, J. 2016. June Hogs. The Oregon Encyclopedia. Portland State University and the Oregon Historical Society.
Healey, M., C. Groot, and L. Margolis. 1991. Life history of chinook salmon (Oncorhynchus tshawytscha). Pacific salmon life histories:311-394.
Heck, K. L., G. van Belle, and D. Simberloff. 1975. Explicit calculation of the rarefaction diversity measurement and the determination of sufficient sample size. Ecology 56(6):1459-1461.
Heled, J., and A. J. Drummond. 2008. Bayesian inference of population size history from multiple loci. BMC Evolutionary Biology 8(1):1.
79 Hess, J. E., and S. R. Narum. 2011. Single-nucleotide polymorphism (SNP) loci correlated with run timing in adult Chinook salmon from the Columbia River basin. Transactions of the American Fisheries Society 140(3):855-864.
Hewes, G. W. 1947. Aboriginal use of fishery resources in northwestern North America. University of California, Berkeley.
Hewes, G. W. 1973. Indian fisheries productivity in pre-contact times in the Pacific salmon area. Northwest Anthropological Research Notes 7(2):133-155.
Hickling, R., D. B. Roy, J. K. Hill, R. Fox, and C. D. Thomas. 2006. The distributions of a wide range of taxonomic groups are expanding polewards. Global change biology 12(3):450-455.
Hicks, B. A. 2004. Marmes Rockshelter: A final report on 11,000 years of cultural use. Washington State University.
Hofkamp, A. R. 2015. Age Determination of Modern and Archaeological Chinook Salmon (Oncorhynchus tshawytcha) Using Vertebrae.
Huber, H. R., J. C. Jorgensen, V. L. Butler, G. Baker, and R. Stevens. 2011. Can salmonids (Oncorhynchus spp.) be identified to species using vertebral morphometrics? Journal of Archaeological Science 38(1):136-146.
Hurlbert, S. H. 1971. The nonconcept of species diversity: a critique and alternative parameters. Ecology 52(4):577-586.
Jordan, L. G., C. A. Steele, and G. H. Thorgaard. 2010. Universal mtDNA primers for species identification of degraded bony fish samples. Molecular Ecology Resources 10(3):581- 581.
Karban, R., and S. Y. Strauss. 2004. Physiological tolerance, climate change, and a northward range shift in the spittlebug, Philaenus spumarius. Ecological Entomology 29(2):251-254.
Kass, R. E., and A. E. Raftery. 1995. Bayes factors. Journal of the american statistical association 90(430):773-795.
Kemp, B. M., R. S. Malhi, J. McDonough, D. A. Bolnick, J. A. Eshleman, O. Rickards, C. Martinez‐ Labarga, J. R. Johnson, J. G. Lorenz, and E. J. Dixon. 2007. Genetic analysis of early Holocene skeletal remains from Alaska and its implications for the settlement of the Americas. American journal of physical anthropology 132(4):605-621.
Kemp, B. M., C. Monroe, K. G. Judd, E. Reams, and C. Grier. 2014. Evaluation of methods that subdue the effects of polymerase chain reaction inhibitors in the study of ancient and degraded DNA. Journal of Archaeological Science 42:373-380.
80 Kemp, B. M., C. Monroe, and D. G. Smith. 2006. Repeat silica extraction: A simple technique for the removal of PCR inhibitors from DNA extracts. Journal of Archaeological Science 33(12):1680-1689.
Kemp, B. M., and D. G. Smith. 2005. Use of bleach to eliminate contaminating DNA from the surface of bones and teeth. Forensic Science International 154(1):53-61.
Kemp, B. M., and D. G. Smith. 2010. Ancient DNA methodology: Thoughts from Brian M. Kemp and David Glenn Smith on “Mitochondrial DNA of protohistoric remains of an Arikara population from South Dakota”. Human Biology 82(2):227-238.
Kemp, B. M., T. A. Tung, and M. L. Summar. 2009. Genetic continuity after the collapse of the Wari empire: Mitochondrial DNA profiles from Wari and post‐Wari populations in the ancient Andes. American journal of physical anthropology 140(1):80-91.
Kenaston, M. R. 1966. The archaeology of the Harder Site, Franklin County, Washington. Washington State University Laboratory of Anthropology.
Kingman, J. F. 1982. On the genealogy of large populations. Journal of Applied Probability:27- 43.
Kinnison, M., P. Bentzen, M. Unwin, and T. Quinn. 2002. Reconstructing recent divergence: evaluating nonequilibrium population structure in New Zealand chinook salmon. Molecular Ecology 11(4):739-754.
Kuhner, M. K. 2009. Coalescent genealogy samplers: windows into population history. Trends in Ecology & Evolution 24(2):86-93.
Lang, B. 2011. Columbia River. Center for Columbia River History.
Larson, S., R. Jameson, M. Etnier, M. Fleming, and P. Bentzen. 2002. Loss of genetic diversity in sea otters (Enhydra lutris) associated with the fur trade of the 18th and 19th centuries. Molecular Ecology 11(10):1899-1903.
Leonard, J. A. 2008. Ancient DNA applications for wildlife conservation. Molecular Ecology 17(19):4186-4196.
Leonhardy, F. C. 1970. Artifact Assemblages and Archaeological Units at Granite Point Locality 1 (45WT41), Southeastern Washington. Washington State University.
Librado, P., and J. Rozas. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25(11):1451-1452.
Lichatowich, J. A., and L. E. Mobrand. 1995. Analysis of chinook salmon in the Columbia River from an ecosystem perspective. Prepared for Bonneville Power Administration.
81 Lyman, R. L. 2006. Zooarchaeology of 45SP266. Eastern Washington University, Cheney, WA.
Lyman, R. L. 2012. Human-behavioral and paleoecological implications of terminal Pleistocene fox remains at the Marmes Site (45FR50), eastern Washington state, USA. Quaternary Science Reviews 41:39-48.
Marshall, A. G. 1977. Nez Perce social groups: An ecological interpretation.
Martin, K. E., C. A. Steele, J. P. Brunelli, and G. H. Thorgaard. 2010. Mitochondrial variation and biogeographic history of Chinook salmon. Transactions of the American Fisheries Society 139(3):792-802.
McCord, B., A. Pionzio, and R. Thompson. 2015. Analysis of the Effect of a Variety of PCR Inhibitors on the Amplification of DNA using RealTimePCR, MeltCurves and STR Analysis. U.S. Department of Justice.
McPhail, J., and C. Lindsey. 1986. Zoogeography of the freshwater fishes of Cascadia (the Columbia system and rivers north to the Stikine). The zoogeography of North American freshwater fishes:615-637.
Melles, S., M. J. FORTIN, K. Lindsay, and D. Badzinski. 2011. Expanding northward: influence of climate change, forest connectivity, and population processes on a threatened species' range shift. Global Change Biology 17(1):17-31.
Monroe, C., C. Grier, and B. M. Kemp. 2013. Evaluating the efficacy of various thermo-stable polymerases against co-extracted PCR inhibitors in ancient DNA samples. Forensic Science International 228(1):142-153.
Moss, M. L., K. G. Judd, and B. M. Kemp. 2014. Can salmonids (Oncorhynchus spp.) be identified to species using vertebral morphometrics? A test using ancient DNA from Coffman Cove, Alaska. Journal of Archaeological Science 41:879-889.
Moss, M. L., D. Y. Yang, S. D. Newsome, C. F. Speller, I. McKechnie, A. D. McMillan, R. J. Losey, and P. L. Koch. 2006. Historical ecology and biogeography of North Pacific pinnipeds: Isotopes and ancient DNA from three archaeological assemblages. Journal of Island & Coastal Archaeology 1(2):165-190.
Nakonechny, L. D. 1998. Archaeological analysis of Area A, Wexpusnime site (45GA61).
Narum, S. R., J. J. Stephenson, and M. R. Campbell. 2007. Genetic variation and structure of Chinook salmon life history types in the Snake River. Transactions of the American Fisheries Society 136(5):1252-1262.
Nei, M. 1982. Evolution of human races at the gene level. Human genetics, part A: the unfolding genome. Alan R. Liss, New York 167.
82 Nei, M. 1987. Molecular evolutionary genetics. Columbia university press.
Nei, M., and F. Tajima. 1981. DNA polymorphism detectable by restriction endonucleases. Genetics 97(1):145-163.
Netboy, A. 1980. The Columbia River salmon and steelhead trout, their fight for survival.
Newton, M. A., and A. E. Raftery. 1994. Approximate Bayesian inference with the weighted likelihood bootstrap. Journal of the Royal Statistical Society. Series B (Methodological):3- 48.
Nielsen, J. L., M. C. Fountain, J. C. Favela, K. Cobble, and B. L. Jensen. 1998. Oncorhynchus at the southern extent of their range: a study of mtDNA control–region sequence with special reference to an undescribed subspecies of O. mykiss from Mexico. Environmental Biology of Fishes 51(1):7-23.
Nielsen, J. L., C. Gan, and W. K. Thomas. 1994. Differences in genetic diversity for mitochondrial DNA between hatchery and wild populations of Oncorhynchus. Canadian Journal of Fisheries and Aquatic Sciences 51(S1):290-297.
Nisbet, J. 2005. The mapmaker's eye: David Thompson on the Columbia Plateau. Pullman: Washington State University Press.
NPS. 1980. General Management plan Coulee Dam National Recreation Area. Pages 39 in N. P. Service, editor.
O'Connor, J. E., G. E. Grant, and T. L. Haluska. 2003. Overview of geology, hydrology, geomorphology, and sediment budget of the Deschutes River basin, Oregon. A Peculiar River:7-29.
Oksanen, J., R. Kindt, P. Legendre, B. O’Hara, M. H. H. Stevens, M. J. Oksanen, and M. Suggests. 2007. The vegan package. Community ecology package 10.
Pääbo, S., H. Poinar, D. Serre, V. Jaenicke-Després, J. Hebler, N. Rohland, M. Kuch, J. Krause, L. Vigilant, and M. Hofreiter. 2004. Genetic analyses from ancient DNA. Annual Review of Genetics 38:645-679.
Pinsky, M. L., S. D. Newsome, B. R. Dickerson, Y. Fang, M. van Tuinen, D. J. Kennett, R. Ream, and E. A. Hadly. 2010. Dispersal provided resilience to range collapse in a marine mammal: insights from the past to inform conservation biology. Molecular Ecology 19(12):2418-2429.
Posada, D., and K. A. Crandall. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14(9):817-818.
83 Pouley, J. O. 2008. Analysis of the Kettle Falls culture chronology temporal gaps. Archeology in Washington 14:3-20.
Quinn, T. P., M. T. Kinnison, and M. J. Unwin. 2001. Evolution of chinook salmon (Oncorhynchus tshawytscha) populations in New Zealand: pattern, rate, and process. Genetica 112(1):493-513.
R Core Team. 2016. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.
Rice, H. S. 1965. The cultural sequence at Windust Caves. Washington State University.
Rich, W. H. 1942. The salmon runs of the Columbia River in 1938. Salem, Or.: Department of Research, Fish Commission of the State of Oregon.
Richardson, F., and D. H. Woodside. 1954. Rediscovery of the nesting of the Dark-rumped Petrel in the Hawaiian Islands. The Condor 56(6):323-327.
Ricker, W. 1981. Changes in the average size and average age of Pacific salmon. Canadian Journal of Fisheries and Aquatic Sciences 38(12):1636-1656.
Riggers, B., K. Tempel, and S. Jacobs. 2003. Inventory of fall Chinook spawning habitat in mainstem reaches of Oregon’s coastal rivers. Cumulative Report. Coastal Salmonid Inventory Project and Coastal Chinook Research and Monitoring Project, Oregon Department of Fish and Game.
Ronquist, F., and J. P. Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19(12):1572-1574.
Ross, A. 1986. Adventures of the First Settlers on the Oregon or Columbia River, 1810-1813. Bison Books.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning, volume 2. Cold spring harbor laboratory press New York.
Schalk, R. F. 1986. Estimating salmon and steelhead usage in the Columbia Basin before 1850: the anthropological perspective. Northwest Environmental Journal 2(2):1-29.
Scholz, A., K. O’Laughlin, D. Geist, D. Peone, J. Uehara, L. Fields, T. Kleist, I. Zozaya, T. Peone, and K. Teesatuskie. 1985. Compilation of information on salmon and steelhead total run size, catch and hydropower related losses in the Upper Columbia River Basin, above Grand Coulee Dam. Upper Columbia United Tribes Fisheries Center Fisheries Technical Report (2).
Scholz, A. T., T. Basler, J. Smith, J. McMillan, and L. Conboy. 2014. Fishes of Eastern Washington: A Natural History, volume 1. Eastern Washington University.
84 Seufert, F., and T. Vaughan. 1980. Wheels of fortune. Oregon Historical Society.
Shedlock, A. M., J. D. Parker, D. A. Crispin, T. W. Pietsch, and G. C. Burmer. 1992. Evolution of the salmonid mitochondrial control region. Molecular phylogenetics and evolution 1(3):179-192.
Smith, C. L. 1979. Salmon fishers of the Columbia. Oregon State University Press.
Steward, J. H. 1938. Basin-plateau aboriginal sociopolitical groups, volume 120. US Government Printing Office Washington, DC.
Suchard, M. A., R. E. Weiss, and J. S. Sinsheimer. 2001. Bayesian selection of continuous-time Markov chain evolutionary models. Molecular biology and evolution 18(6):1001-1013.
Tavaré, S. 1986. Some probabilistic and statistical problems in the analysis of DNA sequences. Lectures on mathematics in the life sciences 17:57-86.
Taylor III, J. E. 2009. Making salmon: An environmental history of the Northwest fisheries crisis. University of Washington Press.
Teit, J. A. 1930. The Salishan tribes of the western plateaus. US Government Printing Office.
Thomas, W. K., R. E. Withler, and A. T. Beckenbach. 1986. Mitochondrial DNA analysis of Pacific salmonid evolution. Canadian Journal of Zoology 64(5):1058-1064.
Thompson, W. F. 1951. An outline for salmon research in Alaska. University of Washington, Fisheries Research Institute.
UCUT. 2015. Upper Columbia River Basin fish reintroduction project; Upper Columbia United Tribes.
US Columbia Basin Tribes, and Canadian First Nations. 2014. Fish passage and reintroduction in the US & Canadian upper Columbia River.
Utter, F. M., D. W. Chapman, and A. R. Marshall. 1995. Genetic population structure and history of chinook salmon of the upper Columbia River. American Fisheries Society Symposium[AM. FISH. SOC. SYMP.]. 1995.
Valentine, K., D. A. Duffield, L. I. Patrick, D. R. Hatch, V. L. Butler, R. L. Hall, and N. Lehman. 2008. Ancient DNA reveals genotypic relationships among Oregon populations of the sea otter (Enhydra lutris). Conservation Genetics 9(4):933-938.
Waknitz, F. W., G. M. Matthews, T. Wainwright, and G. A. Winans. 1995. Status review for Mid- Columbia River summer chinook salmon. US Department of Commerce, National Oceanic and Atmoospheric Administration, National Marine Fisheries Service, Northwest Fisheries Science Center.
85 Walker, D. 1967. Mutual cross-utilization of economic resources in the Plateau from aboriginal Nez Perce fishing practices. Washington State University Laboratory of Anthropology. Report of Investigations (21).
Walker, I. R., and M. G. Pellatt. 2008. Climate change and ecosystem response in the northern Columbia River basin-A paleoenvironmental perspective. Environmental Reviews 16(NA):113-140.
Waples, R. S. 1990. Conservation genetics of Pacific salmon. II. Effective population size and the rate of loss of genetic variability. Journal of Heredity 81(4):267-276.
Waples, R. S., G. R. Pess, and T. Beechie. 2008. Evolutionary history of Pacific salmon in dynamic environments. Evolutionary Applications 1(2):189-206.
Waples, R. S., D. J. Teel, J. M. Myers, and A. M. Marshall. 2004. Life-history divergence in Chinook salmon: Historic contingency and parallel evolution. Evolution 58(2):386-403.
Welch, A. J., A. E. Wiley, H. F. James, P. H. Ostrom, T. W. Stafford, and R. C. Fleischer. 2012. Ancient DNA reveals genetic stability despite demographic decline: 3,000 years of population history in the endemic Hawaiian petrel. Molecular biology and evolution 29(12):3729-3740.
Wilkes, C. 1845. Narrative of the United States' Exploring Expedition: During the Years 1838, 1839, 1840, 1841, 1842, volume 1. Whittaker.
Winans, G. A. 1989. Genetic variability in chinook salmon stocks from the Columbia River Basin. North American Journal of Fisheries Management 9(1):47-52.
Winters, M., J. L. Barta, C. Monroe, and B. M. Kemp. 2011. To clone or not to clone: method analysis for retrieving consensus sequences in ancient DNA samples. PLoS One 6(6):e21247.
Wright, S. 1931. Evolution in Mendelian populations. Genetics 16(2):97-159.
Wright, S. 1949. The genetical structure of populations. Annals of eugenics 15(1):323-354.
86 FIGURES AND TABLES
Figure 1. Map showing locations discussed in the chapter. Ancient sample locations for the
Columbia (1 – 3), Snake (6 – 8), and Spokane (5) sample groups are indicated with black circles.
Contemporary sample locations for the Columbia (near 4) and Snake (near 9) are in blue.
87
Figure 2. Example of salmon vertebrae analyzed in this study. Image on left shows a fully intact vertebra, middle and right images are examples of fragmented, partial vertebrae. Samples shown here are from the Spokane River group.
88
Figure 3. Conceptual schematic of general processing procedure for ancient samples. Test results for amplification indicated with “A” and inhibition indicated with “I” where “+” indicates the presence of a band of the correct size.
89
Figure 4. Overlapping fragments used compile DLoop haplotype for ancient samples. Primer targets (including primers) are indicated
by grey bars and known SNPs for Chinook haplotypes are indicated with vertical black lines. All positions are relative to reference
sequence NC2980. Sequences and annealing temperature for each set listed in Table 2.
90
Figure 5. Three-dimensional haplotype network for Columbia River sample group. Circle size is proportional to relative haplotype frequency in ancient and contemporary subgroup, horizontal lines represent connections between haplotypes.
91
Figure 6. Conceptual outline of the Wright-Fisher model of genetic drift described in this study. A full, annotated version of the script is given in Appendix B.
Goal Model logic Example Effective population size (N) Create a population with effective population size N and N = 200
and haplotype frequencies are haplotypes HA,… HX with frequencies f(HA),...f(HX) f(HA) = 0.25 entered a priori f(HB) = 0.35
f(HC) = 0.40 This is generation 1
Take N samples from the Choose a random number (R) between 0 and 1 and assign a If population and determine their haplotype based on that number and the haplotype 0 < R ≤ 0.25 assign
R ≤ f(HA) assign haplotype A 0.25 < R ≤ 0.6 assign haplotype B R ≤ [f(HA) + f(HB)] assign haplotype B 0.6 < R ≤ 1 assign …HX
92 Make a new population from Count each of the haplotypes from the previous step and HA = 40 f(HA) = 0.20 the haplotypes sampled determine their frequencies HB = 76 f(HB) = 0.38
HC = 84 f(HC) = 0.42 This is generation 2
Take N samples from the Choose a random number (R) between 0 and 1 and assign a If population and determine their haplotype based on that number and the haplotype 0 < R ≤ 0.20 assign
R ≤ f(HA) assign haplotype A 0.20 < R ≤ 0.58 assign haplotype B R ≤ [f(HA) + f(HB)] assign haplotype B 0.58 < R ≤ 1 assign haplotype C …HX
Make a new population from Count each of the haplotypes from the previous step and HA = 22 f(HA) = 0.11 the haplotypes sampled determine their frequencies HB = 50 f(HB) = 0.25
HC = 28 f(HC) = 0.14 This is generation 3
The process is repeated for the number of generations specified a priori; each time making a new population with new haplotype frequencies based on a random sampling of the previous generation.
Figure 7. Regression plots for correlation of vertebrae diameter to fork length. Generated from data provided in Hofkamp (2015).
93
Figure 8. Figure follows on next page. Haplotype networks for haplotypes sampled in ancient and contemporary groups. Spokane is compared to Columbia subgroups. Orientation for haplotypes is constant between networks, circle size is proportional to frequency in the grouping, lines represent mutational connections. Four “evolutionary backbone” types as defined by Martin et al. 2010 are labeled for reference.
94
95
Figure 9. Rarefaction curves for ancient and contemporary samples from the Columbia, Spokane and Snake River sample groups.
Spokane samples are compared to Columbia subgroups as a proxy for single stock comparisons.
96
Figure 10. Bayesian cladogram for haplotypes from full species range as well as novel types detected in this study. Previously identified types are coded according to geographic region north: blue, central: yellow, southern: red, per designations in Martin et al. (2010).
91 20 21 86 17 19 18 68 23 1150 & 3127 YBP (Columbia); 300 - 3000 YBP, 1500 YBP (Snake) 26 500 - 4000 YBP (Snake) 1B 14 15 1A 13 11 12 4A 10 16 2A 22 Contemporary (Columbia) 24 3127 YBP (Columbia) 25 3250 or 7200 YBP (Spokane); & Contemporary (Snake) 27 7200 YBP (Spokane) 28 3250 or 7200 YBP (Spokane) O. mykiss O. clarkii
97
Figure 11. Follows on next page. Distributions of final outcomes for 5000 simulations of the
Wright-Fisher process (drift) given varying effective population sizes (Ne) starting with haplotypes in single temporal ancient subset of Columbia, Snake, and Spokane groups and modeled to contemporary time. Model outcomes are indicated with grey bars, observed values for contemporary data are indicated with black lines. Observed values in Spokane group are for the redirection tributaries, as an example of diversity in a single stock.
98
99
Figure 12. Extended Bayesian Skyline Plots (top) for Columbia and Snake River groups. Columbia was modeled using only sequences from contemporary samples as well as with all ancient and contemporary samples. Snake was modeled with both contemporary and ancient samples. Mean effective population size is indicated with solid black line, median with dashed line, and 95% highest posterior density (HPD) in shaded grey. Graph (bottom) plots the rate of reduction in effective population size indicated in the EBSP plots.
100
Figure 13. Comparison of spatial (top) and temporal (lower four) haplotype sampling.
Orientation for haplotypes is constant between networks, circle size is proportional to frequency in the grouping, lines represent mutational connections. Spatial network (top) is color coded for: southern-red, central-yellow, and northern-blue portions of the species range.
101
Table 1. Summary of haplotypes sampled, haplotype richness (HR) and haplotype richness adjusted to the smallest sample size of 24
(HRADJ), as well as haplotype (h) and nucleotide (π) diversity for ancient and contemporary samples and for selected temporal
groupings. Standard error (SE) and variance (var.) are for sampling process (iterations). [*Data from Martin et al. (2010)]. Table
continues on consecutive pages.
Approximate Haplotype: TSA___ HRADJ age (YBP) N 1a 1b 4a 10 12 17 18 19 22 23 24 25 26 27 28 HR (SE) h (var.) π x100 Spokane River Group 26 2 -- -- 10 -- 9 ------3 -- 1 1 6 6 (0.5) 0.52 (0.004) 0.186 2500 2 1 -- -- 1 ------2 3250 4 ------4 ------1 7200 11 1 -- -- 3 -- 6 ------1 -- 4 0.55 (0.005) 0.195
102 3250 or 7200 9 ------2 -- 3 ------3 -- -- 1 4
Columbia River Group: Ancient 34 3 9 -- 6 -- 13 ------2 1 ------6 6 (0.6) 0.73 (0.002) 0.196 Fort Colvile 100 1 1 ------1 GCDPA Tail Race 3127 28 2 6 -- 6 -- 12 ------1 1 ------6 0.71 (0.002) 0.198 Ksunku (Kettle Falls) 1150 4 -- 3 ------1 ------2 Shonitkwu (Kettle Falls) 7627 1 ------1 ------1 Columbia River Group: Contemporary 240 -- 5 -- 20 1 208 -- -- 6 ------5 3 (0.8) 0.24 (0.001) 0.080 Carson & Leavenworth Hatch. 55 -- 1 -- 4 -- 50 ------3 2 (0.6) 0.17 (0.004) 0.055 Entiat 53 -- 1 -- 10 -- 42 ------3 2 (0.5) 0.34 (0.005) 0.116 Icicle Creek 52 -- 1 -- 2 -- 48 -- -- 1 ------4 3 (0.8) 0.15 (0.004) 0.046 Methow 42 ------3 -- 34 -- -- 5 ------3 3 (0.6) 0.33 (0.007) 0.119 Wenatchee 38 -- 2 -- 1 1 34 ------4 3 (0.8) 0.20 (0.007) 0.062
Approximate Haplotype: TSA___
age (YBP) N 1a 1b 4a 10 12 17 18 19 22 23 24 25 26 27 28 HRHRADJ (SE) h (var.) π x100 Snake River Group: Ancient 24 4 -- -- 3 -- 12 ------3 -- -- 1 -- -- 5 5 (NA) 0.64 (0.005) 0.216 3 Springs Bar 300 - 3000 6 ------1 -- 3 ------1 ------3 Granite Point 1500 - 2500 1 ------1 ------1 Granite Point 2500 - 5000 1 1 ------1 Harder site 1450 3 ------1 ------2 ------2 Hatiuhpuh 500 - 4000 2 ------1 ------1 -- -- 2 Wexpusnime 700 - 1000 9 2 -- -- 1 -- 6 ------3 0.50 (0.016) 0.179 Windust Caves 300 - 4500 2 1 ------1 ------2 Snake River Group: Contemporary 139 11 12 1 6 -- 105 1 1 ------2 ------8 4 (0.9) 0.42 (0.003) 0.141 Chamberlain 10 ------10 ------1 0.00 (0.000) 0.000 Grande Ronde 20 3 ------16 ------1 ------3 0.32 (0.012) 0.115 Imnaha River 9 1 1 -- 2 -- 5 ------4 0.64 (0.016) 0.189 Lemhi 8 1 ------7 ------2 0.25 (0.032) 0.089 Lyons Ferry Hatchery* 22 3 11 1 4 -- 1 1 1 ------7 0.72 (0.007) 0.300 Middle Fork Salmon R. 17 1 ------16 ------2 0.12 (0.010) 0.042
103 South Fork Salmon R. 12 1 ------11 ------2 0.17 (0.018) 0.059 Tucannon* 21 ------21 ------1 0.00 (0.000) 0.000
Upper Salmon 20 1 ------18 ------1 ------3 0.18 (0.011) 0.065 Temporal: Basin Combined Contemporary - 100 YBP 380 12 17 1 26 1 313 1 1 6 -- -- 2 ------0.31 (0.001) 0.096 101 - 2500 YBP 19 3 3 -- 3 -- 7 ------3 ------0.75 (0.003) 0.234 2501 - 3950 YBP 32 2 6 -- 10 -- 12 ------1 1 ------0.70 (0.002) 0.208 3951 - 7950 YBP 12 1 -- -- 3 -- 7 ------1 -- 0.53 (0.006) 0.197
Table 2. Sequence, position, and annealing temperature for primer sets used to determine haplotype. Graphical depiction of primer
location is given in Figure 4; sequences are listed 5' - 3' with positions relative to reference sequence NC2980.
Target position Forward Reverse Annealing temperature (°C) 499-683 TCTTATTGCCCGTTACCCCC TGATTCTTTATAGAATATCA 50 524-716 CCGGGCGTTCTCTATATATGC CATAGTTCCCTGGAATTCAA 54 524-726 CCGGGCGTTCTCTATATATGC TAACGGACCTTATGCACTTG 56 630-793 TTGAATTCCAGGGAACTATG AAGGATCTTTCAGCGTAGGG 54 630-761 TTGAATTCCAGGGAACTATG TTACCGCGCAGAAGCCGGGG 54 732-891 TCTAAGATTTCCCCGGCTTC TGCCAAACTGCTATAAAGTGC 58 732-845 TCTAAGATTTCCCCGGCTTC AAAACATCATGCTGATTTGA 50 816-1008 GCTTTAGTTAAGCTACGCCAG CCAGGAAGTTTCAAATCAGCA 58
104 879-1054 TAGCAGTTTGGCACCGACAG TGCTCGTGGGACTTTCTAGG 60 929-1077 GTAAAGTCAGGACCAAGCCTTT GTATACATTAATAAACTTTCCG 54 953-1141 CCCCTAGCAACACCATTTTC GGGATTAAGGGCATCCTCAC 58 962-1141 ACACCATTTTCCCGCCTAAC GGGATTAAGGGCATCCTCAC 58
Table 3. Polymorphic positions for 12S species (top) and DLoop (bottom) haplotypes sampled for Chinook salmon in this study. Novel 12S types are types 2 - 4, type 1 was described in Jordan et al. (2010). Novel DLoop haplotypes types are TSA22 - TSA28, other types were described by
Martin et al. (2010).
629 631 660 669 671 674 711 730 DQ288271 G A C A C C C G 1 T . T . . . T A 2 T . T . . . T . 3 . . T C . . T A 4 T C T . T T T A
592 607 623 641 848 869 884 887 898 899 955 956 961 967 972 973 976 985 1029 1052 NC2980 G T C G A G G T G C C C A A C - G A T A TSA01A - - . . . . - ...... C . . G . TSA01B - - . A . . - ...... C . . G . TSA04A - - . . . . - ...... A . . G . TSA10 - - . . . . - ...... G . TSA12 - - . . . . - ...... TSA17 - - . A . . - ...... C . C . . G . TSA18 - - . A . . - . T . . . . C . C . . G . TSA19 - - . A G . - ...... C . C . . G . TSA22 - - . A . A - ...... C . . G . TSA23 - - . A . . - . . . A . . C . C . . G . TSA24 - - . . . . - . . . . A . . . C . . G . TSA25 - - ...... C . . G . TSA26 - - . A . . - C . . . . . C . C . . G . TSA27 - - ...... G . TSA28 - - T A . . - ...... C . . G .
105
Table 4. Follows on next page. Quantification of success rates by PCR, extract, and sample
(independent vertebra) for species (12S) as well as the overall portion of samples that resulted in haplotypes (DLoop: Overall) and the success rate of DLoop efforts on samples that were genetically confirmed as Chinook Salmon (DLoop: post 12S). PCR methods are abbreviated as standard (S) and rescue (R).
106
Successful Species DLoop succes PCR method results 12S succes rates rates (sample) Chinook Dloop Post Samples Extracts S R None Total salmon haplo. PCR Extract Sample Overall 12S Ancient Snake River group 172 235 49 15 171 50 24 18 4% 21% 29% 10% 75% Granite Point 25 33 6 5 22 9 3 2 6% 27% 36% 8% 67% Harder site 11 19 5 1 13 3 3 3 3% 16% 27% 27% 100% Hatiuhpuh 14 17 2 4 11 3 2 2 4% 18% 21% 14% 100% Marmes 20 25 0 0 25 0 0 0 0% 0% 0% 0% 0% Three Springs Bar 36 54 14 2 38 14 5 0 6% 26% 39% 0% 0% Wexpusnime 18 26 14 1 11 13 9 9 16% 50% 72% 50% 100% Windust Caves 48 61 8 2 51 8 2 2 3% 13% 17% 4% 100% Ancient Columbia River group 56 62 38 1 23 35 35 34 12% 43% 63% 61% 97% Fort Colville 4 4 1 0 3 1 1 1 5% 25% 25% 25% 100% GCDPA Tail Race 35 36 29 0 7 28 28 28 39% 78% 80% 80% 100%
107 Ksunku (Kettle Falls) 8 11 7 0 4 4 4 4 13% 36% 50% 50% 100% Shonitkwu (Kettle Falls) 9 11 1 1 9 2 2 1 3% 18% 22% 11% 50%
Nason Creek Weir (formalin) 5 10 0 0 10 0 0 0 0% 0% 0% 0% 0% Revelstoke, BC (scales, formalin pos.) 10 15 1 0 14 0 0 0 0% 0% 0% 0% 0% Golden, BC (scales, formalin pos.) 8 12 1 0 11 1 0 0 1% 8% 13% 0% 0% Rock Island Dam (formalin) 8 17 0 0 17 0 0 0 0% 0% 0% 0% 0% Ancient Spokane River group 118 165 14 43 108 45 45 26 6% 27% 38% 22% 58% Ancient Sentinel Gap 19 19 0 0 19 0 0 0 0% 0% 0% 0% 0% Contemporary Snake River group 103 103 96 NA NA NA NA 96 85% 93% 93% See 12S data Contemporary Columbia River group 263 278 240 NA NA NA NA 240 75% 86% 91% See 12S data TOTALS Total Ancient (non-bone) 26 44 2 0 42 1 0 0 0% 2% 4% 0% 0% Total Ancient (bone) 365 481 101 59 321 130 104 78 6% 27% 36% 21% 75% Total Contemporary 103 103 96 NA NA NA NA 96 81% 93% 93% See 12S data
Table 5. Comparison of species indicated by morphometric data [following Huber et al. (2011] versus that determined genetically. Table continues on consecutive pages.
Mean Mean Vert. length height Sample ID type (mm) (mm) Morphometric result Genetic result Columbia 1 2 8.53 13.08 O. tshawytscha (Chinook) Same Columbia 2 2 8.77 11.73 O. tshawytscha (Chinook) Same Columbia 3 2 9.08 12.59 O. tshawytscha (Chinook) Same Columbia 4 2 8.28 11.24 O. tshawytscha (Chinook) Same Columbia 5 2 8.84 10.23 O. mykiss (Steelhead) O. tshawytscha Columbia 6 2 7.80 10.65 O. tshawytscha (Chinook) Same Columbia 7 2 7.51 10.46 O. tshawytscha (Chinook) Same Columbia 8 2 7.70 10.35 O. tshawytscha (Chinook) Same Columbia 9 2 9.23 14.16 O. tshawytscha (Chinook) Same Columbia 10 2 6.75 10.58 O. tshawytscha (Chinook) Same Columbia 11 2 4.27 6.61 O. gorbuscha (Pink) O. tshawytscha Columbia 12 2 6.33 8.22 O. kisutch (Coho) O. tshawytscha Columbia 13 2 7.74 11.01 O. tshawytscha (Chinook) Same Columbia 14 2 7.16 11.22 O. tshawytscha (Chinook) Same Columbia 15 2 7.11 11.21 O. tshawytscha (Chinook) Same Columbia 16 3 8.43 12.30 O. tshawytscha (Chinook) Same Columbia 17 3 9.71 13.34 O. tshawytscha (Chinook) Same Columbia 18 3 8.57 13.24 O. tshawytscha (Chinook) Same Columbia 19 3 8.86 11.39 O. tshawytscha (Chinook) Same Columbia 20 3 9.37 13.08 O. tshawytscha (Chinook) Same Columbia 21 3 6.51 9.71 O. tshawytscha (Chinook) Same Columbia 22 3 9.69 14.55 O. tshawytscha (Chinook) Same Columbia 23 3 7.47 13.65 O. tshawytscha (Chinook) Same Columbia 24 3 6.75 9.79 O. tshawytscha (Chinook) Same Columbia 25 3 6.35 9.97 O. tshawytscha (Chinook) Same Columbia 26 3 8.65 13.06 O. tshawytscha (Chinook) Same Columbia 27 3 8.74 12.90 O. tshawytscha (Chinook) Same Columbia 28 3 7.05 11.98 O. tshawytscha (Chinook) Same Columbia 29 3 8.11 11.55 O. tshawytscha (Chinook) Same
108
Mean Mean Vert. length height Sample ID type (mm) (mm) Morphometric result Genetic result Snake 1 2 5.22 9.80 O. tshawytscha (Chinook) Same Snake 2 2 4.57 10.72 O. tshawytscha (Chinook) Same Snake 3 2 6.20 9.20 O. tshawytscha (Chinook) Same Snake 4 2 6.49 9.37 O. tshawytscha (Chinook) Same Snake 5 2 6.81 11.61 O. tshawytscha (Chinook) Same Snake 6 2 6.71 10.85 O. tshawytscha (Chinook) Same Snake 7 2 6.99 12.14 O. tshawytscha (Chinook) Same Snake 8 2 9.87 15.53 O. tshawytscha (Chinook) Same Snake 9 2 8.49 13.09 O. tshawytscha (Chinook) Same Snake 10 2 9.78 14.09 O. tshawytscha (Chinook) Same Snake 11 2 5.42 12.68 O. tshawytscha (Chinook) Same Snake 12 2 5.04 7.10 O. gorbuscha (Pink) O. tshawytscha Snake 13 2 5.74 7.06 O. mykiss (Steelhead) O. tshawytscha Snake 14 2 6.93 9.22 O. kisutch (Coho) Same Snake 15 2 5.19 8.06 O. gorbuscha (Pink) O. kisutch Snake 16 3 4.14 7.47 O. gorbuscha (Pink) O. tshawytscha Snake 17 3 3.97 8.89 O. gorbuscha (Pink) O. tshawytscha Snake 18 3 5.29 8.18 O. gorbuscha (Pink) O. tshawytscha Snake 19 3 6.50 7.78 O. mykiss (Steelhead) O. tshawytscha Snake 20 3 7.55 12.98 O. tshawytscha (Chinook) Same
109
Table 6. Predicted fork length for samples genetically confirmed as Chinook salmon based on vertebrae diameter and data in Hoffkamp (2015). Sample IDs are prefixed with the sample group.
Diameter Predicted fork Diameter Predicted fork Sample ID Type (mm) length (cm) Sample ID Type (mm) length (cm) Columbia 1 2 5.8 39.8 Snake 1 2 7.1 49.1 Columbia 2 2 7.5 52.4 Snake 2 2 7.1 49.4 Columbia 3 3 8.4 55.9 Snake 3 2 9.2 64.3 Columbia 4 3 8.4 56.1 Snake 4 2 9.4 65.6 Columbia 5 3 8.5 56.7 Snake 5 2 9.8 68.7 Columbia 6 2 9.0 62.9 Snake 6 2 10.7 75.2 Columbia 7 2 9.2 64.4 Snake 7 2 10.9 76.1 Columbia 8 2 9.2 64.5 Snake 8 2 11.6 81.5 Columbia 9 3 9.9 65.9 Snake 9 2 12.1 85.4 Columbia 10 2 9.4 66.1 Snake 10 2 12.7 89.2 Columbia 11 2 9.5 66.3 Snake 11 2 13.1 92.1 Columbia 12 2 9.5 66.4 Snake 12 2 14.1 99.2 Columbia 13 3 10.1 67.4 Snake 13 2 15.5 109.5 Columbia 14 2 9.7 67.7 Columbia 15 2 9.8 68.5 Spokane 1 3 4.4 29.5 Columbia 16 3 10.4 69.2 Spokane 2 2 5.3 36.6 Columbia 17 2 10.0 70.3 Spokane 3 2 11.6 81.5 Columbia 18 3 10.7 71.4 Spokane 4 unk 15.3 101.4 or 107.9 Columbia 19 3 11.0 73.1 Spokane 5 unk 16.3 108.0 or 115.2 Columbia 20 2 10.5 73.5 Spokane 6 3 18.7 123.9 Columbia 21 3 11.2 74.4 Spokane 7 2 18.8 132.9 Columbia 22 3 11.2 74.6 Spokane 8 2 19.5 137.8 Columbia 23 3 11.3 75.0 Columbia 24 3 11.5 76.6 Columbia 25 2 11.2 78.6 Columbia 26 3 11.9 78.7 Columbia 27 2 11.3 79.6 Columbia 28 3 12.5 82.8 Columbia 29 2 12.1 85.4
110
Table 7. Pairwise φST for ancient and contemporary sample groupings. φST values are listed above and color-coded by quartile below, underlined values indicate significance (α = 0.05).
Values of φST for the ancient Columbia River samples are presented as both just those dated to
3127 YBP and for all ancient samples combined. Non-significant values colored as values of 0
(no differentiation) and marked with "NS". [*Data from Martin et al. (2010)]. Table continues on consecutive pages.
111
Contemporary Columbia
Contemporary Columbia Contemporary Hatchery Leavenworth & Carson Wenatchee Methow Icicle Entiat Columbia (3127 YBP) 0.29 0.30 0.26 0.17 0.33 0.12 Ancient Columbia 0.29 0.30 0.26 0.17 0.33 0.13 Ancient Snake 0.14 0.15 0.13 0.08 0.19 0.02 Ancient Spokane 0.52 0.53 0.49 0.38 0.56 0.32
Columbia (3127 YBP) Ancient Columbia Ancient Snake NS Ancient Spokane
Quartiles defined from data spread very high 0.58 max high 0.39 moderate 0.20 median low 0.10
112
Contemporary Snake
Contemporary Snake Contemporary Chamberlain Ronde Grand Imnaha Lemhi Hatchery* Ferry Lyons Fork Middle Fork South Tucannon* Salmon Upper Columbia (3127 YBP) 0.13 0.29 0.09 0.00 0.12 0.08 0.25 0.19 0.36 0.20 Ancient Columbia 0.14 0.28 0.10 0.00 0.13 0.07 0.25 0.20 0.35 0.21 Ancient Snake 0.04 0.14 0.00 0.00 0.00 0.21 0.11 0.06 0.22 0.07 Ancient Spokane 0.06 0.50 0.28 0.04 0.33 0.07 0.46 0.40 0.58 0.42
Columbia (3127 YBP) NS NS Ancient Columbia NS NS Ancient Snake NS NS NS NS NS NS NS Ancient Spokane NS NS
Species Range
American R., CA(South)* R., American BC(Central)* R., Chilliwack (North)* AK R., Gulkana (North)* Russia Pen., Kamchatka (Central)* WA Hatch., Rapids Priest (North)* AK R., Tuluksak CA(South)* R., Tuolumne OR R., (Central)* Willamette (North)* AK R., Yukon Columbia (3127 YBP) 0.46 0.27 0.24 0.35 0.03 0.14 0.48 0.18 0.07 Ancient Columbia 0.46 0.28 0.20 0.34 0.04 0.14 0.49 0.20 0.07 Ancient Snake 0.55 0.36 0.39 0.21 0.08 0.11 0.56 0.22 0.09 Ancient Spokane 0.27 0.09 0.39 0.56 0.05 0.35 0.30 0.14 0.30
Columbia (3127 YBP) NS Ancient Columbia NS Ancient Snake NS Ancient Spokane NS
Quartiles defined from data spread very high 0.58 max high 0.39 moderate 0.20 median low 0.10
113
Table 8. Probability of obtaining haplotype diversity (h) equal or less than that observed in the contemporary group given a model of
genetic drift applied to temporal subset of ancient samples. Probability is based on proportion of outcomes from 5000 simulations of
the Wright-Fisher process starting with the ancient group. Spokane is compared to subgroups of contemporary Columbia samples as
a proxy of single-stock sample sets. Values < 0.05 are underlined.
Probability of h ≤ than that observed, given random genetic
h drift for NeF indicated
Group Ancient Contemp. NGEN 100 500 1000 2500 5000 10K 25K 35K 45K 55K Columbia 0.73 0.24 795 0.999 0.689 0.354 0.076 0.011 0.000 0.000 0.000 0.000 0.000
114 Snake 0.50 0.42 225 0.946 0.576 0.469 0.357 0.270 0.173 0.066 0.036 0.021 0.000
Spokane 0.55 --- 1813 ------Carson and Leavenworth Hatch. --- 0.17 --- 1 0.957 0.764 0.340 0.129 0.028 0.002 0.000 0.000 0.000 Entiat --- 0.34 --- 1 0.968 0.834 0.490 0.266 0.123 0.027 0.008 0.004 0.002 Icicle --- 0.15 --- 1 0.956 0.757 0.328 0.116 0.023 0.001 0.000 0.000 0.000 Methow --- 0.33 --- 1 0.967 0.828 0.477 0.256 0.116 0.024 0.006 0.004 0.001 Wenatchee --- 0.20 --- 1 0.958 0.774 0.362 0.146 0.040 0.003 0.000 0.000 0.000
Table 9. Probability of observing haplotype diversity equal or less than that in contemporary given low effective population size (NeF)
over 50, 100, or 200 year intervals for the Columbia, Snake, and Spokane groups. Values < 0.05 are underlined.
NeF: 100 500 1000 2500 Number of years 50 100 200 50 100 200 50 100 200 50 100 200 Columbia 0.005 0.041 0.174 0.000 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 Snake 0.390 0.475 0.600 0.191 0.280 0.365 0.106 0.187 0.292 0.019 0.069 0.157 Carson and Leavenworth Hatch. 0.014 0.070 0.210 0.000 0.000 0.005 0.000 0.000 0.000 0.000 0.000 0.000 Entiat 0.078 0.185 0.371 0.000 0.008 0.049 0.000 0.000 0.008 0.000 0.000 0.000 Icicle 0.010 0.060 0.195 0.000 0.000 0.004 0.000 0.000 0.000 0.000 0.000 0.000 Methow 0.072 0.178 0.359 0.000 0.007 0.043 0.000 0.000 0.006 0.000 0.000 0.000 Spokane vs. Spokane Wenatchee 0.020 0.085 0.238 0.000 0.000 0.008 0.000 0.000 0.000 0.000 0.000 0.000
115
Table 10. Summary of demographic models implemented in BEAST. Comparisons are given as the difference in likelihood, negative values indicate a better fit of the constant population size model (A) compared to the EBSP model (B).
Bayes Description Model ESS factor 1 A Columbia: Modern only Constant 2249 -- B Columbia: Modern only EBSP 2498 -5.87 2 A Columbia: Ancient and contemporary Constant 391 -- B Columbia: Ancient and contemporary EBSP 207 9.38 3 A Snake: Ancient and contemporary Constant 3392 -- B Snake: Ancient and contemporary EBSP 3664.5 0.65
116
Table 11. Comparison of spatial (range based, from Martin et al. 2010) and temporal haplotype frequencies. Novel haplotypes TSA26
and TSA28 were sampled from vertebrae with broad age estimates (500 - 4000 and 3300 or 7200 YBP, respectively) and are not
included in this data.
North South only North & central Central only Central & South only Novel types N 20 21 01b 17 18 11 13 14 16 12 19 4a 10 01a 15 2A 22 23 24 25 27 0 - 100 380 -- -- 0.045 0.824 0.003 ------0.003 0.003 0.003 0.068 0.032 -- -- 0.016 -- -- 0.005 -- 101 - 2500 19 -- -- 0.158 0.368 ------0.158 0.158 ------0.158 ------2501 - 3950 32 -- -- 0.188 0.375 ------0.313 0.063 ------0.031 0.031 -- -- 3951 - 7950 12 ------0.583 ------0.250 0.083 ------0.083
117
CHAPTER TWO
Rescue PCR: Reagent-rich PCR recipe improves amplification of degraded DNA extracts
1. INTRODUCTION
Polymerase chain reaction (PCR), the in vitro process by which small amounts of template
DNA can be copied and exponentially increased in copy number, has transformed molecular biology (Bartlett and Stirling 2003; Palumbi 1996). Despite several decades of refinement, some challenges remain. In particular, as PCR is dependent on enzymes, it is subject to inhibition. PCR
Inhibition can result in the failure to copy available DNA molecules due to the presence of some extraneous substance or substances, the inhibitor(s). Given that adequate DNA is present in an eluate, inhibition is the most frequent cause of PCR failure (Alaeddini 2012). To this end, analysis of DNA derived from low copy number (LCN), ancient and/or degraded samples can be especially challenging, as such specimens have often spent time buried in the ground and/or in contact with environmentally-based inhibitory substances (Alaeddini 2012; Kemp et al. 2014a;
Schrader et al. 2012). While DNA recovered from these types of samples may be especially prone to PCR inhibition, this phenomenon has also been well documented in studies utilizing clinical, food, and other contemporary sample sources (Al-Soud and Rådström 2001; Alaeddini
2012; Rådström et al. 2004; Rossen et al. 1992; Schrader et al. 2012; Wiedbrauk et al. 1995).
The list of compounds that can act as inhibitors to PCR is as long as it is diverse. Some inhibitors may be introduced during sample processing and/or DNA extraction. Types of such inhibitors include salts (e.g., sodium or potassium chloride), detergents, ethanol, isopropyl, phenol, and even powder from laboratory gloves (Burkardt 2000; Demeke and Jenkins 2010;
118
Katcher and Schwartz 1994; Schrader et al. 2012; Weyant et al. 1990; Wilson 1997). In these cases, proper protocol selection and careful processing may be able to neutralize or minimize the effects of these inhibitors (Rådström et al. 2008; Schrader et al. 2012; Weyant et al. 1990).
However, naturally occurring environmental substances such as iron, copper, lead, as well as substances that exist in the samples themselves (e.g., calcium and collagen in bone and/or connective tissue, melanin in hair and skin, hematin in blood, among others) can also inhibit
PCR [for reviews see Alaeddini (2012), Kemp et al. (2006), and Schrader et al. (2012)]. Similarly, if present, exogenous non-target DNA will be co-extracted with target DNA. This can reduce the efficiency of PCR when present in high enough concentrations (Tebbe and Vahjen 1993; Wilson
1997). Inhibitors of these types are more difficult to exclude from processing as their sample- incorporated nature means they might be co-extracted with DNA, despite even the most impeccable laboratory procedures.
The outcome of the presence of inhibitors has been well documented, but determining the actual mechanism of inhibition has proven more challenging (Alaeddini 2012). Potential mechanisms include interference with cell lysing during DNA extraction as well as interference with polymerase, primer binding sites, and/or template DNA during PCR (Bickley et al. 1996;
Eckhart et al. 2000; Opel et al. 2010; Wilson 1997). These mechanisms may be predictable for some specific inhibitors. For example, calcium, hematin, and tannic acid are indicated to act directly on polymerase, melanin appears to bind with DNA templates, and collagen exhibits both these behaviors (Opel et al. 2010). These classifications may be informative, but are based on controlled experiments where known inhibitors at known concentrations are added to DNA standards. As such, they lack direct application in the laboratory where, in any given DNA
119
eluate, there can exist unknown concentrations of an unknown number of different inhibitory substances.
Practical investigations into inhibition have focused on strategies to remove inhibitors and/or subdue their influence. For example, removal of inhibitors can be accomplished via treatments in which DNA is bound to silica, thus allowing inhibitors to be washed away prior to releasing the DNA back into solution (Kemp et al. 2006; Yang et al. 1998). In fact, subjection of
DNA to repeated rounds of silica extraction has been found to be particularly useful (Grier et al.
2013; Kemp et al. 2014a; Moss et al. 2014). In addition to removal strategies, several methods for the circumvention of inhibition have been demonstrated. The most common is direct dilution of DNA extracts, which likely lowers the level of inhibitors below some “threshold” at which PCR can successfully copy DNA (Alaeddini 2012; Kemp et al. 2006). Modifications of the
PCR recipe have also been demonstrated as a means to amplify DNA in the presence of inhibitors. Adding protein-based facilitators such as bovine serum albumin (BSA) to PCR reactions may bind and inactivate some types of inhibitors (Juen and Traugott 2006; Kreader
1996). Increasing the concentration of polymerase and/or its magnesium cofactor (e.g., in the form of MgCl2) may also aid in overcoming inhibition (Rådström et al. 2008; Wilcox et al. 1993).
Additionally, alternating or blending multiple polymerases for use in PCR recipe has also been demonstrated as a means to effectively overcome inhibition, as certain polymerases appear to have decreased susceptibility to specific types of inhibitors (Al-Soud and Rådström 1998; Belec et al. 1998; Eilert and Foran 2009; Hedman et al. 2010; Monroe et al. 2013).
In practice, any DNA sample can be subject to a potential sundry of inhibitors, the outcomes of which can vary between samples, eluates, and PCR reactions (Huggett et al. 2008).
120
While methods exist to remove or dilute inhibitors, eluates with very low concentrations of
DNA may not benefit from such applications. In the case of removal methods, for each treatment some loss of DNA will occur, along with inhibitor removal, which is a particularly undesired outcome when processing eluates with low DNA concentrations (Barta et al. 2014b;
Kemp et al. 2014b). Similarly, when eluates are diluted, the DNA concentration will be reduced along with inhibitors so this also may not be an effective strategy for processing samples with low DNA concentrations (Alaeddini 2012; Ye et al. 2004). Thus, modifications to the PCR recipe alone may present the best option for these types of samples. Here we present evidence that even low-levels of inhibition can produce false-negatives when DNA concentrations are reduced and, in turn, provide a simple and effective method to overcome such inhibition and obtain
DNA amplification. In homage to Gilbert and Willerslev (2007) who suggested that new polymerases may help “rescue” ancient DNA, we term our new method rescue PCR, a strategy based on a reagent-rich PCR recipe.
2. METHODS
All pre-PCR laboratory work (DNA extraction and PCR set-up) was conducted in the ancient DNA lab at Washington State University. Strict laboratory protocols are in place in this laboratory to closely monitor and minimize contamination to ensure results are authentic
(Kemp and Smith 2010).
121
2.1. Sample sources
A total of 227 vertebrae identified as Oncorhynchus spp. via morphometric analysis were acquisitioned from two archaeological collections. The first collection is comprised of materials from seven excavation locations coinciding with four contemporary dams along the Snake and
Columbia Rivers (locations 1-4 depicted in Figure 1). The first site, Strawberry Island (45FR5) in the McNary Reservoir, is an excavated house pit village with materials dated to 2000 – 200 years before present (YBP) (Schalk et al. 1983). The second site, Windust Caves (45FR46), is located near Ice Harbor Dam. These caves were used as ancient storage and camp shelters with materials dating 9000 – 200 YBP (Jenkins 2011; Rice 1965; Thompson 1985). The caves were inundated by Lake Sacajawea in 1961 after the completion of the lock and dam. Three ancient house pit villages: Harder (45FR40), Hatiuhpuh (45WT134), and Three Springs Bar (45FR39) are located near the Lower Monumental Dam. Materials sampled from these sites are dated to
~1500 YBP (Harder), 4000 – 500 YBP (Hatiuhpuh), and 3000 – 200 YBP (Three Springs Bar)
(Brauner 1990; Browman and Munsell 1969; Daugherty et al. 1967; Funk 1998; Hicks 2004).
Two sites, Granite Point (45WT41) and Wexpusnime (45GA61) are located near Lower Granite
Dam. Granite Point is an ancient camp site with materials dated from 10000 – 200 YBP
(Leonhardy 1969). The Wexpusnime site is comprised of two components, a camp site with materials dated to pre-8000 YBP as well as a house pit village with materials dating to 500 YBP
(Nakonechny 1998). Samples from these seven locations will be collectively referred to as the
Snake River group. The second sample of vertebrae originated from a collection of materials excavated at an ancient fishing site near the Spokane River (45SP266) (location 5 depicted in
122
Figure 1) with two components, approximated at 2500 and 3250 YBP (Galm 1994). These samples will be collectively referred to as the Spokane River group.
2.2. DNA extractions
Three hundred and thirty-four extractions were conducted from the 227 vertebrae using one of two methods. One hundred and fifty-five extractions were generated using the first method (henceforth referred to as Extraction Method 1 or EM1) from approximately 7 - 48 mg of bone carefully removed from the whole. These portions of bone were submerged in 6%
(w/v) sodium hypochlorite (bleach) for 4 min (Barta et al. 2013) and the bleach decanted. The samples were then twice submerged in DNA-free water, with the water poured off following each submersion. Samples were transferred to 1.5 mL tubes, to which aliquots of 500 L of
EDTA (ethylenediaminetetraacetic acid) were added, and gently rocked at room temperature for >48 hours. Samples were extracted in batches of seven with one accompanying extraction negative control per batch. DNA was extracted following the WSU method described by Cui et al. (2013). One hundred and seventy-nine extractions were generated with the second method
(henceforth referred to as Extraction Method 2 or EM2) from approximately 53 - 412 mg of bone carefully removed from the whole. These portions of bone were submerged in 6% (w/v) sodium hypochlorite (bleach) for 4 min (Barta et al. 2013) and the bleach decanted. The samples were then twice submerged in DNA-free water, with the water poured off following submersion. Samples were transferred to 15 mL tubes, to which aliquots of 2 mL of EDTA were added, and gently rocked at room temperature for >48 hours. Samples were extracted in
123
batches of seven with one accompanying extraction negative control. DNA was extracted following a modified protocol of Kemp et al. (2007) described in Moss et al. (2014).
2.3. Initial evaluation for inhibition
All extracts were initially tested for inhibition following Kemp et al. (2014a) (see schematic illustration in their Figure 1). In brief, PCRs were set-up with an aDNA control, one comprised of pooled DNA extracted from ~3500 year old northern fur seal (Callorhinus ursinus) remains
(Barta et al. 2014a; Barta et al. 2013; Winters et al. 2011). This pool was created using individual
DNA extracts previously verified to yield 181 base pair (bp) amplicons of northern fur seal mitochondrial cytochrome B gene using the following primers: CytB-F 5’-
CCAACATTCGAAAAGTTCATCC-3’ and CytB-R 5’- GCTGTGGTGGTGTCTGAGGT-3’ (with an annealing temperature of 60°C) (Moss et al. 2006). This control PCR mix is then “spiked” with the DNA recovered from the salmonid vertebrae, that is DNA to be tested for the presence of sufficient inhibition to prevent the northern fur seal mitochondrial DNA (mtDNA) from amplifying. Note that the northern fur seal primers are incapable of amplifying salmonid mtDNA. One advantage of this approach to monitoring for the presence of PCR inhibitors is that the control is aDNA and exhibits characteristics common in ancient extracts (i.e., signatures of post-mortem chemical degradation, high levels of DNA fragmentation, and low concentrations)
(Barta et al. 2014a; Barta et al. 2013; Winters et al. 2011). Another advantage, given that the degree of PCR inhibition is directly related to the size of DNA to be amplified (Mccord et al.
2015), the northern fur seal mtDNA fragment size targeted by these reactions is similar to that targeted in salmonids (189 bp, see section 2.4 below). All of the ancient salmonid DNA extracts
124
demonstrated to not contain sufficient inhibitors to cause amplification failure of the aDNA control.
2.4. “Standard” PCR
Except in one test (described below in section 2.5.), all PCRs targeted a 189 bp portion of the 12S mitochondrial gene using “universal” fish primers: OST12S-F 5’-
GCTTAAAACCCAAAGGACTTG-3’ and OST12S-R 5’- CTACACCTCGACCTGACGTT-3’ (Jordan et al.
2010). Note that Jordan et al. (2010) described the OST12S-R primer in the incorrect orientation. It has been corrected here. These primers have been demonstrated to be especially effective in amplifying salmonid mtDNA, the sequences of which can be used to differentiate the Pacific salmonids and a variety of other fish to the to the species level (Grier et al. 2013;
Halffman et al. 2015; Jordan et al. 2010; Kemp et al. 2014a).
Polymerase selection was based on results from Monroe et al (2013) indicating Klentaq LA was the least susceptible of nine polymerase or polymerase blends to inhibition associated with
DNA obtained from prehistoric salmonid vertebrae recovered from two archaeological sites in the Pacific Northwest (DgRv-003 and DgRv-006). “Standard” 25 µL PCRs contained: 1X Omni
Klentaq Reaction Buffer mix (containing a final concentration of MgCl2 at 3.5 mM), 0.32 mM dNTPs, 0.24 µM each of forward and reverse primer, 0.3 U of Omni Klentaq LA polymerase, and
2.5 µL of template DNA. PCRs consisted of an initial three-minute denaturation at 94°C, followed by sixty cycles of 94°C (denaturation, 15 s), 55°C (annealing, 15 s), and 68°C
(extension, 15 s). This was followed with a final extension at 68°C for 3 minutes. Negative PCR controls and positive PCR controls (utilizing DNA extracted from contemporary Chinook salmon,
125
added in the post-PCR lab prior to initiating PCR) accompanied all sets of standard PCRs and experimentally modified PCRs (described below).
Successful amplification following standard PCR or any of the experimentally modified
PCRs were confirmed via separation on a 4% agarose gel and approximate size was determined against a 20 bp ladder (Bayou BioLabs). Amplification outcomes were classified as: 1) successful when a single, clear band of the correct size was seen, 2) failure when no band was seen, or 3) non-target (NT). Non-target classifications were further divided into two additional categories, those that were the incorrect size (i.e., based on the relative position on the gel) (non-target size, or NT-S) and those that produced multiple bands (non-target multiple or NT-M).
2.5. Rescue PCR: Tests of varying percentage reagent increases (10%, 25%, and 50%)
Initial rescue PCRs consisted of increasing the buffer mix, dNTPs, primers, and polymerase in equal relative proportion (i.e., +10%, +25%, +50%) with the amount of water reduced to accommodate the increased reagent volumes. Thirty extracts were tested for amplification with increases of 10%, 25%, and 50% for the buffer mix, dNTPs, primers, and polymerase. For example, in comparison to the “standard” PCR described in section 2.4, +50% rescue PCRs contained: 1.5X Omni Klentaq Reaction Buffer mix (containing a final concentration of MgCl2 at
5.25 mM), 0.48 mM dNTPs, 0.36 µM each of forward and reverse primer, 0.45 U of Omni
Klentaq LA polymerase, and 2.5 µL of template DNA. Rescue PCR conditions using the 12S primers were as described in section 2.4.
In addition to the 12S primer set, an additional set of PCRs were used to account for any differential behavior of inhibitors on specific primers, as well as potential template/primer
126
compatibility. These PCRs targeted a 193 bp portion of the salmonid mitochondrial control region (D-loop), spanning nucleotide positions (nps) 816 - 1008 [relative to a reference sequence of Oncorhynchus tshawytscha (154)] with the following primers: 5’-
GCTTTAGTTAAGCTACGCCAG-3’ and reverse 5’-CCAGGAAGTTTCAAATCAGCA-3’. These reaction conditions were as described in section 2.4, with an annealing temperature of 58oC for this primer set.
2.6. Effect of 25% increases of individual reagents and combinations of reagents
As the experiments described in section 2.5 demonstrate that +25% rescue PCR outperformed standard, +10% and +50% rescue PCRs (see results), we tested the efficacy of increasing individual reagents by 25%, as well as combinations of reagents. Sixteen PCR mixes were prepared, one using standard PCR (no increase in reagents), another using rescue PCR (all reagents increased by 25%), and the remaining fourteen using a 25% increase specifically of dNTPs, Omni Klentaq Reaction Buffer mix (including premixed MgCl2), Klentaq LA polymerase, or primers, as well as all possible combinations of these four reagents. Each reaction mix was tested across twelve salmon DNA extracts.
2.7. Comparisons of standard and rescue PCR across samples
Of the 334 extracts, neither standard or rescue PCR permitted amplification from 222 of them. Thus, we focused on the results from the remaining 112 extracts, 68 from the Snake
River group and 44 from the Spokane River group (Table S1). From these 112 extracts, 358 PCR reactions were conducted using either standard (N = 230) or +25% rescue (N = 128) PCR
127
recipes. Note that from this point forward, +25% rescue PCR will simply be referred to as
“rescue PCR”, as to differentiate it from +10% and +50% rescue PCRs. Results were tabulated in one of two ways. First, we established “application of method” for each sample by determining if a given method (rescue or standard PCR) could produce successful PCR amplification in any number of attempts. Second, we determined an “efficiency rate”. Using the subset of samples that amplified using both rescue and standard PCR (N = 55), efficiency rate is based on the number of successful amplifications per PCR attempt.
2.8. Determining mechanism of rescue
Although all extracts passed the initial test for inhibition, as described above under section 2.3., it is possible that some level of inhibitors still exist, but at a threshold below that which would render PCR amplification of the aDNA positive control impossible. We hypothesized that in cases where only very small amounts of DNA are present in the salmonid
DNA extracts, even undetectable low levels of inhibition might be sufficient to hinder amplification of the salmonid mtDNA. To investigate if rescue PCR is capable of overcoming this potential problem, we designed two complementary tests. The first test was designed to simulate an incremental decrease of DNA concentration. Two pools of DNA were created from salmon extracts (N = 24 for each pool) that individually amplified for the 12S fragment using standard PCR. These pools were then diluted 1:1, 1:5, 1:10, 1:25, 1:50, 1:75, 1:100, 1:150, and
1:200 with DNA-free water. Two replicates for each, as well as an undiluted pool, were then subjected to standard and rescue PCR for each dilution value [undiluted (1:0) – 1:200] and checked for amplification (See Figure 2 for schematic illustration of the method). It is important
128
to note that in this test inhibitors would be diluted at a rate proportional to that of the target
DNA and, therefore, results would vary according to the effect of the DNA concentration and not the relationship between inhibitors and the DNA, which here is a constant.
The second test was designed to simulate a decreasing concentration of target DNA relative to that of the inhibitors. We used a modification of the inhibition test described in
Kemp, et al. (2014a). In our modified test, the aDNA control (northern fur seal DNA) was diluted
1:1, 1:5, 1:10, 1:25, 1:50, 1:75, 1:100, 1:150, and 1:200 and each, along with an undiluted control (1:0). These were then tested in replicates of four, each spiked with an individual salmon extract. In this set of experiments, while the inhibitors present in the aDNA control were diluted equally with the targeted northern fur seal DNA, the inhibitors in the pool of salmon DNA should have been introduced equally into each reaction. Using primers specific to the aDNA control (described in section 2.3), amplification targeted the northern fur seal DNA.
As such, the level of total inhibitors (from both the northern fur seal and salmon DNA extracts) to the target DNA (northern fur seal) was increased across the dilutions (See Figure 3 for schematic illustration of the method.) Each combination was tested in replicate for both standard and rescue PCR.
2.9. Statistical comparisons
Chi-square tests of independence were used to test for significant differences at the 0.05 level of probability between treatments. Significance was determined for differences in reagent concentration at the standard, rescue +10%, rescue +25%, and rescue +50% levels as well as
129
across all attempts and when grouped by extraction method or geographical group, as well as for differences in efficiency rate between rescue and standard PCR.
2.10. Sequencing confirmation
Amplicons from 55 extracts, ones subjected to both standard and rescue PCR, were submitted for sequencing in both the forward and reverse directions using the same primers utilized for amplification. Product clean-up and sequencing were performed by Molecular
Cloning Laboratories (South San Francisco, CA). Sequences were aligned and the priming regions were trimmed using Sequencher v 4.8 (Gene Codes; Ann Arbor, MI). Sequence quality scores were determined using data provided by Molecular Cloning Laboratories. Each base was assigned a score between 0 and 60 as part of the sequencing process, with ranges of 20 for low, medium and high confidence. All bases scoring in the medium or high range (21-60) were combined to calculate the percent quality for the sequence as a whole. A sequence with a quality score of 75% indicates that 75% of the bases in the sequence were of medium to high confidence. All sequencing results were compared to the NCBI nucleotide database using the
Basic Local Alignment Search Tool (BLAST) to determine species and gene region.
3. RESULTS AND DISCUSSION
3.1. Tests of varying increase levels
In the test of standard PCR against rescue PCR of various levels of increased reagents
(10%, 25%, and 50%) results were tabulated for each primer set, as well as combined where a success was counted if an extract amplified for either primer set at a given increase (Table 1).
130
For the D-Loop primers, of the 30 extracts tested, 3 amplified (extracts 9, 12, 26) using standard
PCR, 7 amplified (extracts 8, 9, 13, 19, 25-27) using the 10% increase, 16 amplified (extracts 1, 2,
3, 6, 8-10, 13, 20, 22-27, 30) using the 25% increase, and 13 amplified (extracts 1-3, 6, 8, 9, 13,
14, 19) using the 50% increase. For the 12S primer set, 8 amplified (extracts 12, 15, 22-27) using standard PCR, 12 amplified (extracts 9, 12, 15, 17, 20, 22-27, 30) using the 10% increase, 17 amplified (extracts 1, 3, 6, 9-12, 14, 16, 17, 21-23, 25-27, 30) using the 25% increase, and 14 amplified (extracts 2, 4-6, 9, 10, 12, 15, 19, 20, 22, 25, 26, 29) using the 50% increase in reagents. In the combined dataset, where a success was counted if a sample amplified for either primer set at a given increase, standard PCR amplified 9 total extracts, the 10% increase amplified 15 extracts, a 25% increase amplified 22 extracts, and the 50% increase amplified 19 extracts. Both the +25% and +50% rescue PCR treatments resulted in significantly more amplification over standard PCR (P < 0.000 and P = 0.010, respectively). There was no statistically significant difference indicated for differences between the other levels (standard vs. 10%, 10% vs. 25%, 10% vs. 50%, or 25% vs. 50%).
Based on the number of successful amplifications, all rescue PCR treatments (+10%, +25%, and +50%) outperformed standard PCR and the +25% rescue outperformed +10% and +50%.
Despite a lower overall success rate, the 50% increase permitted amplification of four of the samples that could not be amplified using the lower reagent concentrations. However, this higher reagent concentration also resulted in 3 non-target (NT) amplifications (indicated by multiple bands) using the D-Loop primers (extracts 22, 27, 30) and 2 NT amplifications using the
12S primers (extracts 13, 27), for a total of 4 independent extracts producing non-target amplification (extracts 13, 22, 27, 30). In all four cases, lower reagent concentrations were able
131
to amplify target DNA and the 50% increase resulted in multiple bands. Therefore, we conclude that a 50% reagent increase may be a good strategy to attempt on a set of samples if additional amplifications are desired after attempting rescue PCR at +25%.
3.2. Effect of 25% increases of individual reagents and combinations of reagents
In this test, two extracts (extract 1 and 2) failed to produce target DNA amplicons and none of the extracts benefitted from increasing single reagents (Table 2). For increases in combinations of two reagents, six extracts benefitted from an increase in dNTPs & MgCl2
(extracts 3-8), four from MgCl2 & polymerase (extracts 4-7), eight from MgCl2 & primers
(extracts 3, 5-11), one from dNTPs & polymerase (extract 5), five from dNTPs & primers
(extracts 4-6, 9, 11), and six from polymerase & primers (extracts 4-7, 11, 12). For combinations of three reagent increases, two extracts benefitted from increasing dNTPs & MgCl2 & polymerase (extract 7, 8), eight from dNTPs & MgCl2 & primers (extracts 4-7, 9-12), six from
MgCl2 & polymerase & primers (extracts 3, 5-7, 9, 11), and four from dNTPs & polymerase & primers (extract 4, 5, 8, 11). Nine of the ten extracts that indicated amplification of target DNA for any treatment (i.e., not including extracts 1 and 2) benefitted from rescue PCR, or increasing all four reagents (extracts 3-5, 7-12).
Two commonly employed methods to increase PCR success rates on difficult samples is to increase either the concentration of MgCl2 or amount of polymerase (Alaeddini 2012; Opel et al. 2010; Schrader et al. 2012). These strategies are designed to overcome inhibitory substance that act directly on the polymerase enzyme or that sequester the magnesium cofactor.
However, it is possible that inhibitors could act on any component of PCR to prevent
132
amplification (Alaeddini 2012). This principal, and the great complexity of inhibitor action, is evidenced in our results. No single reagent increase resulted in amplification for any of the twelve extracts tested and results were mixed for combinations of two and three reagent increases. For example, amplification from Extract 3 was made possible by: 1) increased percent combination of dNTPs & MgCl2, 2) increased percent combination of MgCl2 & polymerase & primers, as well as 3) when all reagents were increased. Intriguingly, this same extract did not amplify with the increased percent combination of dNTPs & MgCl2 & polymerase. These incongruent amplification results were seen throughout these experiments, highlighting the stochastic and complex nature of inhibition. Consequently, increasing only a portion of the PCR reagents may not provide a consistently successful strategy. In the case of rescue PCR, which is an increase in all reagent components, many possible inhibitory combinations are accounted for resulting in a reduction of stochastic effect, and an increase in overall amplification success.
3.3. Sequencing results
For the subset of 55 extracts used for sequencing confirmation, 38 amplified with standard PCR and 44 with rescue PCR (Table 3). The average sequence quality score for each method was approximately equal, with 80.1% and 81.4% confidence scores for standard and rescue PCR generated sequences, respectively. In cases where templates generated under both standard and rescue PCR were sequenced (extracts 1 – 27) the sequences generated using standard PCR were identical to those generated using rescue PCR in every case. All the standard
PCR (N=38) amplifications were confirmed as target DNA and 91% (40 of 44) of those generated
133
from rescue PCR were confirmed as target DNA. Four sequences generated using rescue PCR
(extracts 52 – 55) matched to human DNA.
Rescue PCR appears to be more prone to amplification of non-target DNA than is standard
PCR. This was directly evidenced in the subset of samples that were selected for sequencing. In this experiment, six non-target amplifications were seen in 55 attempts, all originating from rescue PCR. Of these, two were obvious on the gel, observed as the presences of multiple bands. However, the remaining four were bands that appeared to be of the correct size but sequencing results revealed the DNA as human in origin. As determined from the sequences, the actual length of these amplicons ranged from 184 – 195 bp compared to the 189 bp expected from target DNA. Other cases of non-target DNA, indicated by incorrect size or by the observation of multiple bands is noted throughout the tests in this study and has been seen as we continue to apply the rescue method in our laboratory. Previous studies utilizing increased polymerase or MgCl2 have indicated that this modification can lead to increased non-specific binding of primers (Edwards et al. 2004). However, the risk of non-target amplification may be greater for our 12S primer set, as this region of mtDNA that is well conserved across species
(Melton and Holland 2007; Yang et al. 2014). Conserved areas of the 12S portion of mtDNA have been noted for animal species including amphibians, fish, and mammals (Yang et al. 2014).
Less conserved genetic targets may lower the rate of non-target amplification. Cases of single- band indicated non-target DNA using +25% rescue PCR were only observed in extracts where standard PCR failed, indicating that rescue PCR is likely amplifying non-target DNA in the absence of target DNA, not instead of target DNA.
134
3.4. Application of method and efficiency rates between standard and rescue PCR
It was possible to amplify 49% (55 of 112) of the extracts in this study using either standard or rescue PCR (Table 4). The remaining 51% were amplifiable only with rescue PCR. The proportion of samples that were amplifiable using either standard or rescue PCR differed significantly (P <
0.000). To determine if extraction method or sample collection had an influence on the proportion of samples that could be amplified under either PCR recipe, we grouped extracts categorically and determined success rates for each method within the groupings. Standard
PCR was used to successfully amplify 50% (24 of 48) of the samples generated using extraction method 1 and 48% (31 of 64) of the extracts generated using extraction method 2. Differences in the proportion of amplifications possible with standard PCR or rescue PCR was significant for both extraction methods (P < 0.000 in both cases). For extracts organized by geographical group, standard PCR was able to amplify 60% (41 or 68) of the Snake River extracts and 32% (14 or 44) of the Spokane River extracts. Statistical significance was indicated for the differences in amplification success rates between standard and rescue PCR in both the Snake and Spokane groups (P < 0.000 in both cases).
All PCR preparations are subject to some level of stochasticity. Each aliquot of a DNA extract will have varying amounts of inhibitors and DNA and, thus, mixed results may be seen across multiple PCR reactions from a single extract. In fact, this is a commonly cited observation in the laboratory; amplification of aDNA can be sporadic. Mixed outcomes are likely to occur frequently in eluates where the concentration of target DNA and inhibitors exist near the threshold where amplification or failure are equally likely. By random chance one draw from the extract may contain inhibitors above the threshold of amplification while the next has
135
inhibitors below the threshold, allowing amplification to complete. This effect was commonly observed in our study, with multiple PCR reactions necessary to obtain amplification. We quantified this effect using the calculation of an efficiency rate, or the number of successful amplifications per PCR attempted. Standard PCR had an efficiency rate of 59% (58 of 98 attempts resulted in amplification) while rescue PCR was 88% (58 of 66 attempts resulted in amplification) (Table 4). Statistical significance was indicated (P < 0.000) for the differences in efficiency between the PCR types (P = 0.005).
When the subset of amplicons was submitted for sequencing (section 3.3.) four of the
44 amplifications generated with rescue PCR appeared to be the correct size on the agarose gel, but were confirmed as human DNA (Table 3). This potential miscall rate (9%) was applied to the amplification counts for rescue PCR and reevaluated for statistical significance (Table 4). In all cases, significant differences between standard and rescue PCR were maintained.
In the present study, rescue PCR significantly outperformed standard PCR. This enhancement does not appear to be a function of the extraction methods employed here, indicating that success of rescue PCR is independent of how the DNA was extracted and purified. The benefit of rescue PCR was particularly evident in the case of the Spokane River collection. If processed using the standard PCR protocol these samples would have only produced 14 amplifications from the 44 DNA extracts in this group. However, rescue PCR permitted amplification from an additional 30 extracts, resulting in amplification from 100% of the samples.
136
3.5. Mechanism of rescue
There was no difference in the amplification success of standard and rescue PCR for a given dilution treatment of the pooled salmon DNA (Figure 4). For salmon pool 1 (SP1), at least faint amplification (even if inadequate for sequencing) was observed for the undiluted and 1:1 dilution using both standard and rescue PCR. As visually assessed, neither PCR method produced amplification at any of the further dilutions. For salmon pool 2 (SP2), amplification was achieved from the undiluted samples through the 1:25 dilutions for both standard and rescue PCR treatments. No amplification was possible for either standard or rescue PCR at any further dilution.
For the test inducing increases of inhibitors relative to the amounts of northern fur seal
DNA, more striking differences are observed between the normal and rescue PCR results
(Figure 5). The positive controls (aDNA control diluted but not spiked with salmon DNA) indicate that both rescue and standard PCR can produce amplification up to the 1:150 dilutions.
When additional inhibitors were introduced using the salmonid DNA spike, standard PCR could produce amplification up to the 1:10 dilutions. Rescue PCR was able to produce amplification through the 1:150 dilutions. Amplification strength does appear to drop off after the 1:25 dilutions and only very faint amplification was possible from the 1:50 – 1:150 dilutions.
Attempts to determine the mechanism of rescue PCR was based on two complimentary tests focused on the concentration of target DNA and inhibitors. When the concentration of
DNA was reduced in equal proportion to the inhibitors present, the amplified products (or lack thereof), appear consistent between standard and rescue PCR across all dilutions (1:0 – 1:200).
Under these conditions, rescue PCR had no positive affect on amplification success. However,
137
when the amount of target DNA was reduced at a rate different from that of inhibitors, rescue
PCR was able to produce amplification when standard PCR could not. As the aDNA control sample is diluted, the target DNA and inhibitors from that sample are diluted at equal rates. The salmon DNA spike is added to all dilution treatments equally, introducing some level of additional inhibitors without increasing the amount of northern fur seal target DNA. Thus, each dilution treatment has a decreasing amount of target DNA relative to the amount of total inhibitors. In our tests, both standard and rescue PCR were able to amplify the target DNA up to a 1:10 level dilution. However, only rescue PCR was able to produce any amplification in the remaining dilutions. Non-spiked controls were run for all dilutions indicating that sufficient DNA existed for amplification by both standard and rescue PCR up to the 1:150 dilutions and so amplification failure in these reactions is a function of inhibitory action(s). Based on our results we conclude that rescue PCR circumvents the problem associated with overcoming the combination of reduced DNA concentrations relative to the amount of inhibitors present in a sample.
3.6. Conclusions
Despite wide-ranging efforts to remove inhibitors, it is likely that some amount of inhibitors are present in all DNA eluates. Samples with very low DNA concentrations are unlikely to benefit from some of the previously described methods to further reduce inhibitors.
Specifically, direct dilution of sample reduces DNA along with inhibitors while additional silica treatments will result in further losses of DNA. Samples with such low copy numbers may not
138
be able to withstand such reductions. Rescue PCR provides a means by which to amplify DNA without risking loss or reduction of the target.
Our results demonstrate a clear ability of reagent-rich PCR mixes to rescue DNA (i.e., rescue PCR). However, the application of this method should not be applied blindly or with the notion that it solves all problems. Rescue PCR represents a simple and robust addition to the suite of options currently available to work with DNA in the presence of inhibition, especially ancient and degraded DNA. This method appears to have particular value when applied to samples where the relationship between DNA and inhibitors concentration may be of particular importance. For the study of some ancient and/or degraded specimens, even the increased cost associated with a reagent-rich PCR recipe is probably offset by the improved efficiency of rescue PCR over standard PCR.
139
LITERATURE CITED
Al-Soud, W. A., and P. Rådström. 1998. Capacity of nine thermostable DNA polymerases to mediate DNA amplification in the presence of PCR-inhibiting samples. Applied and Environmental Microbiology 64(10):3748-3753.
Al-Soud, W. A., and P. Rådström. 2001. Purification and characterization of PCR-inhibitory components in blood cells. Journal of Clinical Microbiology 39(2):485-493.
Alaeddini, R. 2012. Forensic implications of PCR inhibition—a review. Forensic Science International: Genetics 6(3):297-305.
Barta, J. L., C. Monroe, S. J. Crockford, and B. M. Kemp. 2014a. Mitochondrial DNA preservation across 3000 year old northern fur seal ribs is not related to bone density: Implications for forensic investigations. Forensic Science International 239:11-18.
Barta, J. L., C. Monroe, and B. M. Kemp. 2013. Further evaluation of the efficacy of contamination removal from bone surfaces. Forensic Science International 231(1):340- 348.
Barta, J. L., C. Monroe, J. E. Teisberg, M. Winters, K. Flanigan, and B. M. Kemp. 2014b. One of the key characteristics of ancient and forensic DNA, low copy number, may be a product of its extraction. Journal of Archaeological Science 46:281-289.
Bartlett, J. M. S., and D. Stirling. 2003. A short history of the polymerase chain reaction. Pages 3-6 in J. M. S. Bartlett, and D. Stirling, editors. Methods in Molecular Biology: PCR Protocols, 2nd edition. Humana Press, Totowa, NJ.
Belec, L., J. Authier, M. C. Eliezer‐Vanerot, C. Piédouillet, A. S. Mohamed, and R. K. Gherardi. 1998. Myoglobin as a polymerase chain reaction (PCR) inhibitor: a limitation for PCR from skeletal muscle tissue avoided by the use of Thermus thermophilus polymerase. Muscle & Nerve 21(8):1064-1067.
Bickley, J., J. Short, D. McDowell, and H. Parkes. 1996. Polymerase chain reaction (PCR) detection of Listeria monocytogenes in diluted milk and reversal of PCR inhibition caused by calcium ions. Letters in Applied Microbiology 22(2):153-158.
Brauner, D. 1990. Archaeological data recovery at Hatiuhpuh, 45WT134, Whitman County, Washington. Department of Anthropology, Oregon State University.
Browman, D. L., and D. A. Munsell. 1969. Columbia Plateau prehistory: Cultural development and impinging influences. American Antiquity:249-264.
Burkardt, H.-J. 2000. Standardization and quality control of PCR analyses. Clinical Chemistry and Laboratory Medicine 38(2):87-91.
140
Cui, Y., J. Lindo, C. E. Hughes, J. W. Johnson, A. Hernandez, B. M. Kemp, J. Ma, R. Cunningham, B. Petzelt, J. Mitchell, D. Archer, J. S. Cybulski, and R. S. Malhi. 2013. Ancient DNA analysis of mid-Holocene individuals from the Northwest Coast of North America reveals different evolutionary paths for mitogenomes. PLoS One 8(7):e66948.
Daugherty, R. D., B. A. Purdy, and R. Fryxell. 1967. The descriptive archaeology and geochronology of the Three Springs Bar archaeological site, Washington. Washington State University Laboratory of Anthropology.
Demeke, T., and G. R. Jenkins. 2010. Influence of DNA extraction methods, PCR inhibitors and quantification methods on real-time PCR assay of biotechnology-derived traits. Analytical and Bioanalytical Chemistry 396(6):1977-1990.
Eckhart, L., J. Bach, J. Ban, and E. Tschachler. 2000. Melanin binds reversibly to thermostable DNA polymerase and inhibits its activity. Biochemical and Biophysical Research Communications 271(3):726-730.
Edwards, C. J., D. E. MacHugh, K. M. Dobney, L. Martin, N. Russell, L. K. Horwitz, S. K. McIntosh, K. C. MacDonald, D. Helmer, and A. Tresset. 2004. Ancient DNA analysis of 101 cattle remains: Limits and prospects. Journal of Archaeological Science 31(6):695-710.
Eilert, K. D., and D. R. Foran. 2009. Polymerase resistance to polymerase chain reaction inhibitors in bone. Journal of Forensic Sciences 54(5):1001-1007.
Funk, P. S. 1998. Tucannon and Harder phase economic faunal analysis at Three Springs Bar (45FR39), Franklin County, Washington.
Galm, J. R. 1994. A design for management of cultural resources in the Lake Roosevelt basin of northeastern Washington, volume 100. Archaeological and Historical Services, Eastern Washington University.
Gilbert, M. T. P., and E. Willerslev. 2007. Rescuing ancient DNA. Nature Biotechnology 25(8):872-874.
Grier, C., K. Flanigan, M. Winters, L. G. Jordan, S. Lukowski, and B. M. Kemp. 2013. Using ancient DNA identification and osteometric measures of archaeological Pacific salmon vertebrae for reconstructing salmon fisheries and site seasonality at Dionisio Point, British Columbia. Journal of Archaeological Science 40(1):544-555.
Halffman, C. M., B. A. Potter, H. J. McKinney, B. P. Finney, A. T. Rodrigues, D. Y. Yang, and B. M. Kemp. 2015. Early human use of anadromous salmon in North America at 11,500 y ago. Proceedings of the National Academy of Sciences 112(40):12344-8.
Hedman, J., A. Nordgaard, C. Dufva, B. Rasmusson, R. Ansell, and P. Rådström. 2010. Synergy between DNA polymerases increases polymerase chain reaction inhibitor tolerance in forensic DNA analysis. Analytical Biochemistry 405(2):192-200.
141
Hicks, B. A. 2004. Marmes Rockshelter: A final report on 11,000 years of cultural use. Washington State University.
Huggett, J. F., T. Novak, J. A. Garson, C. Green, S. D. Morris-Jones, R. F. Miller, and A. Zumla. 2008. Differential susceptibility of PCR reactions to inhibitors: an important and unrecognised phenomenon. BMC Research Notes 1(1):70.
Jenkins, S. L. 2011. An analysis of the faunal remains from Windust Cave C (45FR46), Washington. Washington State University.
Jordan, L. G., C. A. Steele, and G. H. Thorgaard. 2010. Universal mtDNA primers for species identification of degraded bony fish samples. Molecular Ecology Resources 10(3):581- 581.
Juen, A., and M. Traugott. 2006. Amplification facilitators and multiplex PCR: Tools to overcome PCR-inhibition in DNA-gut-content analysis of soil-living invertebrates. Soil Biology and Biochemistry 38(7):1872-1879.
Katcher, H., and I. Schwartz. 1994. A distinctive property of Tth DNA polymerase: enzymatic amplification in the presence of phenol. Biotechniques 16(1):84-92.
Kemp, B. M., R. S. Malhi, J. McDonough, D. A. Bolnick, J. A. Eshleman, O. Rickards, C. Martinez- Labarga, J. R. Johnson, J. G. Lorenz, E. J. Dixon, T. E. Fifield, T. H. Heaton, R. Worl, and D. G. Smith. 2007. Genetic analysis of early Holocene skeletal remains from Alaska and its implications for the settlement of the Americas. American Journal of Physical Anthropology 132:605-621.
Kemp, B. M., C. Monroe, K. G. Judd, E. Reams, and C. Grier. 2014a. Evaluation of methods that subdue the effects of polymerase chain reaction inhibitors in the study of ancient and degraded DNA. Journal of Archaeological Science 42:373-380.
Kemp, B. M., C. Monroe, and D. G. Smith. 2006. Repeat silica extraction: A simple technique for the removal of PCR inhibitors from DNA extracts. Journal of Archaeological Science 33(12):1680-1689.
Kemp, B. M., and D. G. Smith. 2010. Ancient DNA methodology: Thoughts from Brian M. Kemp and David Glenn Smith on “Mitochondrial DNA of protohistoric remains of an Arikara population from South Dakota”. Human Biology 82(2):227-238.
Kemp, B. M., M. Winters, C. Monroe, and J. L. Barta. 2014b. How much DNA is lost? Measuring DNA loss of short-tandem-repeat length fragments targeted by the PowerPlex 16® System using the Qiagen MinElute Purification Kit. Human Biology 86(4):313-329.
Kreader, C. A. 1996. Relief of amplification inhibition in PCR with bovine serum albumin or T4 gene 32 protein. Applied and Environmental Microbiology 62(3):1102-1106.
142
Leonhardy, F. C. 1969. Artifact assemblages and archaeological units at Granite Point Locality 1 (45WT41), Southeastern Washington.
McCord, B., A. Pionzio, and R. Thompson. 2015. Analysis of the effect of a variety of PCR inhibitors on the amplification of DNA using RealTimePCR, melt curves and STR analysis. U.S. Department of Justice.
Melton, T., and C. Holland. 2007. Routine forensic use of the mitochondrial 12S ribosomal RNA gene for species identification. Journal of Forensic Sciences 52(6):1305-1307.
Monroe, C., C. Grier, and B. M. Kemp. 2013. Evaluating the efficacy of various thermo-stable polymerases against co-extracted PCR inhibitors in ancient DNA samples. Forensic Science International 228(1):142-153.
Moss, M. L., K. G. Judd, and B. M. Kemp. 2014. Can salmonids (Oncorhynchus spp.) be identified to species using vertebral morphometrics? A test using ancient DNA from Coffman Cove, Alaska. Journal of Archaeological Science 41:879-889.
Moss, M. L., D. Y. Yang, S. D. Newsome, C. F. Speller, I. McKechnie, A. D. McMillan, R. J. Losey, and P. L. Koch. 2006. Historical ecology and biogeography of North Pacific pinnipeds: Isotopes and ancient DNA from three archaeological assemblages. Journal of Island & Coastal Archaeology 1(2):165-190.
Nakonechny, L. D. 1998. Archaeological analysis of Area A, Wexpusnime site (45GA61).
Opel, K. L., D. Chung, and B. R. McCord. 2010. A study of PCR inhibition mechanisms using real time PCR. Journal of Forensic Sciences 55(1):25-33.
Palumbi, S. R. 1996. Nucleic acids II: The polymerase chain reaction. Pages 205-247 in D. M. Hillis, C. Moritz, and B. K. Mable, editors. Molecular Systematics. Sinauer Associates, Sunderland, MA.
Rådström, P., R. Knutsson, P. Wolffs, M. Lövenklev, and C. Löfström. 2004. Pre-PCR processing. Molecular Biotechnology 26(2):133-146.
Rådström, P., C. Löfström, M. Lövenklev, R. Knutsson, and P. Wolffs. 2008. Strategies for overcoming PCR inhibition. Cold Spring Harbor Protocols 2008(3):pdb. top20.
Rice, H. S. 1965. The cultural sequence at Windust Caves. Washington State University.
Rossen, L., P. Nørskov, K. Holmstrøm, and O. F. Rasmussen. 1992. Inhibition of PCR by components of food samples, microbial diagnostic assays and DNA-extraction solutions. International Journal of Food Microbiology 17(1):37-45.
Schalk, R. F., R. R. Mierendorf, and D. L. Olson. 1983. The 1978 and 1979 Excavations at Strawberry Island in the McNary Reservoir. DTIC Document.
143
Schrader, C., A. Schielke, L. Ellerbroek, and R. Johne. 2012. PCR inhibitors–occurrence, properties and removal. Journal of Applied Microbiology 113(5):1014-1026.
Tebbe, C. C., and W. Vahjen. 1993. Interference of humic acids and DNA extracted directly from soil in detection and transformation of recombinant DNA from bacteria and a yeast. Applied and Environmental Microbiology 59(8):2657-2665.
Thompson, R. S. 1985. Paleoenvironmental investigations at Seed Cave (Windust Cave H- 45FR46), Franklin County, Washington. DTIC Document.
Weyant, R., P. Edmonds, and B. Swaminathan. 1990. Effect of ionic and nonionic detergents on the Taq polymerase. Biotechniques 9(3):308-309.
Wiedbrauk, D. L., J. C. Werner, and A. M. Drevon. 1995. Inhibition of PCR by aqueous and vitreous fluids. Journal of Clinical Microbiology 33(10):2643-2646.
Wilcox, A. W., D. L. Fisher, L. Mitchell, M. Wadhams, M. M. Holland, P. S. Sledzik, and V. W. Weedn. 1993. Extraction, evaluation, and amplification of DNA from decalcified and undecalcified United States Civil War bone. Journal of Forensic Science 38(1):60-68.
Wilson, I. G. 1997. Inhibition and facilitation of nucleic acid amplification. Applied and Environmental Microbiology 63(10):3741.
Winters, M., J. L. Barta, C. Monroe, and B. M. Kemp. 2011. To clone or not to clone: method analysis for retrieving consensus sequences in ancient DNA samples. PLoS One 6(6):e21247.
Yang, D. Y., B. Eng, J. S. Waye, J. C. Dudar, and S. R. Saunders. 1998. Technical note: Improved DNA extraction from ancient bones using silica-based spin columns. American Journal of Physical Anthropology 105(4):539-543.
Yang, L., Z. Tan, D. Wang, L. Xue, M.-x. Guan, T. Huang, and R. Li. 2014. Species identification through mitochondrial rRNA genetic analysis. Scientific Reports 4.
Ye, J., A. Ji, E. J. Parra, X. Zheng, C. Jiang, X. Zhao, L. Hu, and Z. Tu. 2004. A simple and efficient method for extracting DNA from old and burned bone. Journal of Forensic Sciences 49(4):754-759.
144
FIGURES AND TABLES
Figure 1. Excavation site locations and approximate ages for samples used in this study.
Samples were comprised of two collections, those from in and near the Snake River basin (1 –
4) and those from a location near the Spokane River (5).
145
Figure 2. Schematic of experimental set-up to investigate effect of rescue PCR on decreasing concentrations of DNA. A pooled sample of salmon DNA was subject to nine dilutions and amplification was attempted using standard and rescue PCR.
146
Figure 3. Schematic of experimental set-up to investigate the success of rescue PCR with changing DNA-inhibitor ratios. An aDNA control consisting of pooled DNA from northern fur seal was subject to nine dilutions; each dilution was then spiked with undiluted salmon DNA. Amplification was attempted using standard and rescue PCR mixes created using primers designed to target the northern fur seal DNA and not salmonid DNA.
147
Figure 4. Amplification results for test inducing decreasing DNA concentrations. Two pools of salmon DNA (SP1 and SP2) were diluted and amplification was attempted using standard and rescue PCR at +25%. Two replicates were run for each dilution of a DNA pool. Results indicate little difference between amplification capabilities of standard and rescue PCR. (Note: Results are from a single gel image that has been reordered to support interpretation, no other alterations were made to the images).
Dilution No dilution 1:1 1:5 1:10 1:25
DNA Pool SP1 SP2 SP1 SP2 SP1 SP2 SP1 SP2 SP1 SP2
Standard PCR Standard Rescue PCR Rescue
Dilution 1:50 1:75 1:100 1:150 1:200
DNA Pool SP1 SP2 SP1 SP2 SP1 SP2 SP1 SP2 SP1 SP2
Standard PCR Standard Rescue PCR Rescue
148
Figure 5. Figure follows on next page. Amplification results for test inducing DNA-inhibitor ratio change. An aDNA positive control was diluted and the spiked with undiluted salmon DNA (four distinct DNA samples, numbered 1-4 here). Primers used in the PCR target the aDNA control.
Amplification was then attempted using standard and rescue PCR at +25%. Two replicates were done for each dilution/spike combination. Positive controls, indicated by “+C”, are the dilution treatment with no salmon DNA spike. Results indicate amplification success of rescue PCR in cases where standard PCR fails. (Note: results are from three gel images that have been combined and reordered to facilitate interpretation, no other alterations were made to the images).
149
Dilution No dilution 1:1 1:5 1:10 1:25
DNA Spike 1 2 3 4 +C 1 2 3 4 +C 1 2 3 4 +C 1 2 3 4 +C 1 2 3 4 +C
Standard PCR Standard Rescue PCR Rescue
Dilution 1:50 1:75 1:100 1:150 1:200 150 DNA Spike 1 2 3 4 +C 1 2 3 4 +C 1 2 3 4 +C 1 2 3 4 +C 1 2 3 4 +C
Standard PCR Standard Rescue PCR Rescue
Table 1. Table follows on next page. Results of PCR amplification tested with standard PCR as well as 10%, 25%, and 50% increases using two primer sets. Combined results indicate if an extract amplified using either primer set for a given PCR protocol. Amplification of target DNA
(inferred by band size) indicated with "+", no amplification indicated with "-", and "NT-M" indicates non-target DNA (multiple bands present on gel). No bands of incorrect size were seen in this experiment.
151
Primer set 1 (D-Loop) Primer set 2 (12S) Combined 0 10% 25% 50% 0 10% 25% 50% 0 10% 25% 50% Extract 1 - - + + - - + - - - + + Extract 2 - - - + - - - + - - - + Extract 3 - - + + - - + - - - + + Extract 4 - - + - - - - + - - + + Extract 5 ------+ - - - + Extract 6 - - + + - - + + - - + + Extract 7 ------Extract 8 - + + + - - - - - + + + Extract 9 + + + + - + + + + + + + Extract 10 - - + - - - + + - - + + Extract 11 ------+ - - - + - Extract 12 + - - - + + + + + + + + Extract 13 - + + + - - - NT-M - + + NT-M Extract 14 - - - + - - + - - - + + Extract 15 - - - - + + - + + + - + Extract 16 ------+ - - - + - Extract 17 - - - - - + + - - + + - Extract 18 ------Extract 19 - + - + - - - + - + - + Extract 20 - - + - - + - + - + + + Extract 21 ------+ - - - + - Extract 22 - - + NT-M + + + + + + + NT-M Extract 23 - - + - + + + - + + + - Extract 24 - - + + + + - - + + + + Extract 25 - + + + + + + + + + + + Extract 26 + + + + + + + + + + + + Extract 27 - + + NT-M + + + NT-M + + + NT-M Extract 28 - - - + ------+ Extract 29 ------+ - - - + Extract 30 - - + NT-M - + + - - + + NT-M Target + 3 7 16 13 8 12 17 14 9 15 22 19
152
Table 2. Results of 16 PCR treatments testing the effect of increasing individual reagents as well as all possible combinations or
reagents on 12 unique DNA extracts. Symbol "+" indicates amplification of the target DNA occurred (inferred by band size), "-"
indicates no amplification occurred, non-target amplification is indicated by either "NT-S" (when inferred by size) or "NT-M" (when
inferred by multiple bands).
Reagent(s) increased by 25% dNTPs
MgCl2 (+ buffer) Polymerase Primers 153 Extract 1 ------NT-S - NT-S NT-M - NT-S NT-S NT-M NT-S
Extract 2 - - - - - NT-S NT-M NT-M NT-M NT-M NT-M NT-M NT-M NT-M NT-M NT-M Extract 3 - - - - - + - - - + - - - - + + Extract 4 - - - - - + NT-M + + NT-M + NT-M + + NT-M + Extract 5 - - - - - + + + + + + NT-M + + + + Extract 6 - - - - - + NT-M + + + + NT-M + NT-M + NT-M Extract 7 - - - - - + NT-M - + + + + + NT-M + + Extract 8 - - - - - + - - - + NT-S + - + NT-M + Extract 9 ------+ - + - - + - + + Extract 10 ------+ - - + - - + Extract 11 ------+ - + + - + + + + Extract 12 ------+ - + - - +
Table 3. Results from attempts to amplify and sequence 55 extracts using both standard and rescue PCR at +25%. Extracts that failed to amplify in this test are indicated by "NA", non-target
DNA evidenced by multiple bands is indicated by "NT-M". All others were amplifications that appeared to be of the correct size and were submitted for sequencing. Raw sequence quality score and NCBI BLAST results including species and genetic region for match, number of matches expected by chance (E value), and percent similarity between the sequence generated and the BLAST match (Ident) are given for all attempts. Table continues on consecutive pages.
Sequence quality Results Match quality Standard Rescue Species (genetic region) E value Ident (%) Extract 1 83.8 93.2 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 2 85.0 89.7 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 3 48.0 89.0 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 4 39.2 30.1 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 5 72.3 86.5 Oncorhynchus tshawytscha (12S) 4.00E-25 100 Extract 6 85.5 79.7 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 7 84.5 85.8 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 8 75.3 90.6 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 9 85.5 79.9 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 10 84.3 85.6 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 11 86.7 86.0 Catostomus catostomus (12S) 2.00E-68 99 Extract 12 87.9 85.8 Catostomus catostomus (12S) 2.00E-68 99 Extract 13 91.9 93.2 Ptychocheilus oregonensis (12S) 4.00E-90 99 Extract 14 83.8 86.2 Ptychocheilus oregonensis (12S) 4.00E-90 99 Extract 15 83.9 80.1 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 16 71.6 81.9 Catostomus catostomus (12S) 2.00E-68 99 Extract 17 87.9 83.9 Catostomus catostomus (12S) 3.00E-70 99 Extract 18 81.1 78.0 Catostomus catostomus (12S) 2.00E-68 99 Extract 19 84.2 81.4 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 20 95.2 87.8 Oncorhynchus kisutch (12S) 3.00E-70 100 Extract 21 73.4 80.8 Oncorhynchus kisutch (12S) 3.00E-70 100 Extract 22 79.3 78.8 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 23 82.3 85.5 Catostomus catostomus (12S) 2.00E-68 99 Extract 24 76.6 91.0 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 25 90.4 86.6 Oncorhynchus tshawytscha (12S) 2.00E-68 99 Extract 26 90.6 72.6 Catostomus catostomus (12S) 3.00E-70 99
154
Sequence quality Results Match quality Standard Rescue Species (genetic region) E value Ident (%) Extract 27 93.4 88.6 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 28 52.2 NA Oncorhynchus tshawytscha (12S) 2.00E-68 99 Extract 29 34.8 NA Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 30 82.8 NA Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 31 89.7 NA Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 32 81.4 NA Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 33 89.2 NA Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 34 84.0 NA Catostomus catostomus (12S) 7.00E-67 99 Extract 35 85.1 NA Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 36 76.8 NA Catostomus catostomus (12S) 2.00E-68 99 Extract 37 91.5 NT-M Ptychocheilus oregonensis (12S) 4.00E-90 99 Extract 38 92.1 NT-M Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 39 NA 89.6 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 40 NA 88.8 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 41 NA 62.4 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 42 NA 71.1 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 43 NA 89.1 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 44 NA 87.7 Catostomus catostomus (12S) 3.00E-70 99 Extract 45 NA 81.3 Catostomus catostomus (12S) 3.00E-70 99 Extract 46 NA 83.3 Catostomus catostomus (12S) 3.00E-70 99 Extract 47 NA 81.2 Oncorhynchus tshawytscha (12S) 2.00E-68 99 Extract 48 NA 82.8 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 49 NA 88.2 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 50 NA 86.8 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 51 NA 81.7 Oncorhynchus tshawytscha (12S) 9.00E-71 100 Extract 52 NA 50.3 human (chromosomal) 4.00E-37 98 Extract 53 NA 93.0 human (chromosomal) 3.00E-25 94 Extract 54 NA 72.9 human (12S) 1.00E-64 99 Extract 55 NA 55.2 human (chromosomal) 9.50E-19 97.5
155
Table 4. Comparisons in application of method and efficiency between standard and rescue PCR at +25%. Successful application is
based on the ability of standard and/or rescue PCR to generate amplification in a given extract. Calculated for all extracts (Ext.) as
well as grouped by DNA extraction method (EM 1 and EM 2) and geography. Efficiency is the number of successful amplifications per
attempted PCR across extracts that amplified using both PCR methods. Values for rescue PCR are give both as indicated by gel
amplification and with a 9% reduction to account for miscalls of non-target DNA generating a band of the correct size.
Application of Method 156 Efficiency EM 1 EM 2 Snake River Spokane River
Ext. Amplified Ext. Amplified Ext. Amplified Ext. Amplified Ext. Amplified Ext. Amplified Standard 112 55 49% 48 24 50% 64 31 48% 68 41 60% 44 14 32% 98 58 59% Rescue (as indicated) 112 100% 48 100% 64 100% 68 100% 44 100% 66 58 88% Rescue (9% reduction) 102 91% 44 91% 58 91% 62 91% 40 91% 53 80% P (as indicated) < 0.000 < 0.000 < 0.000 < 0.000 < 0.000 < 0.000 P (with reduction) < 0.000 < 0.000 < 0.000 < 0.000 < 0.000 0.005
CHAPTER THREE
Archival archrival: Evidence of contamination via non-human DNA in formalin-preserved
specimens
1. INTRODUCTION
Heterochronously sampled specimens are especially informative for addressing questions related to evolution, adaptation, conservation, and ecology [e.g., (Chan et al. 2006; Hadly et al.
2004; Iwamoto et al. 2012; Orlando et al. 2002; Orlando et al. 2013; Pedersen et al. 2016)].
Museum archives collectively hold millions of such samples, many of which have been meticulously cataloged with associated contextual. Thus, these collections potentially represent treasuries of temporally diverse molecular data. The majority of wet collections held in museums consist of formalin-fixed, ethanol-preserved specimens (Schander and Kenneth
2003). Formalin, a dilution of formaldehyde, was first formulated in 1859 and subsequently became the fixative of choice for ecological and medical preservation applications (Schander and Kenneth 2003). Upon the application of formalin to biological tissues, formaldehyde combines with functional groups of amino acids causing protein denaturation (Schander and
Kenneth 2003). Subsequent hydrogen bonds between oxygen atoms and primary amines crosslink proteins with formaldehyde, forming methylene bridges between the functional groups (French and Edsall 1945; Schander and Kenneth 2003). Formalin efficiently penetrates biomolecules (Hopwood 1967), does not cause shrinkage (Jackson 1978), and has no risk of over fixation (Vachot and Monnerot 1996) making it ideal for preserving biological material for visual examination.
157
Although formalin is an ideal fixative for preserving morphological characteristics, that is not the case for the preservation of DNA. In addition to protein-protein crosslinks, formalin has been shown to link DNA to DNA and DNA to protein (Chang and Loew 1994; Chaw et al. 1980;
Crisan and Mattson 1993; Ma and Harris 1988; Schander and Kenneth 2003; Trifonov et al.
1967). DNA molecules that are linked to proteins may not be released during DNA extraction and/or subsequently interfere with polymerase activity, preventing DNA amplification via polymerase chain reaction (PCR) (Chang and Loew 1994; Karlsen et al. 1994; Schander and
Kenneth 2003). Several studies have indicated that the DNA recovered from formalin treated specimens is limited to short targets, possibly due to direct fragmentation (Campos and Gilbert
2012; Hykin et al. 2015; Libório et al. 2005; Schander and Kenneth 2003; Wong et al. 2014) or to alteration of DNA structure which limits recovery of longer targets (Karlsen et al. 1994).
Formaldehyde will also combine with nucleotides containing cytosine, guanine, and adenine to form a reactive methylene compound that prevents primer annealing, renaturation, and replication required for PCR (Karlsen et al. 1994; Schander and Kenneth 2003). Moreover, comparison of fixed and unfixed material from single samples detected artefactual point mutations in the fixed treatment group (De Giorgi et al. 1994; Williams et al. 1999). If undetected, such artifacts may lead one to incorrect conclusions over phylogenetic relationships and estimates of heterozygosity.
Methodological studies aimed to improve PCR amplification of genetic material from formalin-fixed specimens can largely be placed into one of three main categories, those that employ: 1) pre-extraction treatments, 2) treatments during extraction, or 3) post-extraction- repair. Pre-extraction treatments are commonly aimed at breaking DNA-protein crosslinks.
158
Such treatments include microwave radiation (Sato et al. 2001), alkali and chemically based treatments (Chatigny 2000; Gilbert et al. 2007; Johnson et al. 1995; Shedlock et al. 1997; Shi et al. 2002; Shi et al. 2004), extended exposure to proteinase K (Bramwell and Burns 1988; Bunker and Locker 1989; Dubeau et al. 1986; Schander and Kenneth 2003), or the use of heat (Gilbert et al. 2007; Trembath-Reichert et al. 2013; Wu et al. 2002). Treatments applied during extraction tend to focus on varying buffer composition (Krafft et al. 1995; Longy et al. 1997;
Shedlock et al. 1997) or the introduction of glycine, which may bind free formalin reduced during the extraction procedure (Shedlock et al. 1997). Post-extraction efforts utilize Taq DNA polymerase to repair nicked DNA (Bonin et al. 2003; Bonin et al. 2005). The list of studies summarized here encompasses only a small portion of those available. However, notably, the number of methodological studies far outweighs those that have successfully utilized such methodologies [see Wirgin et al. (1997) and Raja et al. (2011)].
We applied seven methodological variations to a variety of formalin preserved samples
(Figure 1 and Table 1). These activities were conducted in an effort to obtain DNA from a series of formalin-preserved samples, one of the goals of a parallel study that we were undertaking.
As such, no formalized, comparative study was conducted on these methods [for such descriptions see Gilbert et al. (2007) and Sato et al. (2001), among others]. However, cumulative results of our efforts did produce a comparative dataset for the methods applied and we summarize those results here. We also identify an additional obstacle in the pursuit of recovering molecular data from formalin-fixed specimens: contamination. This manuscript outlines the various methods applied to a series of samples as well as a subsequent pilot study into contamination risks associated with collections of formalin-preserved material.
159
2. METHODS
2.1. DNA Extractions
All DNA extractions and PCR preparations were conducted in the ancient DNA lab at
Washington State University (WSU), a laboratory located in a separate building from wherein
PCR and post-PCR activities are conducted. The ancient DNA laboratory is a dedicated workspace for processing degraded and low copy number (LCN) DNA samples, wherein strict protocols are employed to closely monitor and minimize the inadvertent introduction of contamination to ensure results are authentic (Kemp and Smith 2010).
Formalin
Seven extraction methods were applied to a set of 78 subsamples taken from 15 unique specimens (Table 1). A conceptual schematic of the methods is provided in Figure 1 and methodological differences are summarized below.
Four variations of extraction protocols utilizing heat treatments, in tandem with commercially available silica-based DNA extractions kits were applied following suggestions from Gilbert et al. (2007) (Methods 1A, 1B, 2A, and 2B depicted in Figure 1). Method 1A and 1B utilized the DNAeasy animal tissue kit (Qiagen; Valencia, CA) with the following modifications to the manufacturer’s protocol: Samples were initially rinsed in 1 mL TE buffer followed by two rinses using 1 mL of DNA free water. Lysis reagents, omitting proteinase K, were added following manufacturer’s instructions. Following method 1A, samples were incubated for 3 hours at 56°C to initiate tissue breakdown. Samples were then incubated at 95°C for 10 minutes and then immediately transferred to -20°C for 2 minutes. After this incubation, the
160
manufacturer’s protocol was again resumed through the DNA elution step, with a modified time of incubation on the filter of 3 minutes. Method 1A was applied to 11 subsamples of four unique specimens. Method 1B was as just described for Method 1A, omitting the 3-hour incubation treatment at 56°C. Method 1B was applied to seven subsamples from three unique specimens.
Methods 2A and 2B followed the QIAamp DNA Mini Kit protocol for animal tissue
(Qiagen; Valencia, CA) with the following modifications: Initially samples were submerged, with gentle rocking, for 24 hours at room temperature in either TE (method 2A) or EDTA (method
2B). Lysis reagents, omitting proteinase K, were added and samples incubated at 95°C for 10 minutes before immediate transfer to -20°C for 2 min. Proteinase K was then added following the manufacture’s protocol (20 µL for samples ≤ 0.025 g, increased proportionally for larger sample weights) and samples were incubated for 3 hours at 56°C. The remainder of the extraction followed the QIAamp tissue protocol. Method 2A was applied to 21 subsamples from
13 unique specimens and method 2B was applied to 18 subsamples from 11 unique specimens.
Methods 3A and 3B were modified phenol:chloroform extraction procedures. Method
3A followed Shiozawa et al. (1992) who reported successful DNA extraction from formalin- preserved trout tissues. This method utilizes an extended period of digestion. Samples were washed with submersion in TE9 buffer [500 mM Tris, 20 mM EDTA, 10 mM NaCl; (Goelz et al.
1985)] for 24 hours, with the buffer decanted and replaced every eight hours. Following washing, samples were digested in a buffer comprised of 10 mL TE9, 0.1 g SDS, and 5.0 mg proteinase K for 24 hours at 56°C. After 24 hours, an additional 0.1 g SDS and 5.0 mg proteinase
K were added and the sample incubated for 50 hours at 50°C. An equal volume of
161
phenol:chloroform:isoamyl alcohol (25:24:1) was added to the digestion buffer and the tubes were inverted several times to mix and then centrifuged at 10,000 rpm for 10 min. The aqueous phase was transferred to a new tube containing one volume of phenol:chloroform:isoamyl alcohol (25:24:1) and the centrifugation repeated. The resulting aqueous phase was transferred to a new tube containing one volume of chloroform:isoamyl alcohol (24:1) and the centrifugation repeated. The aqueous phase was then transferred to new tubes to which one volume of 3 M sodium acetate was added. To this combined volume, one volume of 95% ethanol was added and the DNA precipitated overnight at -20°C. Tubes were then centrifuged at 10,000 rpm for 10 min and the supernat carefully poured off and left to air dry via inversion.
After pellets were completely dry, DNA was resuspended in 100 µL of DNA-free water heated to
55°C. Method 3A was applied to ten subsamples from seven unique specimens.
Method 3B combined protocols described by Trembath-Reichert et al. (2013) and Kemp et al. (2007). Prior to extraction, samples were washed four times via 5-minute incubations in
DNA-free water. After washing, samples were placed in 3 mL EDTA and incubated with gentle rocking at room temperature for approximately seven days. Three milligrams proteinase K was added and samples were incubated for 3 hours at 65°C and then placed in a refrigerator overnight. Following this, an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1) was added to the EDTA and briefly vortexed to mix. Tubes were then centrifuged at 3,000 rpm for 5 min and the aqueous phase transferred to new tubes containing one volume phenol:chloroform:isoamyl alcohol (25:24:1). Tubes were vortexed and centrifuged as just described and the resulting aqueous phase was transferred to new tubes containing one volume chloroform:isoamyl alcohol (24:1). Tubes were vortexed briefly and centrifuged at
162
3,000 rpm for 3 min. The aqueous phase was transferred to a new tube to which one half volume of room temperature 5 M ammonium acetate one combined volume (equaling the aqueous phase and the ammonium acetate) of room temperature 100% isopropanol, as suggested by Hänni et al. (1995). DNA was precipitated overnight at room temperature and then pelleted via centrifugation at 3,000 rpm for 30 minutes. Isopropanol was gently poured off and tubes air dried via inversion for 15 minutes. DNA pellets were washed with 1 mL of 80% ethanol and vortexed. DNA was re-pelleted by an additional centrifugation at 3,000 rpm for 30 minutes before gently decanting the ethanol and air drying via inversion. Once completely dry,
DNA was resuspended in 100 µL of DNA-free water heated to 55°C. Method 3B was applied to three subsamples, each from a unique specimen.
Method 4 followed Nishiguchi et al. (2002) which was reported to have resulted in successful extraction of DNA from formalin preserved cyprinid fishes by Raja et al. (2011).
Samples were initially submerged, with gentle rocking, in 1 mL TE buffer overnight at room temperature. Digestion was performed via incubation in 500 µL of STE buffer (0.2% SDS and 250
µL of 10 M ammonium acetate). Samples were then ground manually using polypropylene pellet pestles (Sigma-Aldrich) and incubated for one hour at room temperature. Samples were then centrifuged at 14,000 rpm for 5 min. Resulting supernat was transferred to a new tube containing 2 volumes of ice cold (-20°C) 100% ethanol. Tubes were inverted several times and then placed in -20°C freezer for 1 h, or until DNA had precipitated. Tubes were then centrifuged at 14,000 rpm for 15 min and the supernat transferred to new tubes containing 1 volume of ice cold (-20°C) 70% ethanol. Tubes were then centrifuged at 14,000 rpm for 10 min to pellet DNA.
Ethanol was carefully poured off and tubes dried via inversion. Pellets were resuspended in 50
163
µL of TE buffer overnight at 4°C. Tubes were centrifuged 12,000 rpm for 20 min and the ethanol decanted. Pellets were washed using 1 mL of 70% ethanol and centrifuged at 12,000 rpm for 5 min. The ethanol was then carefully poured off and the tubes dried by inversion. After the pellets were completely dry, DNA was resuspended in 100 µL of DNA-free water heated to
55°C. Method 4 was applied to eight subsamples from four unique specimens.
2.2. PCR amplification and sequencing
Amplification via PCR was pursued using five variations of the standard PCR protocol
(described below). Each sample was subjected to two amplification attempts using standard
PCR as well as a “3X template PCR”, “rescue PCR” at +25% and +50%, and a “pre-amplification
PCR” protocol. Standard PCR was conducted in 25 µL reactions of: 1X Omni Klentaq Reaction
Buffer mix (containing a final concentration of MgCl2 at 3.5 mM), 0.32 mM dNTPs, 0.24 µM each of forward and reverse primer, 0.3 U of Omni Klentaq LA polymerase, and 2.5 µL of template DNA. PCRs consisted of an initial denaturation at 94°C for 3 min, followed by sixty cycles of 94°C for 15 s, 55°C for 15 s, and 68°C for 15 s. This was followed with a final extension at 68°C for 3 minutes. Twenty-five microliter rescue PCRs at +25% and +50% employ all reagents at +25% and +50% concentrations over standard PCR, respectively, without modification to the amount of template DNA introduced to the reactions (Johnson and Kemp in review). “3X template PCR” consisted of increasing the volume of DNA template three-fold over that used in standard PCR, resulting in a final volume of 7.5 µL template in 25 µL reactions.
“Pre-amplification PCR” represents a modification to standard PCR, described by Smith et al.
(2011). Two PCR reactions were initially prepared in the dedicated ancient DNA laboratory
164
space. The pre-amplification PCR consisted of 10 µL volumes containing a final concentration of
1X Omni Klentaq Reaction Buffer mix (with incorporated MgCl2 at 1.4 mM), 0.13 mM dNTPs,
0.96 µM each of forward and reverse primer, 0.12 U Omni Klentaq LA polymerase, and 1 µL of template DNA. The second PCR reaction was set up as a standard PCR (see above) minus template DNA. This second PCR reaction was placed in the freezer while the 10 µL pre- amplification PCRs were conducted as follows: 1) initial denaturation at 94°C for 3 min, 2) fourteen cycles at 94°C for 30 s, 55°C for 30 s, and 68°C for 30 s, and 3) final extension at 68°C for 10 min. Following completion of the 10 µL pre-amplification PCRs, 2.5 µL of these reaction volumes were added as template to the reserved standard PCR mix and the reaction was conducted as described above for standard PCR.
Amplification of mitochondrial DNA (mtDNA) from fish specimens (Cyprinus carpio,
Oncorhynchus mykiss, and Oncorhynchus tshawytscha) was conducted with the primers described in Jordan et al. (2010). It is important to note here that in their original study Jordan et al. (2010) reported the wrong orientation for the reverse primer. Described correctly, the primer sequences are: (i) “OST12S-F” 5’-GCTTAAAACCCAAAGGACTTG-3’ and (ii) “OST12S-R” 5’-
CTACACCTCGACCTGACGTT-3’. Amplification of smoothskin octopus (Benthoctopus leioderma) mtDNA was performed using primers described in Kore et al. (in review): (i) forward 5’-
AGACTAGGATTAGAGACCCTATTA-3’ and (ii) reverse 5’-GTTTAAAACGAGGATCATGAAATC-3’. The annealing temperature of 60°C was employed for this primer set.
Successful amplifications were initially confirmed via separation of amplicons on a 3% agarose gel against a 20 bp size-standard ladder (Bayou BioLabs). Product clean-up and sequencing in both directions were performed by Molecular Cloning Laboratories (South San
165
Francisco, CA). Sequences were aligned using Sequencher v 4.8 (Gene Codes; Ann Arbor, MI).
All sequencing results were compared to the NCBI nucleotide database using the Basic Local
Alignment Search Tool (BLAST) to determine species and to confirm target region.
2.3. Contamination
Following detection of non-target DNA (see section 3.1.) in the DNA eluates, a pilot study to investigate potential DNA contamination in formalin preserved specimens was conducted.
Twenty storage vessels (Figure 2), each containing a species of fish, were chosen at random from the fish collection of the Conner Museum at Washington State University (Table
3). From each vessel, 1 mL of ethanol and 0.040 – 0.265 g of paper from the lid was collected
(Figure 2). Collection tools (scissors and forceps) were decontaminated via submersion in 6%
(w/v) sodium hypochlorite (bleach) for 30 seconds, followed by two pour-over rinses with DNA- free water. Disposable gloves were discarded and replaced between each collection.
DNA extraction from the ethanol samples was conducted by ethanol precipitation following Nishiguchi et al. (2002). DNA extraction from paper samples was conducted by dividing the total sample into three portions of equal weight (approximately 0.013 – 0.088 g), to which three variations of extraction method 2A were applied: (1) method 2A without modification, (2) omitting initial TE wash, and (3) omitting both initial TE wash and pre- extraction heat treatment. Because this experiment was designed to detect contamination, one extraction negative control accompanied each sample extraction. PCR amplification, confirmation of amplification, and sequencing were performed as described above for the formalin trials.
166
3. RESULTS
3.1. Formalin trials
Specimen information, subsample material and weight, extraction method, and PCR results are summarized in Table 1. Extraction methods 2A, 2B, and 3A yielded mtDNA from four of the fifteen the specimens tested (Table 1 and Table 4). In total, eight of 78 extraction attempts were successful, five with method 2A, two with method 2B, and one with method 3A.
Method 2A was the most widely successfully, yielding sequenceable DNA from three (5242,
46711, CASIZ 31213) of the four unique specimens that produced amplifiable DNA. Method 2B was successful for two unique specimens and method 3A for one. For two of the specimens from which DNA was successfully obtained, only contaminating DNA was observed, matching a species other than that identified as the specimen by the museum (Table 4). Specimen 46711 was Chinook salmon (Oncorhynchus tshawytscha) however the DNA recovered was a close match to mottled sculpin (Cottus bairdii), a species distantly related to salmon. Specimen 5242 was also Chinook salmon, however the DNA recovered from this specimen was human in origin.
No amplifiable DNA was obtained from any of the negative controls.
3.2. Contamination trials
One lid sample produced sequenceable DNA (Table 3 and Table 4). This DNA, amplified from all three extraction treatments, was contaminating DNA. Whereas the sample vessel contained northern pike (Esox lucius), only rainbow trout (Oncorhynchus mykiss) mtDNA was observed. No amplifiable mtDNA was obtained from any of the other lid samples, ethanol samples, or negative controls.
167
4. DISCUSSION
The data presented in this study represents a summary of our efforts to extract DNA from a variety of formalin-preserved specimens. While no systematic experimental design was implemented to test methods against one another, these data to provide valuable insights, particularly regarding the potential for contamination in fixed museum specimens.
The majority of formalin trials produced negative results. Amplification was only possible for eight of the 78 extractions despite having employed varying PCR protocols (Table 1).
Extraction methods 2A, 2B, and 3A produced sequenceable DNA, with method 2A being the most widely successful. Methods 2A and 2B both utilize pre-extraction heat treatment at 95°C followed by extraction via the QIAamp DNA Mini Kit (Qiagen; Valencia, CA). The methods vary only in the reagents used for pre-extraction washing, TE for method 2A and EDTA for method
2B. Although method 2A was the most widely successful, effective application of the methods varied by specimen. Considering the four unique specimens that produced sequenceable DNA, for one (CASIZ 31213) method 2A was successful while method 3A was not. However, for another (TL-01), the opposite was achieved (3A was successful and 2A was not). In two specimens (CASIZ 31213 and 46711) both methods 2A and 2B were successful, however for a third (specimen 5242) 2A was successful where 2B was not. These results indicate that a relationship between specimen and extraction method is present. Such a relationship may be due to differences in fixation treatment between specimens. The specifics of fixation treatments vary widely in regards to the fixative chemistry, duration of fixation, and post- fixation ethanol transfer (Tang 2006). Such variability can impact the successful recovery of
DNA from such specimens. Factors that can influence success include: 1) the chemical
168
composition, temperature, and pH of the fixative, 2) length of fixation treatment, 3) size and permeability of specimen, and 4) overall duration of storage (Crisan and Mattson 1993; Raja et al. 2011; Schander and Kenneth 2003). Data on fixation treatment is rarely, if ever included with specimens in collections, making it difficult, if not impossible to parse the effect of fixation methods. Information on fixation treatment was not available for the specimen in the present study.
Perhaps the most interesting result of our attempts to extract DNA from fixed specimens was the amplification of non-target organismal DNA over target from two samples. Both specimens were Chinook salmon but mtDNA from one specimen closely matched mottled sculpin while the other matched to human chromosomal DNA (Table 4). Previous indications of human contamination in formalin-preserved specimens have been noted (Chase et al. 1998;
Schander and Kenneth 2003) and the primers used to amplify fish mtDNA are known to amplify human autosomal and mouse (Mus musculus) mitochondrial DNA (Grier et al. 2013; Kemp et al.
2014). Human contamination can certainly be introduced from prior handling of materials, from the researchers themselves, and/or also from reagents and lab consumables (Barta et al. 2013;
Handt et al. 1994; Kemp and Smith 2005; Pääbo 1989). However, the amplification of non- target, non-human DNA appears to potentially be an as-yet-unreported phenomenon for formalin-fixed specimens. Morphologically, mottled sculpin and Chinook salmon are highly dissimilar and, even in the case of juvenile specimens, quite unlikely to be confused. The contaminating DNA was independently extracted from two fin clips, which were provided in to us in separate PCR tubes. Subsequent to identification of mottled sculpin mtDNA, four additional independent PCRs were conducted from the second extract. All sequences obtained
169
matched again to mottled sculpin, suggesting that the predominant DNA molecules obtainable from the specimen were that of the non-target species. Therefore, it is unlikely that either specimen misidentification, DNA damage, or erroneous sequencing results led to the indication of non-target DNA. No samples of mottled sculpin, ancient or contemporary have been processed in our laboratory. Thus, we conclude that the contaminating DNA was introduced prior to receipt of the samples.
To further investigate the potential incidence of contamination in the museum environment, we collected samples of lid paper and ethanol from a variety of specimen jars in the Conner Museum at Washington State University (a different curation facility from that which housed the previously discussed specimens). Contamination was indicated for one of the vessels where DNA from rainbow trout was observed from a sample of lid paper taken from a jar containing northern pike. It is possible that DNA from rainbow trout came in contact with the lid paper during prior use, as collection vessels are occasionally reused in the collection. In the WSU collection, and likely in many other collections, the risk of DNA contamination has not been a consideration. Therefore, no active standard protocol to keep lids with jars or to always keep specimens with their original jars exists (Dr. K. Cassidy, WSU; personal communication).
Contaminating DNA was obtained from all three extraction variations, including one where the sample was rinsed in for 24 hours in TE prior to extraction. Potential sources of contaminating
DNA are numerous, with post-PCR products likely represent the highest risk as the PCR process creates billions of copies of DNA. However, when target DNA is highly degraded the risk of environmentally-based DNA (e.g., touch DNA) increases (Van Oorschot et al. 2010). Potential sources of contamination for wet collections include jars, jar lids and lid papers (Figure 2), reuse
170
of ethanol, tools used for moving samples between jars or subsampling them for study, identification slips placed with samples (Figure 2), and collection tubes. Further systematic testing of potential sources, under conditions designed to minimize additional introduction of contamination, could yield insight into the origin of such contamination. However, identification of source is not a replacement for practices to minimize risk. For example, if tools were identified as a source, this does not mean switching to "sterile" tools while continuing to reuse a common collection of jars and lids is advisable.
Acquiring DNA from samples is always subject to the risk of contamination. However, due to the majority-rules nature of PCR, most contaminating DNA molecules will be overcome by target DNA molecules and have no effect on the final outcome (Erlich 1989). However, in the case of highly degraded extracts (such as formalin-preserved samples), the risk of contamination increases (Pääbo et al. 2004; Schander and Kenneth 2003). In highly degraded or cross-linked samples, non-target DNA molecules may be more readily available than target DNA molecules. The indication that non-target DNA present in formalin-fixed samples may be preferentially amplified and sequenced over that of the target specimen is concerning for future studies of these sample sources. Whereas we were able to easily identify the contamination because it arose from distantly related organisms, should the contaminating
DNA have originated from a source of the same species, it would be far more difficult, if not impossible, to identify. Further, contamination introduced during museum curation or subsampling is not detectable with negative controls in the extraction or PCR procedures.
The results presented here represent a cautionary tale for the collection, fixation, and storage of specimens. At minimum, reuse of jars and lids is not advised and subsampling should
171
be conducted in a dedicated space with clean supplies and tools. We advise that curators critically evaluate practices specific to their collections and use common sense to identify and minimize potential risks. Investigators should interpret positive amplification with caution and critically evaluate and authenticate data collected from fixed specimens.
172
LITERATURE CITED
Barta, J. L., C. Monroe, and B. M. Kemp. 2013. Further evaluation of the efficacy of contamination removal from bone surfaces. Forensic Science International 231(1):340- 348.
Bonin, S., F. Petrera, B. Niccolini, and G. Stanta. 2003. PCR analysis in archival postmortem tissues. Molecular Pathology 56(3):184-186.
Bonin, S., F. Petrera, J. Rosai, and G. Stanta. 2005. DNA and RNA obtained from Bouin’s fixed tissues. Journal of Clinical Pathology 58(3):313-316.
Bramwell, N., and B. Burns. 1988. The effects of fixative type and fixation time on the quantity and quality of extractable DNA for hybridization studies on lymphoid tissue. Experimental Hematology 16(8):730-732.
Bunker, M. L., and J. Locker. 1989. Warthin’s tumor with malignant lymphoma: DNA analysis of paraffin-embedded tissue. American Journal of Clinical Pathology 91(3):341-344.
Campos, P. F., and T. M. Gilbert. 2012. DNA extraction from formalin-fixed material. Ancient DNA: Methods and Protocols:81-85.
Chan, Y. L., C. N. Anderson, and E. A. Hadly. 2006. Bayesian estimation of the timing and severity of a population bottleneck from ancient DNA. PLoS Genetics 2(4):e59.
Chang, Y.-T., and G. H. Loew. 1994. Reaction mechanisms of formaldehyde with endocyclic imino groups of nucleic acid bases. Journal of the American Chemical Society 116(8):3548-3555.
Chase, M. R., R. J. Etter, M. A. Rex, and J. M. Quattro. 1998. Extraction and amplification of mitochondrial DNA from formalin-fixed deep-sea mollusks. Biotechniques 24(2):243- 247.
Chatigny, M. 2000. The extraction of DNA from formalin-fixed, ethanol-preserved reptile and amphibian tissues. Herpetological Review 31(2):86-87.
Chaw, Y. F. M., L. E. Crane, P. Lange, and R. Shapiro. 1980. Isolation and identification of cross- links from formaldehyde-treated nucleic acids. Biochemistry 19(24):5525-5531.
Crisan, D., and J. C. Mattson. 1993. Retrospective DNA analysis using fixed tissue specimens. DNA and Cell Biology 12(5):455-464.
De Giorgi, C., M. F. Sialer, and F. Lamberti. 1994. Formalin-induced infidelity in PCR-amplified DNA fragments. Molecular and Cellular Probes 8(6):459-462.
173
Dubeau, L., L. A. Chandler, J. R. Gralow, P. W. Nichols, and P. A. Jones. 1986. Southern blot analysis of DNA extracted from formalin-fixed pathology specimens. Cancer Research 46(6):2964-2969.
Erlich, H. A. 1989. Polymerase chain reaction. Journal of Clinical Immunology 9(6):437-447.
French, D., and J. T. Edsall. 1945. The reactions of formaldehyde with amino acids and proteins. Advances in Protein Chemistry 2:277-335.
Gilbert, M. T. P., T. Haselkorn, M. Bunce, J. J. Sanchez, S. B. Lucas, L. D. Jewell, E. Van Marck, and M. Worobey. 2007. The isolation of nucleic acids from fixed, paraffin-embedded tissues– which methods are useful when? PloS One 2(6):e537.
Goelz, S. E., S. R. Hamilton, and B. Vogelstein. 1985. Purification of DNA from formaldehyde fixed and paraffin embedded human tissue. Biochemical and Biophysical Research Communications 130(1):118-126.
Grier, C., K. Flanigan, M. Winters, L. G. Jordan, S. Lukowski, and B. M. Kemp. 2013. Using ancient DNA identification and osteometric measures of archaeological Pacific salmon vertebrae for reconstructing salmon fisheries and site seasonality at Dionisio Point, British Columbia. Journal of Archaeological Science 40(1):544-555.
Hadly, E. A., U. Ramakrishnan, Y. L. Chan, M. Van Tuinen, K. O'Keefe, P. A. Spaeth, and C. J. Conroy. 2004. Genetic response to climatic change: insights from ancient DNA and phylochronology. PLoS Biology 2(10):e290.
Handt, O., M. Höss, M. Krings, and S. Pääbo. 1994. Ancient DNA: Methodological challenges. Experientia 50(6):524-529.
Hänni, C., T. Brousseau, V. Laudet, and D. Stehelin. 1995. Isopropanol precipitation removes PCR inhibitors from ancient bone extracts. Nucleic Acids Research 23(5):881.
Hopwood, D. 1967. Some aspects of fixation with glutaraldehyde. A biochemical and histochemical comparison of the effects of formaldehyde and glutaraldehyde fixation on various enzymes and glycogen, with a note on penetration of glutaraldehyde into liver. Journal of Anatomy 101(Pt 1):83.
Hykin, S. M., K. Bi, and J. A. McGuire. 2015. Fixing formalin: A method to recover genomic-scale DNA sequence data from formalin-fixed museum specimens using high-throughput sequencing. PloS One 10(10).
Iwamoto, E. M., J. M. Myers, and R. G. Gustafson. 2012. Resurrecting an extinct salmon evolutionarily significant unit: archived scales, historical DNA and implications for restoration. Molecular Ecology 21(7):1567-1582.
174
Jackson, V. 1978. Studies on histone organization in the nucleosome using formaldehyde as a reversible cross-linking agent. Cell 15(3):945-954.
Johnson, B. M., and B. M. Kemp. in review. Rescue PCR: Reagent-rich PCR improves amplification of degraded DNA extracts. Journal of Archaeological Science: Reports.
Johnson, D., S. Levanat, and A. E. Bale. 1995. Direct molecular analysis of archival tumor tissue for loss of heterozygosity. Biotechniques 19(2):190.
Jordan, L. G., C. A. Steele, and G. H. Thorgaard. 2010. Universal mtDNA primers for species identification of degraded bony fish samples. Molecular Ecology Resources 10(3):581- 581.
Karlsen, F., M. Kalantari, M. Chitemerere, B. Johansson, and B. Hagmar. 1994. Modifications of human and viral deoxyribonucleic acid by formaldehyde fixation. Laboratory Investigation; A Journal of Technical Methods and Pathology 71(4):604-611.
Kemp, B. M., R. S. Malhi, J. McDonough, D. A. Bolnick, J. A. Eshleman, O. Rickards, C. Martinez- Labarga, J. R. Johnson, J. G. Lorenz, E. J. Dixon, T. E. Fifield, T. H. Heaton, R. Worl, and D. G. Smith. 2007. Genetic analysis of early Holocene skeletal remains from Alaska and its implications for the settlement of the Americas. American Journal of Physical Anthropology 132:605-621.
Kemp, B. M., C. Monroe, K. G. Judd, E. Reams, and C. Grier. 2014. Evaluation of methods that subdue the effects of polymerase chain reaction inhibitors in the study of ancient and degraded DNA. Journal of Archaeological Science 42:373-380.
Kemp, B. M., and D. G. Smith. 2005. Use of bleach to eliminate contaminating DNA from the surface of bones and teeth. Forensic Science International 154(1):53-61.
Kemp, B. M., and D. G. Smith. 2010. Ancient DNA methodology: Thoughts from Brian M. Kemp and David Glenn Smith on “Mitochondrial DNA of protohistoric remains of an Arikara population from South Dakota”. Human Biology 82(2):227-238.
Kore, L., B. M. Johnson, C. Lindsey, and K. Onthank. in review. First record of 'Benthoctopus' leioderma (Cephalopoda: Enteroctopodidae) in shallow water. Marine Biodiversity Records.
Krafft, A., J. H. Lichy, T. P. Lipscomb, B. A. Klaunberg, S. Kennedy, and J. K. Taubenberger. 1995. Postmortem diagnosis of morbillivirus infection in bottlenose dolphins (Tursiops truncatus) in the Atlantic and Gulf of Mexico epizootics by polymerase chain reaction- based assay. Journal of Wildlife Diseases 31(3):410-415.
Libório, T. N., A. Etges, A. d. C. Neves, R. A. Mesquita, and F. D. Nunes. 2005. Evaluation of the genomic DNA extracted from formalin-fixed, paraffin-embedded oral samples archived
175
for the past 40-years. Jornal Brasileiro de Patologia e Medicina Laboratorial 41(6):405- 410.
Longy, M., B. Duboue, P. Soubeyran, and D. Moynet. 1997. Method for the purification of tissue DNA suitable for PCR after fixation with Bouin's fluid: Uses and limitations in microsatellite typing. Diagnostic Molecular Pathology 6(3):167-173.
Ma, T.-H., and M. M. Harris. 1988. Review of the genotoxicity of formaldehyde. Mutation Research/Reviews in Genetic Toxicology 196(1):37-59.
Nishiguchi, M. K., P. Doukakis, M. Egan, D. Kizirian, A. Phillips, L. Prendini, H. C. Rosenbaum, E. Torres, Y. Wyner, and R. DeSalle. 2002. DNA isolation procedures. Pages 249-287 in Techniques in Molecular Systematics and Evolution. Springer.
Orlando, L., D. Bonjean, H. Bocherens, A. Thenot, A. Argant, M. Otte, and C. Hänni. 2002. Ancient DNA and the population genetics of cave bears (Ursus spelaeus) through space and time. Molecular Biology and Evolution 19(11):1920-1933.
Orlando, L., A. Ginolhac, G. Zhang, D. Froese, A. Albrechtsen, M. Stiller, M. Schubert, E. Cappellini, B. Petersen, and I. Moltke. 2013. Recalibrating Equus evolution using the genome sequence of an early Middle Pleistocene horse. Nature 499(7456):74-78.
Pääbo, S. 1989. Ancient DNA: Extraction, characterization, molecular cloning, and enzymatic amplification. Proceedings of the National Academy of Sciences 86(6):1939-1943.
Pääbo, S., H. Poinar, D. Serre, V. Jaenicke-Després, J. Hebler, N. Rohland, M. Kuch, J. Krause, L. Vigilant, and M. Hofreiter. 2004. Genetic analyses from ancient DNA. Annual Review of Genetics 38:645-679.
Pedersen, M. W., A. Ruter, C. Schweger, H. Friebe, R. A. Staff, K. K. Kjeldsen, M. L. Mendoza, A. B. Beaudoin, C. Zutter, and N. K. Larsen. 2016. Postglacial viability and colonization in North America’s ice-free corridor. Nature 537(7618):45-49.
Raja, M., M. Muralidharan, and M. Arunachalam. 2011. Comparative analysis of DNA extracted from fish species preserved in formalin in two different periods. Turkish Journal of Biology 35(3):331-336.
Sato, Y., R. Sugie, B. Tsuchiya, T. Kameya, M. Natori, and K. Mukai. 2001. Comparison of the DNA extraction methods for polymerase chain reaction amplification from formalin- fixed and paraffin-embedded tissues. Diagnostic Molecular Pathology 10(4):265-271.
Schander, C., and H. M. Kenneth. 2003. DNA, PCR and formalinized animal tissue–a short review and protocols. Organisms Diversity & Evolution 3(3):195-205.
176
Shedlock, A. M., M. Haygood, T. W. Pietsch, and P. Bentzen. 1997. Enhanced DNA extraction and PCR amplification of mitochondrial genes from formalin-fixed museum specimens. Biotechniques 22(3):394-400.
Shi, S.-R., R. J. Cote, L. Wu, C. Liu, R. Datar, Y. Shi, D. Liu, H. Lim, and C. R. Taylor. 2002. DNA extraction from archival formalin-fixed, paraffin-embedded tissue sections based on the antigen retrieval principle: heating under the influence of pH. Journal of Histochemistry & Cytochemistry 50(8):1005-1011.
Shi, S.-R., R. Datar, C. Liu, L. Wu, Z. Zhang, R. J. Cote, and C. R. Taylor. 2004. DNA extraction from archival formalin-fixed, paraffin-embedded tissues: Heat-induced retrieval in alkaline solution. Histochemistry and Cell Biology 122(3):211-218.
Shiozawa, D., J. Kudo, R. Evans, S. Woodward, and R. Williams. 1992. DNA extraction from preserved trout tissues. The Great Basin Naturalist:29-34.
Smith, M. J., C. E. Pascal, Z. Grauvogel, C. Habicht, J. E. Seeb, and L. W. Seeb. 2011. Multiplex preamplification PCR and microsatellite validation enables accurate single nucleotide polymorphism genotyping of historical fish scales. Molecular Ecology Resources 11(s1):268-277.
Tang, E. P. 2006. Path to effective recovering of DNA from formalin-fixed biological samples in natural history collections: Workshop summary. National Academies Press.
Trembath-Reichert, E., A. Green-Saxena, and V. J. Orphan. 2013. Whole cell immunomagnetic enrichment of environmental microbial consortia using rRNA-targeted magneto-FISH. Microbial Metagenomics, Metatranscriptomics, and Metaproteomics 531:21-44.
Trifonov, E., Y. Lazurikin, and M. Frank-Kamenstsky. 1967. Formaldehyde-induced uncoiling of DNA. Molecular Biology 1:142-151.
Vachot, A., and M. Monnerot. 1996. Extraction, amplification and sequencing of DNA from formaldehyde-fixed specimens. Ancient Biomolecules 1(3):16. van Oorschot, R. A., K. N. Ballantyne, and R. J. Mitchell. 2010. Forensic trace DNA: A review. Investigative Genetics 1(1):1.
Williams, C., F. Pontén, C. Moberg, P. Söderkvist, M. Uhlén, J. Pontén, G. Sitbon, and J. Lundeberg. 1999. A high frequency of sequence alterations is due to formalin fixation of archival specimens. The American Journal of Pathology 155(5):1467-1471.
Wirgin, I., L. Maceda, J. Stabile, and C. Mesing. 1997. An evaluation of introgression of Atlantic coast striped bass mitochondrial DNA in a Gulf of Mexico population using formalin‐ preserved museum collections. Molecular Ecology 6(10):907-916.
177
Wong, S. Q., J. Li, A. Y. Tan, R. Vedururu, J.-M. B. Pang, H. Do, J. Ellul, K. Doig, A. Bell, and G. A. McArthur. 2014. Sequence artefacts in a prospective series of formalin-fixed tumours tested for mutations in hotspot regions by massively parallel sequencing. BMC Medical Genomics 7(1):1.
Wu, L., N. Patten, C. T. Yamashiro, and B. Chui. 2002. Extraction and amplification of DNA from formalin-fixed, paraffin-embedded tissues. Applied Immunohistochemistry & Molecular Morphology 10(3):269-274.
178
FIGURES AND TABLES
Figure 1. Figure follows on next page. Summary of methods applied to fixed specimens.
Methods varied in pre-extraction washing, lysis, and extraction procedure. Methods 1A, 1B, 2A,
2B are based on Gilbert et al. (2007), method 3A follows that described by Shiozawa et al.
(1992), method 3B combines the protocols described by Trembath-Reichert et al. (2013) and
Kemp et al. (2007), method 4 follows Nishiguchi et al. (2002).
179
180
Figure 2. Storage vessel containing fixed specimens (left) and empty jar ready for reuse (right).
Lid paper (right) on empty jar has indications of previous use evidenced by wrinkled surface and yellow discoloration from contact with liquid.
181
Table 1. Summary of samples, species, collection data, tissue type and weight, extraction method, and amplification by PCR
procedure for fixed specimens. Successful PCR amplification is indicated by "+" whereas “-" indicates failure of amplification, and
"n/a" indicates the PCR procedure was not applied. PCR Procedures are abbreviated as" standard - "St", 3X template - "3X", rescue
+25% - "R25), rescue +50% - "R50", and pre-amp - "PA". Table continues on consecutive pages.
Identification Collection Extraction PCR & Sequencing (source) Species location; year Material Weight (g) Method S 3X R25 R50 PA Sequence result 4964 (1) Oncorhynchus Rock Island Fin clip 0.004 2A - - - - - tshawytscha Dam; 1939 Fin clip 0.007 4 - - - - - 5089 (1) Oncorhynchus Rock Island Fin clip 0.004 1B - - - - - 182 tshawytscha Dam, WA; Fin clip 0.001 4 - - - - - 5242 (1) Oncorhynchus Rock Island Fin clip 0.004 2A + n/a n/a n/a n/a Non-target species tshawytscha Dam, WA; Fin clip 0.008 2B - - - - - 5244 (1) Oncorhynchus Rock Island Fin clip 0.002 2A - - - - - tshawytscha Dam, WA; Fin clip 0.003 2B - - - - - 5247 (1) Oncorhynchus Rock Island Fin clip 0.006 2A - - - - - tshawytscha Dam, WA; Fin clip 0.003 4 - - - - - 5249 (1) Oncorhynchus Rock Island Fin clip 0.002 2A - - - - - tshawytscha Dam, WA; Fin clip 0.001 2B - - - - - 5250 (1) Oncorhynchus Rock Island Fin clip 0.001 1A - - - - - tshawytscha Dam, WA; Fin clip 0.006 2B - - - - - 46711 (1) Oncorhynchus Below Rock Fin clip 0.007 2A + + n/a n/a n/a Non-target species tshawytscha Island Dam, Fin clip 0.006 2B + n/a n/a n/a n/a Non-target species WA; 1959 Fin clip 0.01 3B - - - - -
Identification Collection Extraction PCR & Sequencing (source) Species location; year Material Weight (g) Method S 3X R25 R50 PA Sequence result 96-123 (4) Oncorhynchus Leavenworth Fin clip 0.021 1A - - - - - tshawytscha Hatchery, Fin clip 0.002 1A - - - - - WA; 1954 Fin clip 0.02 1B - - - - - Fin clip 0.001 1B - - - - - Fin clip 0.056 2A - - - - - Fin clip 0.005 2A - - - - - Fin clip 0.007 2A - - - - - Fin clip 0.046 2A - - - - - Fin clip 0.04 2B - - - - - Fin clip 0.005 2B - - - - - Fin clip 0.005 2B - - - - - Fin clip 0.028 2B - - - - - Fin clip 0.016 3A - - - - -
183 Gut 0.021 3A - - - - -
96-124 (4) Oncorhynchus Leavenworth Fin clip 0.016 2A - - - - - tshawytscha Hatchery, Fin clip 0.059 2A - - - - - WA; 1954 Fin clip 0.008 2B - - - - - Fin clip 0.061 2B - - - - - Fin clip 0.019 3A - - - - - Gut 0.013 3A - - - - - 96-126 (4) Oncorhynchus Leavenworth Fin clip 0.007 1A - - - - - tshawytscha Hatchery, Fin clip 0.002 1A - - - - - WA; 1954 Fin clip 0.002 1B - - - - - Fin clip 0.001 1B - - - - - Fin clip 0.001 1A - - - - - Fin clip 0.001 1A - - - - - Fin clip 0.003 1B - - - - - Fin clip 0.001 1B - - - - - Fin clip 0.006 2A - - - - -
Identification Collection Extraction PCR & Sequencing (source) Species location; year Material Weight (g) Method S 3X R25 R50 PA Sequence result 96-126 cont. Oncorhynchus Leavenworth Fin clip 0.043 2A - - - - - tshawytscha Hatchery, Fin clip 0.008 2B - - - - - WA; 1954 Fin clip 0.068 2B - - - - - Fin clip 0.018 3A - - - - - Fin clip 0.015 3B - - - - - CASIZ 31213 (2) Benthoctopus Simpson Bay, Tissue 0.1 2A + n/a n/a n/a n/a Matched species leioderma Alaska; 1974 Tissue 0.027 2A + n/a n/a n/a n/a Matched species Tissue 0.025 2A + n/a n/a n/a n/a Matched species Tissue 0.016 2B + + n/a n/a n/a Matched species Tissue 0.036 3A - - - - - Tissue 0.017 3B - - - - - FR-001 Oncorhynchus Leavenworth Fin clip 0.001 1A - - - - - tshawytscha Hatchery, Fin clip 0.003 1A - - - - - 184 WA; 1954 Fin clip 0.001 1A - - - - - Fin clip 0.001 1A - - - - - Fin clip 0.008 2A - - - - - Fin clip 0.009 2B - - - - - Fin clip 0.018 3A - - - - - TL-01 (3) Oncorhynchus Three Mile Fin clip 0.011 2A - - - - - mykiss Creek, OR; Tissue 0.007 2A - - - - - 1995 Fin clip 0.063 3A - - - - - Flesh 0.104 3A - + - - + Matched species Fin clip 0.058 4 - - - - - Flesh 0.063 4 - - - - - Gill 0.08 4 - - - - - Vertebra 0.092 4 - - - - - Vertebra 0.084 4 - - - - -
Identification Collection Extraction PCR & Sequencing (source) Species location; year Material Weight (g) Method S 3X R25 R50 PA Sequence result TWWC-003 (4) Cyprinus carpio Unknown Vertebra 0.054 2A - - - - - Scales 0.015 2B - - - - - Flesh 0.156 2B - - - - - Vertebra 0.039 2B - - - - - Vertebra 0.053 3A - - - - - Source: (1) Burke Museum, University of Washington, (2) California Academy of Sciences, (3) Dr. G.H. Thorgaard, Washington State University, (4) Conner Museum, Washington State University
185
Table 2. Summary of extraction variables tested on single specimens.
Sample Variable tested Method(s) Result 4964 Application of method 2A, 4 No amplification 5089 Application of method 1B, 4 No amplification 5242 Application of method 2A, 2B Amplificiation using 2A 5244 Application of method 2A, 2B No amplification 5247 Application of method 2A, 4 No amplification 5249 Application of method 2A, 2B No amplification 5250 Application of method 1A, 2B No amplification 46711 Application of method 2A, 2B, 3B Identical results for amplification and sequencing using 2A and 2B 96-123 Application of method 1A, 1B, 2A, 2B, 3A No amplification 96-123 Amount of tissue 1A, 1B, 2A, 2B No amplification 186 96-123 Tissue type 3A No amplification
96-124 Application of method 2A, 2B, 3A No amplification 96-124 Amount of tissue 2A, 2B No amplification 96-124 Tissue type 3A No amplification 96-126 Application of method 1A, 1B, 2A, 2B, 3A, 3B No amplification 96-126 Amount of tissue 2A, 2B No amplification CASIZ 31213 Application of method 2A, 2B, 3A, 3B Identical results for amplification and sequencing using 2A and 2B CASIZ 31213 Amount of tissue 2A Identical results for amplification and sequencing FR-001 Application of method 1A, 2A, 2B, 3A No amplification TL-01 Tissue type 2A, 3A, 4 Only flesh amplified (3A), starting weight is also higher TL-01 Application of method 2A, 3A, 4 Amplification using method 3A TWWC-003 Tissue type 2B No amplification TWWC-003 Application of method 2A, 2B, 3A No amplification
Table 3. Specimen and processing data for test of potential contaminating DNA in museum storage vessels. Extraction and
amplification was attempted for three subsamples of lid paper of equal weight and one subsample of ethanol (EtOH). Amplification
is indicated by "+" for successful amplification or "-" when no amplification was indicated.
Total PCR amplification paper Lid paper EtOH Identification information Species of specimen Year weight (g) 1 2 3 Sequence result Walla Walla Collection #18 Esox lucius unk. 0.088 + + + - Non-target species 96-236, 96-241 Oncorhynchus kisutch 1954 0.031 - - - - ID 03 Apocope oscula unk. 0.087 - - - - 48-146 Nocomis biguttatus unk. 0.072 - - - -
187 97-932 Oncorhynchus clarkii 1956 0.039 - - - -
Walla Walla Collection #13 Oncorhynchus kisutch unk. 0.060 - - - - Walla Walla Collection #38 Oncorhynchus clarkii unk. 0.074 - - - - Walla Walla Collection #08 Prosopium williamsoni unk. 0.071 - - - - Walla Walla Collection Oncorhynchus keta unk. 0.015 - - - - 97-939 Oncorhynchus clarkii unk. 0.047 - - - - 97-1128 Oncorhynchus keta unk. 0.029 - - - - 97-1275 Oncorhynchus keta unk. 0.060 - - - - 97-1008 Oncorhynchus keta unk. 0.035 - - - - 97-466 Ptychocheilus oregonensis 1971 0.051 - - - - 97-457 Richardsonius balteatus 1971 0.024 - - - - 48-477 Richardsonius balteatus 1948 0.067 - - - - 97-635-645 Richardsonius balteatus 1974 0.045 - - - - 97-534 Ptychocheilus oregonensis 1971 0.013 - - - - 56-65 Amia calva 1957 0.037 - - - - 97-731 Ptychocheilus oregonensis 1974 0.045 - - - -
Table 4. Summary data for sequences obtained from amplifications described in the chapter.
Sequence and BLAST results Extraction tissue Query Sample Species type and method N Length Quality Genetic spp. DNA Type cover E value Ident Accession 46711 Oncorhynchus Fin clip (1) 2B 2 192 91.7 Cottus bairdii mtDNA 98 2.00E-89 99 KP013090 tshawytscha Fin clip (2) 2A 10 188 96.62 Cottus bairdii mtDNA 98 4E-91 100 KP013090 TL-01 Oncorhynchus Tissue 3A 4 191 96.5 Oncorhynchus mtDNA 98 8E-93 100 KP013084 mykiss mykiss 5242 Oncorhynchus Fin clip 2A 2 186 92.65 Homo sapiens Chromosomal 98 5E-85 98 AC241530 tshawytscha WWC-18 Esox lucius Lid paper 2A (1) 2 148 89.2 Oncorhynchus mtDNA 100 4.00E-70 100 FJ710971 mykiss
188 Lid paper 2A (2) 4 148 91.2 Oncorhynchus mtDNA 100 4.00E-70 100 FJ710972 mykiss
Lid paper 2A (3) 4 148 78.3 Oncorhynchus mtDNA 100 4E-70 100 FJ710973 mykiss CASIZ Benthoctopus Tissue 2A 2 172 95.4 Benthoctopus mtDNA 100 3.00E-55 99 FJ603552 31213 leioderma spp. Tissue 2A 2 172 98.3 Benthoctopus mtDNA 100 3.00E-55 99 FJ603552 spp. Tissue 2A 2 172 97.1 Benthoctopus mtDNA 100 3.00E-55 99 FJ603552 spp. Tissue 2B 2 172 92.2 Benthoctopus mtDNA 100 3.00E-55 99 FJ603552 spp.
CHAPTER FOUR
Mitochondrial survey of sockeye salmon reveals limited genetic diversity and recent within
species divergence
1. INTRODUCTION
The evolution of Pacific salmonids is tightly linked with the dynamic environment they occupy (Waples et al. 2008). Initial speciation from the most recent common ancestor shared with the Salvelinus and Salmo genera is believed to have occurred during geologic upheaval in the Miocene era (Montgomery 2000). Following this speciation, glacial action in the Pleistocene era isolated populations, facilitating additional within-species divergence and forming the base for much of the contemporary genetic structure (Bernatchez and Wilson 1998). This period, which began about 1.8 mya and ended about 12 kya, consisted of approximately 19 climate cycles. Each of these cycles consisted of a 60 to 90 kyr glaciated period and a 40 to 10 kyr unglaciated period (Pielou 2008). During glaciated periods, habitat was available in refugia, or unglaciated areas with suitable habitat for survival. During glacial retreats, populations at the margins of the refugia would have expanded into deglaciated areas following a variety of routes
(Hewitt 1996; Mcphail and Lindsey 1986).
For salmonids, a number of hypotheses regarding the location and number of
Pleistocene refugia have been proposed. Major refugia around the Pacific Rim may have included the area south of the ice (southern portions of modern day Washington State and areas further south), Alaska west of the peninsula, as well as Japan and the coast of Kamchatka in Russia (Barr and Solomina 2014; Hess and Tasa 2013) with a possible intermittent refugium
189
near Puget Sound (i.e. Salish Sea) and the Chehalis Basin as well as another near the Olympic
Peninsula (Waples et al. 2008). A large body of support exists for a southern, Columbia River refugium and a potential northern location based in Beringia (Utter et al. 1980). This pattern has been supported for rainbow trout (Oncorhynchus mykiss) (Mccusker et al. 2000), as well as
Chinook salmon (O. tshawytscha) (Beacham et al. 2006; Gharrett et al. 1987), coho salmon (O. kisutch) (Gharrett et al. 2001; Mcphail and Lindsey 1986; Smith et al. 2001), and sockeye salmon (O. nerka) (Bickham et al. 1995; Mcphail and Lindsey 1986; Mcphail and Lindsey 1970;
Varnavskaya et al. 1994a).
Successful postglacial dispersal from refugia was inextricably linked to life history requirements and flexibility in regards to those requirements. For example, species that could complete their entire lifecycle in freshwater may have been more successful holding in refugia while those with a reduced reliance on freshwater would be the first to colonize rivers formed after glacial retreat. Phylogeographic surveys of Oncorhynchus species have supported such a model. Surveys of rainbow trout, a species with facultative anadromy, indicate that genetic structure developed in fragmented, refugial lakes during glacial periods (Brunelli et al. 2010;
Mccusker et al. 2000). In contrast, Chinook salmon, which are anadromous with a flexible, but obligate lotic freshwater requirement, lack such allopatric signatures. Instead, this species likely survived in areas south of the glacier until late in the Pleistocene and then expanded into the north during an interglacial period. Colonizers were then isolated in additional, northern based refugia for the remainder of the Pleistocene (Gharrett et al. 1987; Martin et al. 2010).
Sockeye salmon display life histories intermediate to Chinook salmon and rainbow trout.
Three major forms, termed ecotypes, exist for this species: a sea/river type, a lake type, and
190
kokanee (Wood et al., 2008). The sea/river type spends the least amount of time in freshwater, using this habitat primarily for spawning activities while the lake type spends about half its life in freshwater. The kokanee form is not anadromous, completing its full life cycle in the lentic environment. It is notable that kokanee and lake-type forms of sockeye often co-exist in the same area while remaining genetically distinct (Wood et al., 2008). The recurrent evolution (RE) hypothesis proposed by Wood et al. (2008) proposes that these differences arose during the
Pleistocene. Under this hypothesis, an ancestral straying sea/river type of sockeye salmon existed in large, relatively undifferentiated refugial metapopulations during the Pleistocene.
This type was less dependent on the lake for rearing and could successfully spawn along glacial margins (Wood et al. 2008). The lake and kokanee forms then evolved through adaptive radiation from the sea/river type during interglacial periods. Such evolution happened repeatedly, as each subsequent glacial advance would “reset” much of the genetic structure back to the less-differentiated, metapopulation type structure typical of the sea/river ecotype
(Wood et al. 2008).
We sought to develop a dataset for sockeye salmon that could be compared to existing data for Chinook salmon and rainbow trout. Such a dataset provides the opportunity to contrast phylogeographic patterns in the light of these species’ respective life history differences. We examined genetic variation in mtDNA for 468 sockeye salmon from thirty localities ranging from the Lower Columbia Basin in North America to Hokkaido Island, Japan.
We compare this dataset to that developed for Chinook salmon (Martin et al. 2010), as well as rainbow trout (Brunelli et al. 2010). We hypothesized that the phylogenetic patterns of sockeye would be intermediate to that of Chinook salmon and rainbow trout, reflecting the different
191
life-histories of these species. Results from our data did not support this hypothesis, instead matching closely with the recurrent evolution hypothesis and indicating a more recent within species divergence than that indicated for the other two salmonid species.
2. METHODS
2.1. DNA extraction, amplification, and sequencing
Tissue samples (fin clips, spleen, and/or liver) were provided by various sources from thirty localities covering a majority of the species range (Table 1 & Figure 1). Genetic material was extracted and amplified following the methods described in Martin et al. (2010). Primer sequences were amplification forward: (5’-CCCGCCCCTGAAAGCCGAAT-3’), reverse: (5’-
TAGCAAGGCGTCGTGGGCTG-3’), and internal sequencing: (5’-TAGCAAGGCGTCGTGGGCTG-3’).
The final sequenced product encompassed 443 bp of the control region (D-Loop), 68 bp of the tRNA phenylalanine region (t-phe), and 81 bp of the 12S ribosomal RNA (12S). Sequences were aligned using Sequencher 3.1.1 (Gene Codes Corporation, Ann Arbor, MI). Haplotypes seen in three or fewer samples were confirmed by sequencing a second, independent PCR reaction of the template DNA.
2.2. Data analysis and comparisons in Oncorhynchus spp.
Estimates of genetic diversity and differentiation as well as demographic and phylogenetic reconstruction were developed for sockeye salmon. This data is compared to published surveys of Chinook salmon, (Martin et al. 2010) and rainbow trout (Brunelli et al. 2010) that utilized the identical portion of the mitochondrial genome. All comparative analyses generated using the
192
published data were performed under the same methods described for sockeye salmon.
Sample localities for all three data sets were analyzed independently and in regional subgroupings and larger geographic based collectives of localities from the northern, central, and southern portions of the species range (Table 1). Chinook salmon followed the groups provided by the original authors (Table 2) and rainbow trout were grouped as northern: localities 1- 13, central localities 14 – 34, and southern localities 35 – 55, numbers reference population numbers from Brunelli et al. (2010) (Table 3).
2.2.1 Genetic diversity and differentiation
Sockeye salmon
Haplotype (h) and nucleotide diversity (π) were estimated using DNAsp (Librado and
Rozas 2009) for each of sample localities, regional subgroups (indicated in Table 1), as well as for the larger geographical groupings. Population differentiation was evaluated using an exact test of population differentiation implemented in Arlequin (Excoffier and Lischer 2010). Similar to FST, this method tests for the non-random distributions of haplotypes in populations. For the exact test a contingency (RxC) table is developed using the haplotype frequencies for each population and then combined with Markov chain sampling to test the null hypothesis of panmixia (Raymond and Rousset 1995; Ryman et al. 2006). Differentiation was examined between individual sample localities as well as for the subgroups indicated in Table 1 using
10,000 Markov chain steps and 1,000 dememorization steps with significance determined at α
≤ 0.05. Population-level metrics were not calculated for locations with five or fewer sampled individuals.
193
Chinook salmon and rainbow trout
Haplotype diversity in Chinook salmon for each of the 11 sample localities was provided in Martin et al. (2010). Haplotype diversity for rainbow trout as well as nucleotide diversity, and exact tests of population differentiation for both Chinook salmon and rainbow trout were calculated for comparison.
2.2.2 Phylogenetic reconstruction and population demography
Sockeye salmon
Haplotype relationships for sockeye were estimated using Bayesian analysis implemented in MrBayes 3.2 (Ronquist and Huelsenbeck 2003). By default, this program ignores positions that include gaps (indels) so these were coded as binary characters and included in the analysis. Following Martin et al. (2010) the three gene regions were treated independently, and the GTR+I+Γ model was utilized per results from jModelTest (Posada and
Crandall 1998). The program was run for 10E6 generations and sampled every 100th generation.
Two independent, simultaneous runs were used and the standard deviation of the runs compared. Standard deviation at the end of 10E5 runs was 0.003 indicating that the sample trees were similar and sampling was occurring from a stationary posterior distribution. The first
25% of runs were discarded as burn-in and the sampling taken from the remaining runs. The phylogenetic tree produced in MrBayes was transformed to a rooted cladogram in FigTree v.1.4.0 (Institute of Evolutionary Biology; Edinburgh, Scotland).
Estimates of divergence timing between the major clades identified in the phylogeny were made using the coalescent model implemented in MDIV (Nielsen and Wakeley 2001).
194
Initially, several test runs were completed with varying priors as suggested in Nielsen and
Wakeley (2001). Pair combinations of 1, 2, 3, 5, 7, 10, and 15 for migration (M) and divergence timing (T) were tested for each clade. The priors that best generated a bell-shaped posterior and minimized the number of data points in the upper-tail of the distribution were used for the final runs. Using these priors, three independent runs were completed using a chain length of
5E6 with a 10% burn-in. The estimates of Θ, M, and T from each run were compared for similarity to ensure the number of chains was sufficient to result in convergence to a stationary distribution (i.e. convergence of the ergodic averages) (Nielsen and Wakeley 2001). Final estimates of divergence time were based on the equation Tdiv=TΘ/2µ following Martin et al.
(2010). In summary, T and Θ were estimated in MDIV and µ was estimated using µ=K/2T* where K equals the rate of divergence and T* equals the divergence time between two sequences. Assuming a generation time of 4 years and a divergence rate of 0.75-1% per million years (Shedlock et al. 1992; Thomas et al. 1986), lower and upper rates for sockeye were estimated at 6.645E-6 and 8.86E-6 substitutions per site, per generation, respectively.
Distributions of Θ were used to calculate 95% confidence intervals for each rate of divergence
(0.75% and 1%).
A haplotype network was constructed in program Network version 4.613 (Bandelt et al.
1999). Character weights were set to 20 for indels and 30 for transversions to account for the rarity of such events; transitions and other characters were left at default weight ten.
Ambiguous connections were resolved following the rules described in Crandall and Templeton
(1993) and Templeton and Sing (1993). The relationship between population structure and population history was tested using a nested clade analysis (NCA) (Templeton 2008; Templeton
195
et al. 1987; Templeton et al. 1995). The haplotype network was nested using the expanded procedure described in Crandall (1996). Geographical associations were tested using Geodis
(Posada et al. 2000). Significance was determined at p < 0.05 treewise-level using the Dunn-
Sidak correction for multiple tests, implemented in GeoDis. Inferences regarding demographic history were made using the most recent key. Nested-clade analysis has been criticized for its use of dichotomous-type key that can provide inferences without statistical significance
(Garrick et al. 2008; Knowles 2008; Knowles and Maddison 2002; Petit 2008). However, the method has been validated both empirically and theoretically and modifications to reduce the false-positive rate have been implemented (Templeton 2008; Templeton 2009). The updated methodology was used in this study and results are interpreted in tandem with other supporting data.
Further tests for population expansion were conducted using distributions of pairwise differences, i.e., mismatch distributions and estimates of the exponential growth rate (g). When populations have recently expanded we expect the distribution of sequence differences to be unimodal. In contrast, a bimodal distribution is expected in populations that have experienced a recent bottleneck or where balancing selection is present, and populations that have remained stable for long periods of time will have stochastic (i.e., multimodal distributions) (Rogers and
Harpending 1992). Mismatch distributions were generated using Arlequin (Excoffier et al. 2005;
Excoffier and Lischer 2010) with 1,000 bootstrap replications. Exponential growth rates were estimated using program Lamarc (Kuhner 2006) following the method described in Martin et al.
(2010).
196
Chinook salmon and rainbow trout
The Bayesian phylogeny, haplotype network, nested-clade analysis, and estimates of exponential growth rates (g) provided in Martin et al. (2010) were used for the comparisons in
Chinook salmon. A correction to the divergence timing was developed and confirmed with the corresponding author.
Brunelli et al. (2010) provided a haplotype network for rainbow trout, this network was recreated and color coded by geographical group to match the comparisons for sockeye salmon and Chinook salmon (Figure S1). Haplotypes included in the original network for Apache trout
(O. apache) and golden trout (O. mykiss aguabonita) were omitted from the nested network.
Bayesian phylogeny and estimates of divergence timing were developed directly for this study.
3. RESULTS
3.1. Genetic diversity and differentiation
Sockeye salmon
Twenty-two haplotypes were identified for sockeye. Haplotypes were deposited in
GenBank (Accession KU519295-KU519316). One haplotype (ONER01) was common to twenty- nine of the thirty sample locations (Table 4). Five of the twenty-two haplotypes were singletons, eight were unique to a single sample location, eleven were unique to a regional subgroup, and ten were only found in a single larger geographical grouping. Haplotype diversity and nucleotide diversity (x100) among populations ranged from 0 to 0.71 and 0 to 0.276, respectively (Table 4). When sample localities were grouped by geographic region, haplotype
197
diversity was estimated to be 0.38, 0.36, and 0.40 in the northern, central, and southern portions of the range (Table 5). Nucleotide diversity (X100) was estimated at 0.117, 0.084, and
0.141 for these same groups.
The exact test of population differentiation, which investigates the probability that haplotypes sampled from sub-populations are drawn randomly from a larger population, indicated a lack of significant differences in haplotype frequencies for most localities (Table 2 &
Table S6). Twenty of the twenty-eight sample locations compared had no indication of significant differentiation (supporting the hypothesis of panmixia between the two groups) for
50% or more of the comparisons. All sample localities in Alaska, Northern British Columbia, and
Asia (subgroups AES, ASA, SAN) fell into this category. Cultus and Ozette Lake were the most differentiated (96% and 100%, respectively).
Chinook salmon and rainbow trout
Martin et al. (2010) reported haplotype diversity for Chinook salmon ranging from 0.00 to 0.85, with an overall value of 0.82 (Table 2 and Table 5). Nucleotide diversity (x100) estimated from the data ranged from 0.031 to 0.479 with an overall value of 0.382. Grouped sample localities had respective estimates of haplotype diversity 0.41, 0.69, and 0.74 and nucleotide diversity (x100) 0.122, 0.342, and 0.136 in the northern, central, and southern groups (Table 5). Exact tests indicated 95% of the sample localities had significantly different haplotype frequencies (Table 5 & Table S7).
Using the rainbow trout data set (Brunelli et al. 2010) we estimated haplotype and nucleotide (x100) diversity to range from 0 to 0.87 and 0 to 0.313, respectively with values of
198
0.77 and 0.319 overall (Table 3 and Table 5). When sample localities were grouped, diversity was estimated at 0.56, 0.66, and 0.86 (h) and 0.273, 0.230, and 0.198 (π x100) for the northern, central, and southern groups respectively (Table 5). Exact tests indicated 61% of the population comparisons had significantly different haplotype frequencies (Tables 5 and S8).
3.2. Phylogenetic reconstruction and population demography
Sockeye salmon
Bayesian analysis indicated at least weak support for five clades (Figure 1). One highly- supported clade (ONER04 and 15) was found only in upper-Columbia subgroup. This southern clade was the most derived in the Bayesian phylogeny while the least derived types (ONER17,
18, 19, 20, 22) were only detected in samples from Alaska and further westward. Divergence timing for the largest, monophyletic clade was estimated at 122 – 163 kya (Figure 4).
Confidence intervals (95%) based on the distribution of Θ were 112 – 132 kya for the 0.75% mutation rate and 149 – 176 kya for the 1% mutation rate.
The haplotype network generated has a modified star-shaped pattern where one haplotype is common throughout many populations and several less frequent haplotypes are one or two mutations removed from this type (Figure 2). This pattern is indicative of recent and rapid expansion (Ferreri et al. 2011). The nested clade analysis resulted in six one-step clades and two two-step clades. The null hypothesis of no geographical association between haplotypes and sample locations was rejected for the full cladogram as well as three of the nested clades. Clade 2-1 indicated contiguous range expansion for these lineages with
199
haplotypes in clade 1-1 exhibiting signatures of isolation-by-distance and restricted gene flow from those in clade 1-2.
Mismatch distributions for the major clades identified indicate patterns of demographic expansion (Figure 3). A model of population expansion (null model) could not be rejected for the two lower-level clade levels (1-1, p=0.80 and 1-2, p=0.47), for the upper-level clade (2-1, p=0.64) or for all samples combined (p=0.62). Estimates of g for each of the significantly associated clades were 1-1: 3042 ± 246, 1-2: 10,876 ± 1103, and 2-1: 5152 ± 436.
Chinook salmon and rainbow trout
Bayesian phylogeny, haplotype network, and nested-clade-analysis for the Chinook dataset were completed by Martin et al (2010). Divergence for the two major clades was estimated in Martin et al. (2010) at 97 – 260 kya. However, a mathematical error was incorporated when calculating the upper mutation rate (1%); the corrected divergence estimate is 195 – 260 kya (95% confidence intervals: 179 – 210 kya and 239 – 280 kya). The Bayesian phylogeny indicated four clades with at least weak support (Figure 4). The haplotype network follows a linear pattern with six one-step clades nested in three two-step clades.
Bayesian analysis of the rainbow trout haplotypes revealed nine clades with at-least weak support (Figure 4). The most derived haplotypes were sampled in the southern portion of the range but both derived and ancestral types were sampled in all geographic portions. Divergence timing for the largest clade was estimated at 225 – 300 kya (95% confidence interval 208 – 243 kya and 277 – 323 kya). The haplotype network for this species was the most complex, including two star-type clusters with many other connected types. Nested-clade analysis of the
200
network revealed six clades significantly associated with geography (Figure S1). A single one- step clade (1-8) was attributed to allopatric fragmentation. Three two-step clades were significant, the largest indicated to have restricted gene-flow and isolation-by-distance with the others both indicated to be a product of contiguous range expansion.
4. DISCUSSION
4.1. Genetic structure and population differentiation
Sockeye salmon populations in this study tended to share a common haplotype (ONER01).
This type was sampled in all localities (omitting the Kamchatka River, which only had a single sample from this locality) (Table 4). This single haplotype accounted for at least 60% of the fish in each of the sub-groupings as well as 72% of the fish sampled overall. In turn, sockeye salmon haplotype diversity tended to be low, with 22 of 28 localities indicating values less than 0.5. The subgroup with the highest genetic diversity was Asia, where two of the three localities had diversity greater than 0.6. Fish from the third locality, Abira River, were the product of artificial stocking in the 1900s when non-anadromous sockeye from Lake Shikotsu, Hokkaido were used to develop the anadromous population in the river today. The sockeye originally in Lake
Shikotsu were also initially stocked from another lake, Lake Akan, Hokkaido and then eleven additional transplants were made between 1925 and 1940 from Lake Urumobetsu, Hurup
(Kurile Islands) (Winans et al. 1996). Such events may have altered the genetic signal for this population. The larger geographical groupings (northern, central, and southern) contained similar ranges of values and each was very similar to the overall value (0.38) for all sample localities considered together (Table 5).
201
More variation was indicated in measures of nucleotide diversity. As with haplotype diversity, the Asian subgroup was notably higher than the other subgroups. The upper-
Columbia samples also indicated high nucleotide diversity relative to other sub-groups. For all indices considered, sockeye salmon had less diversity indicated than both Chinook salmon and rainbow trout (Table 5). This result confirms early comparative studies for Oncorhynchus spp. which found sockeye and coho salmon (O. kisutch) to be the least genetically diverse of the genus (Allendorf and Utter 1979; Utter et al. 1973; Varnavskaya et al. 1994a; Wood et al. 1994).
The limited genetic diversity may be related to the unique life-history of the species. Sockeye salmon home more reliably than other Oncorhynchus (Hanamura 1966; Quinn et al. 1987;
Quinn 1985). This increased homing fidelity lends itself toward inbreeding and reproductive isolation, reducing genetic variation within sockeye salmon populations (Altukhov and
Salmenkova 1991; Quinn et al. 1987).
Homing-based reproductive isolation in sockeye salmon lends itself to the expectation that populations will be genetically differentiated from each other due to inbreeding and genetic drift. However, only 44% of the exact comparisons indicated statistically significant differences from that expected under panmixia. Four of the eight subgroups indicated greater differentiation between localities in the subgroup than from those in other subgroupings (Table
S6). Despite a generally mixed genetic signal, larger geographical groupings are identifiable for the species. Multiple studies have supported differentiation for sockeye salmon populations in
Russia, western Alaska, southeastern Alaska, northern British Columbia, southern British
Columbia, and Washington within the species (Utter et al. 1980; Varnavskaya et al. 1994b;
Wood et al. 1994). Indeed, unique haplotypes were sampled from localities in five of these
202
groups: Russia (locations ASA-E, ASA-G, and ASA-V), western Alaska (AES-T), southeastern
Alaska (SAN-T), southern British Columbia (SSR-N), and Washington (CBU-W) (Table 4).
The upper-Columbia localities were the least monomorphic of all the regional groups and, along with the Asian localities, had the highest estimates of haplotype and nucleotide diversity.
This result agrees with studies of Chinook salmon, which indicated that the Columbia River
Basin had higher measures of genetic diversity than other areas (Martin et al. 2010; Waples et al. 2004). However it contrasts with a microsatellite-based survey of sockeye salmon, which indicated that populations in the Columbia River were less diverse than those in other regions
(Beacham et al. 2006). Such a discrepancy is not unexpected; dischordance between mtDNA and nuclear DNA has been indicated for many species (Bensch et al. 2006; Toews and Brelsford
2012). Due to its clonal, maternal inheritance, mitochondrial DNA is useful for upper-level taxonomic, phylogenetic, and biogeographic studies while nuclear DNA may better resolve population-level differences and contemporary genetic connections. The differences here highlight the importance of using multiple markers to study the evolutionary and contemporary histories of a species.
4.2. Implications for refugia
Previous phylogeographic studies strongly support two genetic sources, one from refugia in Beringia (north) and the other from refugium in Cascadia (south) (Beacham et al. 2006;
Mcphail and Lindsey 1986; Mcphail and Lindsey 1970; Taylor et al. 1996; Wood et al. 2008).
Support for a northern and southern refugia was also present in our data. Two major clades (as evidenced by the highest Bayesian support values and corroboration in the nested haplotype
203
network) were identified in our study: (1) a northern, ancestral clade, (2) a more widely- distributed clade with many haplotypes. The northern clade, centered around haplotype
ONER19, is prevalent in the Asian samples and was also sampled further south into northern
British Columbia. This clade may represent members of a northern based refugium that colonized southward. The largest clade, centered around haplotype ONER01 was sampled through the species range (Figure 1 and Figure 2). Two haplotypes in the clade (ONER01 and
ONER21) were sampled in the southern, central, and northern portions of the range. However, the majority (11 of the remaining 12 haplotypes) were exclusive to the southern and central localities indicating that this clade may be the product of a southern-based refugium that colonized northward. The existence of a third, coastal refugium in British Columbia where lowered sea levels during glaciation exposed habitat in the Vancouver Island and Queen
Charlotte Island areas has also been suggested (Beacham et al. 2006; Bickham et al. 1995). Our data set lacked sampling in the Queen Charlotte Islands that could be used to implicitly test the third refugium hypothesis so we could not evaluate these claims in our study. However, evidence for such a refugium has been provided for multiple species (Byun et al. 1999; Byun et al. 1997; Conroy and Cook 2000; Mccusker et al. 2000; Ogilvie et al. 1989; Talbot and Shields
1996).
Comparisons of the Bayesian phylogenies between the three species were particularly informative. For all three species, two large, monophyletic clades are present. In rainbow trout, haplotypes from each of these clades were sampled throughout the range and no strong northern or southern associations are present. Comparisons of this phylogeny to the previous surveys of rainbow trout indicate that the most derived haplotypes were sampled exclusively in
204
coastally proximate populations of rainbow trout (Brunelli et al. 2010; Mccusker et al. 2000). In contrast, strong geographic distinction was present for the Chinook salmon phylogeny.
Southern types are the most ancestral, transitioning to the most derived types in the northern portions of the range and regional haplotype sorting is present (Figure 4).
Divergence between the monophyletic ancestral and derived clades for both rainbow trout and Chinook salmon correlate with the Yarmouth interglacial period occurred approximately 200 – 300 kya. Thus, although these two species followed different geographical patterns of recolonization, major migration events for both species appear to have occurred during the same interglacial period in the late Pleistocene. The south-to-north pattern indicated for Chinook salmon directly opposes that in sockeye salmon, where the ancestral clade is comprised of northern based haplotypes and the most derived clade was sampled exclusively in southern localities. Divergence timing for these clades is more recent than that estimated for
Chinook salmon and rainbow trout (122 and 163 kya) and correlates with the Sangamon interglacial period, which occurred approximately 110 – 130 kya.). Estimates of divergence timing rely on application of the molecular clock, a subject of some debate [reviewed in
Bromham and Penny (2003), Kumar (2005)]. Major criticism of the clock hypothesis include variation of evolutionary rates among groups of organisms (Thomas et al. 2006), differences in rates for the sequences of some proteins (Donoghue and Smith 2003), as well as the need for accurate calibration (Pulquerio and Nichols 2007). The clock estimate utilized in this study was developed specific for the control region (DLoop) region of mtDNA for Oncorhynchus spp. and calibrated with both molecular and fossil evidence (Shedlock et al. 1992). Therefore, we propose that although the clock may not correspond exactly to the prediction of neutral theory,
205
it can provide a reasonable comparative estimates of within-species divergence times for those species considered here.
The comparative pattern indicated for sockeye salmon is supportive of the poorly differentiated, metapopulation structure predicted by the RE hypothesis. Genetic diversity is low and, despite evidence for stronger homing in the species, most populations cannot be statistically differentiated from those found in other geographical areas. Phylogeographic analysis supports northern and southern separation, but the distinction is not as strong as that indicated for Chinook salmon. It is likely that sockeye salmon recolonized new areas during the
Yarmouth interglacial along with Chinook salmon and rainbow trout. However, divergence timing maps to the last interglacial period, fitting with a “resetting” of the mitochondrial genetic signal late in the Pleistocene. After the last glacial maximum, glaciers receded from the southern edges first, and sockeye salmon populations in this region likely had more opportunity for subsequent divergence from the ancestral form. Indeed, the highly supported and most derived haplotypes in this study were identified in samples taken from the Okanogan and
Wenatchee systems where lake-type life history forms are common (Chapman et al. 1995;
Gustafson and Winans 1999; Wood et al. 2008).
4.3. Conclusions
It is likely that several species of Oncorhynchus experienced cycles of colonization and extirpation during the Pleistocene. For sockeye salmon, the unique dependence on freshwater lentic environments may have limited the success of recolonizations. Relatively low genetic diversity, limited population structuring, and signatures of population expansion all indicate
206
that sockeye experienced a recent within-species divergence. Dating of major clade divergence in comparison with Chinook salmon and rainbow trout indicates that sockeye divergence occurred after that of the other two species, with most of the contemporary mtDNA variation developing subsequent to the last inter-glacial period. Furthermore, our study indicated the upper-Columbia River contains derived and genetically diverse populations of sockeye salmon, a result that opposes that previously reported for this species. This location may be an important conservation target for this species in addition those identified in other studies using different genetic markers.
207
LITERATURE CITED
Allendorf, F. W., and F. M. Utter. 1979. Population Genetics. Pages 407-454 in W. S. Hoar, D. J. Randall, and J. R. Brett, editors. Fish physiology. Acedemic Press, New York, NY.
Altukhov, Y. P., and E. Salmenkova. 1991. The genetic structure of salmon populations. Aquaculture 98(1-3):11-40.
Bandelt, H.-J., P. Forster, and A. Röhl. 1999. Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16(1):37-48.
Barr, I. D., and O. Solomina. 2014. Pleistocene and Holocene glacier fluctuations upon the Kamchatka Peninsula. Global and Planetary Change 113:110-120.
Beacham, T. D., B. McIntosh, C. MacConnachie, K. M. Miller, R. E. Withler, and N. Varnavskaya. 2006. Pacific rim population structure of sockeye salmon as determined from microsatellite analysis. Transactions of the American Fisheries Society 135(1):174-187.
Bensch, S., D. E. Irwin, J. H. Irwin, L. Kvist, and S. Åkesson. 2006. Conflicting patterns of mitochondrial and nuclear DNA diversity in Phylloscopus warblers. Molecular Ecology 15(1):161-171.
Bernatchez, L., and C. C. Wilson. 1998. Comparative phylogeography of Nearctic and Palearctic fishes. Molecular Ecology 7(4):431-452.
Bickham, J., C. Wood, and J. Patton. 1995. Biogeographic implications of cytochrome b sequences and allozymes in sockeye (Oncorhynchus nerka). Journal of Heredity 86(2):140-144.
Bromham, L., and D. Penny. 2003. The modern molecular clock. Nature Reviews Genetics 4(3):216-224.
Brunelli, J. P., C. A. Steele, and G. H. Thorgaard. 2010. Deep divergence and apparent sex-biased dispersal revealed by a Y-linked marker in rainbow trout. Molecular Phylogenetics and Evolution 56(3):983-990.
Byun, A., B. Koop, and T. Reimchen. 1999. Coastal refugia and postglacial recolonization routes: a reply to Demboski, Stone, and Cook. Evolution 53(6):2013-2015.
Byun, S., B. Koop, and T. Reimchen. 1997. North American black bear mtDNA phylogeography: implications for morphology and the Haida Gwaii glacial refugium controversy. Evolution:1647-1653.
Chapman, D., C. Peven, A. Giorgi, T. Hillman, F. Utter, M. Hill, J. Stevenson, and M. Miller. 1995. Status of sockeye salmon in the mid-Columbia region. Report prepared by BioAnalysts for Chelan, Douglass, and Grant County Public Utility Districts.
208
Conroy, C. J., and J. A. Cook. 2000. Phylogeography of a post‐glacial colonizer: Microtus longicaudus (Rodentia: Muridae). Molecular Ecology 9(2):165-175.
Crandall, K. A. 1996. Multiple interspecies transmissions of human and simian T-cell leukemia/lymphoma virus type I sequences. Molecular Biology and Evolution 13(1):115- 131.
Crandall, K. A., and A. R. Templeton. 1993. Empirical tests of some predictions from coalescent theory with applications to intraspecific phylogeny reconstruction. Genetics 134(3):959- 969.
Donoghue, P. C., and M. P. Smith. 2003. Telling the evolutionary time: molecular clocks and the fossil record. CRC Press.
Excoffier, L., G. Laval, and S. Schneider. 2005. Arlequin (version 3.0): an integrated software package for population genetics data analysis. Evolutionary bioinformatics online 1:47.
Excoffier, L., and H. E. Lischer. 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10(3):564-567.
Ferreri, M., W. Qu, and B. Han. 2011. Phylogenetic networks: a tool to display character conflict and demographic history. African Journal of Biotechnology 10(60):12799-12803.
Garrick, R., R. Dyer, L. Beheregaray, and P. Sunnucks. 2008. Babies and bathwater: a comment on the premature obituary for nested clade phylogeographical analysis. Molecular Ecology 17(6):1401-1403.
Gharrett, A., S. Shirley, and G. Tromble. 1987. Genetic relationships among populations of Alaskan chinook salmon (Oncorhynchus tshawytscha). Canadian Journal of Fisheries and Aquatic Sciences 44(4):765-774.
Gharrett, A. J., A. K. Gray, and V. Brykov. 2001. Phylogeographic analysis of mitochondrial DNA variation in Alaskan coho salmon (Oncorhynchus kisutch). Fishery Bulletin 99(4).
Gustafson, R. G., and G. A. Winans. 1999. Distribution and population genetic structure of river‐ and sea‐type sockeye salmon in western North America. Ecology of Freshwater Fish 8(3):181-193.
Hanamura, N. 1966. Salmon of the North Pacific Ocean--Part III. A review of the life history of North Pacific salmon: Sockeye salmon in the far east. Int. North Pac. Fish. Comm. Bull 18:1-27.
Hess, D., and D. G. Tasa. 2013. McKnight's physical geography: a landscape appreciation. Pearson Higher Ed.
209
Hewitt, G. M. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biological journal of the Linnean Society 58(3):247-276.
Knowles, L. L. 2008. Why does a method that fails continue to be used? Evolution 62(11):2713- 2717.
Knowles, L. L., and W. P. Maddison. 2002. Statistical phylogeography. Molecular Ecology 11(12):2623-2635.
Kuhner, M. K. 2006. LAMARC 2.0: maximum likelihood and Bayesian estimation of population parameters. Bioinformatics 22(6):768-770.
Kumar, S. 2005. Molecular clocks: four decades of evolution. Nature Reviews Genetics 6(8):654- 662.
Librado, P., and J. Rozas. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25(11):1451-1452.
Martin, K. E., C. A. Steele, J. P. Brunelli, and G. H. Thorgaard. 2010. Mitochondrial variation and biogeographic history of Chinook salmon. Transactions of the American Fisheries Society 139(3):792-802.
McCusker, M. R., E. Parkinson, and E. B. Taylor. 2000. Mitochondrial DNA variation in rainbow trout (Oncorhynchus mykiss) across its native range: testing biogeographical hypotheses and their relevance to conservation. Molecular Ecology 9(12):2089-2108.
McPhail, J., and C. Lindsey. 1986. Zoogeography of the freshwater fishes of Cascadia (the Columbia system and rivers north to the Stikine). The zoogeography of North American freshwater fishes:615-637.
McPhail, J. D., and C. C. Lindsey. 1970. Freshwater fishes of northwestern Canada and Alaska. Fisheries Research Board of Canada Bulletin 173, Ottawa.
Montgomery, D. R. 2000. Coevolution of the Pacific salmon and Pacific Rim topography. Geology 28(12):1107-1110.
Nielsen, R., and J. Wakeley. 2001. Distinguishing migration from isolation: a Markov chain Monte Carlo approach. Genetics 158(2):885-896.
Ogilvie, R., G. Scudder, and N. Gessler. 1989. Disjunct vascular flora of northwestern Vancouver Island in relation to Queen Charlotte Islands’ endemism and Pacific coast refugia. The outer shores:127-130.
Petit, R. J. 2008. The coup de grâce for the nested clade phylogeographic analysis? Molecular Ecology 17(2):516-518.
210
Pielou, E. C. 2008. After the ice age: the return of life to glaciated North America. University of Chicago Press.
Posada, D., K. Crandall, and A. Templeton. 2000. GeoDis: a program for the cladistic nested analysis of the geographical distribution of genetic haplotypes. Molecular Ecology 9(4):487-488.
Posada, D., and K. A. Crandall. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14(9):817-818.
Pulquerio, M. J., and R. A. Nichols. 2007. Dates from the molecular clock: how wrong can we be? Trends in Ecology & Evolution 22(4):180-184.
Quinn, T., C. Wood, L. Margolis, B. Riddell, and K. Hyatt. 1987. Homing in wild sockeye salmon (Oncorhynchus nerka) populations as inferred from differences in parasite prevalance and allozyme allele frequencies. Canadian Journal of Fisheries and Aquatic Sciences 44(11):1963-1971.
Quinn, T. P. 1985. Homing and the evolution of sockeye salmon.
Raymond, M., and F. Rousset. 1995. An exact test for population differentiation. Evolution:1280-1283.
Rogers, A. R., and H. Harpending. 1992. Population growth makes waves in the distribution of pairwise genetic differences. Molecular biology and evolution 9(3):552-569.
Ronquist, F., and J. P. Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19(12):1572-1574.
Ryman, N., S. Palm, C. André, G. R. Carvalho, T. G. Dahlgren, P. E. Jorde, L. Laikre, L. C. Larsson, A. Palmé, and D. E. Ruzzante. 2006. Power for detecting genetic divergence: differences between statistical methods and marker loci. Molecular Ecology 15(8):2031-2045.
Shedlock, A. M., J. D. Parker, D. A. Crispin, T. W. Pietsch, and G. C. Burmer. 1992. Evolution of the salmonid mitochondrial control region. Molecular phylogenetics and evolution 1(3):179-192.
Smith, C. T., R. J. Nelson, C. C. Wood, and B. F. Koop. 2001. Glacial biogeography of North American coho salmon (Oncorhynchus kisutch). Molecular Ecology 10(12):2775-2785.
Talbot, S. L., and G. F. Shields. 1996. A phylogeny of the bears (Ursidae) inferred from complete sequences of three mitochondrial genes. Molecular phylogenetics and evolution 5(3):567-575.
211
Taylor, E. B., C. J. Foote, and C. Wood. 1996. Molecular genetic evidence for parallel life-history evolution within a Pacific salmon (sockeye salmon and kokanee, Oncorhynchus nerka). Evolution:401-416.
Templeton, A. R. 2008. Nested clade analysis: an extensively validated method for strong phylogeographic inference. Molecular Ecology 17(8):1877-1880.
Templeton, A. R. 2009. Statistical hypothesis testing in intraspecific phylogeography: nested clade phylogeographical analysis vs. approximate Bayesian computation. Molecular Ecology 18(2):319-331.
Templeton, A. R., E. Boerwinkle, and C. F. Sing. 1987. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping. I. Basic theory and an analysis of alcohol dehydrogenase activity in Drosophila. Genetics 117(2):343-351.
Templeton, A. R., E. Routman, and C. A. Phillips. 1995. Separating population structure from population history: a cladistic analysis of the geographical distribution of mitochondrial DNA haplotypes in the tiger salamander, Ambystoma tigrinum. Genetics 140(2):767-782.
Templeton, A. R., and C. F. Sing. 1993. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping. IV. Nested analyses with cladogram uncertainty and recombination. Genetics 134(2):659-669.
Thomas, J. A., J. J. Welch, M. Woolfit, and L. Bromham. 2006. There is no universal molecular clock for invertebrates, but rate variation does not scale with body size. Proceedings of the National Academy of Sciences 103(19):7366-7371.
Thomas, W. K., R. E. Withler, and A. T. Beckenbach. 1986. Mitochondrial DNA analysis of Pacific salmonid evolution. Canadian Journal of Zoology 64(5):1058-1064.
Toews, D. P., and A. Brelsford. 2012. The biogeography of mitochondrial and nuclear discordance in animals. Molecular Ecology 21(16):3907-3930.
Utter, F. M., F. W. Allendorf, and H. O. Hodgins. 1973. Genetic variability and relationships in Pacific salmon and related trout based on protein variations. Systematic Biology 22(3):257-270.
Utter, F. M., D. Campton, S. Grant, G. Milner, J. Seeb, and L. Wishard. 1980. Population structures of indigenous salmonid species of the Pacific Northwest. Salmonid ecosystems of the North Pacific 285:304.
Varnavskaya, N. V., C. C. Wood, and R. J. Everett. 1994a. Genetic variation in sockeye salmon (Oncorhynchus nerka) populations of Asia and North America. Canadian Journal of Fisheries and Aquatic Sciences 51(S1):132-146.
212
Varnavskaya, N. V., C. C. Wood, R. J. Everett, R. L. Wilmot, V. S. Varnavsky, V. V. Midanaya, and T. Quinn. 1994b. Genetic differentiation of subpopulations of sockeye salmon (Oncorhynchus nerka) within lakes of Alaska, British Columbia, and Kamchatka, Russia. Canadian Journal of Fisheries and Aquatic Sciences 51(S1):147-157.
Waples, R. S., G. R. Pess, and T. Beechie. 2008. Evolutionary history of Pacific salmon in dynamic environments. Evolutionary Applications 1(2):189-206.
Waples, R. S., D. J. Teel, J. M. Myers, and A. M. Marshall. 2004. Life-history divergence in Chinook salmon: Historic contingency and parallel evolution. Evolution 58(2):386-403.
Winans, G., P. Aebersold, and R. Waples. 1996. Allozyme variability in selected populations of Oncorhynchus nerka with special consideration of Redfish Lake, Idaho. Transactions of the American Fisheries Society 125:645-653.
Wood, C. C., J. W. Bickham, R. John Nelson, C. J. Foote, and J. C. Patton. 2008. Recurrent evolution of life history ecotypes in sockeye salmon: implications for conservation and future evolution. Evolutionary Applications 1(2):207-221.
Wood, C. C., B. E. Riddell, D. T. Rutherford, and R. E. Withler. 1994. Biochemical genetic survey of sockeye salmon (Oncorhynchus nerka) in Canada. Canadian Journal of Fisheries and Aquatic Sciences 51(S1):114-131.
213
FIGURES AND TABLES
Figure 1. Bayesian analysis and distribution of haplotypes identified for sockeye salmon. Corresponding sample localities are indicated in Table 1 and Table 4.
.
214
Figure 2. Nested haplotype network for sockeye salmon. Circles represent haplotypes and are proportional to frequency, each line represents a single mutation, dashed lines represent ambiguous connections, small black circles indicate likely but unsampled types. Network is colored by geographic collective: north indicated with blue, central with yellow, and south with red.
215
Figure 3. Distribution of pairwise differences (mismatch distributions) for significantly associated clades identified for sockeye salmon. Grey bars indicate the observed differences and black lines indicate the expected values under a model of population expansion.
216
Figure 4. Comparative cladograms for sockeye salmon, Chinook salmon, and rainbow trout showing geographic-based trends in clade distribution as well as estimates of divergence timing for the major clade splits. Haplotypes are color coded according to the portion of the species range where they were sampled: north indicated with blue, central with yellow, and south with red.
217
Figure 5. Nested haplotype network for rainbow trout with inferences indicated for significantly associated clades. Circles represent haplotypes and are proportional to frequency, each line represents a single mutation, dashed lines represent ambiguous connections, small black circles indicate likely but unsampled types. Network was developed to match Brunelli et al. (2010) and re- colored by geographic collective: north indicated with blue, central with yellow, and south with red.
218
Table 1. Locality information for sockeye salmon samples. Regional subbgroupings and larger geographic collectives are indicated in
the table. Coordinates are given as N/W decimal degrees.
Sample Location Abbrev. Coordinates Sample Location Abbrev. Coordinates Asian (ASA) - Northern Salish Sea Region (SSR) - Central 1. Abira River; Hokkaido Is., Japan [5] ASA-A 44.9 147.52 16. Baker Lake [3] SSR-B 48.66 -121.68 2. Etamink Bay, Kamchatka Pen. [5] ASA-E 51.45 157.1 17. Cedar River, L. Washington [5] SSR-C 47.5 -122.22 3. Gavrushka Bay, Kuril Lake, Russia [5] ASA-G 51.48 157.04 18. Issaquah Creek, L.Washington [5] SSR-I 47.56 -122.06 4. Kamchatka River, Russia [5] ASA-K 56.23 162.42 19. Nooksack Drainage [3] SSR-N 48.78 -122.58 5. Vichenkiya River, Kamchatka Pen. [5] ASA-V 51.45 157.1 20. Whatcom Lake [1] SSR-W 48.73 -122.33 Alaska - Excluding Southeast Locations (AES) - Northern Olympic Peninsula (OPN) - Central 6. Goodnews River [4] AES-G 59.14 -161.47 21. Big Creek, Lake Quinault [5] OPN-B 47.53 -123.77
219 7. Kenai River [4] AES-K 60.48 -151.08 22. Ozette Lake [3] OPN-O 48.15 -124.67
8. Togiak River [4] AES-T 59.08 -160.34 23. Pleasant Lake [3] OPN-P 48.06 -124.33 SE Alaska and Northern British Columbia (SAN) - Central Columbia Basin - Upper (CBU) - Southern 9. Gingut Creek, Nass River; BC [5] SAN-G 55.18 -129.29 24. Cassimer Bay, Okanogan R. [3] CBU-C 48.1 -119.72 10. Iskut-Stikine; BC [4] SAN-I 56.75 -131.73 25. Osoyoos Lake, Okanogan R. [6] CBU-O 48.99 -119.44 11. Situk River; AK [7] SAN-S 59.63 -139.41 26. Skaha Lake, Okanogan R. [5,6] CBU-S 49.35 -119.58 12. Taku River; AK [4] SAN-T 58.58 -133.65 27. Wenatchee, Lake and upper R. [5] CBU-W 47.82 -120.76 Fraser River, British Columbia (FBC) - Central Columbia Basin - Mid-lower (CBL) - Southern 13. Adams River [5] FBC-A 50.93 -119.64 28. Link Creek [1] CBL-L 44.42 -121.73 14. Cultus Lake [8] FBC-C 49.06 -121.98 29. Redfish Lake [2] CBL-R 44.14 -114.91 15. Horsefly River [5] FBC-H 52.33 -121.41 30. Suttle Lake [1] CBL-S 44.42 -121.74 Source: [1] A. Matala, CRITFC [2] L. Park, NOAA [3] C. Dean, WDFW [4] C. Habicht, ADFG [5] E. Iwamoto, NOAA [6] R. Bussanich, ONA [7] G. Thorgaard, WSU [8] R. Devlin, DFO
219
Table 2. Sample locality, haplotype diversity (h), and nucleotide diversity (π) (x100) for Chinook salmon samples. Geographic collectives are indicated in the table. Data from Martin et al.
(2010).
Sample location N h π (x100) Northern group 1 Kamchatka Pen. 31 0.19 0.035 2 Yukon R. 20 0.62 0.127 3 Tuluksak R. 26 0.68 0.159 4 Gulkana R. 22 0.17 0.031 Central group 5 Chilliwack R. 23 0.56 0.201 6 Priest Rapids Hatch. 22 0.75 0.276 7 Lyons Ferry Hatch. 22 0.72 0.316 8 Tucannon R. 21 0.00 0.000 9 Willamette R. 24 0.74 0.479 Southern group 10 American R. 27 0.49 0.110 11 Tuolumne R. 23 0.68 0.164
220
Table 3. Sample locality, haplotype diversity (h), and nucleotide diversity (π) (x100) for rainbow trout samples. Geographic collectives
are indicated in the table. Diversity was not calculated for localities with five or less individuals. Data from Brunelli et al. (2010).
π π π Sample location N h (x100) Sample location N h (x100) Sample location N h (x100) Northern group Central group Southern group 1 Voyampolka 2 na na 14 Blackwater R. 8 0.25 0.042 35 Alsea R. 1 na na 2 Sedanka 2 na na 15 Tzenzaicut Lake 9 0.39 0.065 36 Warm Springs R. 17 0.67 0.198 3 Zhupanova R. 3 na na 16 Pennask Lake 9 0.22 0.074 37 Bake Oven Ck. 13 0.78 0.174 4 Swanson R. 9 0.67 0.313 17 Cowichan R. 4 na na 38 Rogue R. 6 0.87 0.288 5 Sashin Crk 4 na na 18 W. Fork Trout Ck. 8 0.00 0.000 39 Buck Ck. 1 na na 6 Ealue Lake 4 na na 19 Kootenay Lake 2 na na 40 Bridge Ck. 2 na na 7 Moosevale Crk 1 na na 20 Basin Ck. 9 0.50 0.083 41 Upper Williamson R. 4 na na 221 8 Turnagain R. 9 0.00 0.000 21 Fisher River 6 0.60 0.166 42 Witham Ck. 5 na na 9 Yakoun R. 7 0.48 0.237 21 Little Sheep Ck. 16 0.62 0.118 43 Threemile Ck. 4 na na 10 Copper R. 7 0.48 0.237 22 Hoh R. 6 0.33 0.166 44 Bridge Ck. 3 na na 11 Canyon Crk 1 na na 23 White R. 2 na na 45 Mud Ck. 3 na na 12 Zymoetz R. 1 na na 24 Wells Hatchery 3 na na 46 Thomas Ck. 3 na na 13 Morice R. 1 na na 25 NF Little Deep Ck. 8 0.61 0.196 47 Honey Ck. 2 na na 26 Touchet R. 5 na na 48 N. Fork Little Deep 1 na na 27 Abernathy Ck. 12 0.44 0.178 49 W. Little Owyhee R. 2 na na 28 Washougal R. 4 na na 50 Sheepheaven Ck. 2 na na 29 Kalama R. 12 0.68 0.297 51 Hayspur Hatch. 16 0.72 0.288 30 Hood R. 9 0.58 0.184 52 S. Tacoma Hatch. 7 0.00 0.000 32 Clearwater R. 9 0.50 0.083 53 Spokane Hatch. 9 0.67 0.221 33 Rapid R. 7 0.52 0.142 54 Scott Ck. 5 na na 34 Pahsimeroi Hatch. 8 0.75 0.172 55 Whale Rock Res. 6 0.33 0.221
Table 4. Haplotype composition, haplotype diversity (h), and nucleotide diversity (π) (x100) for sockeye salmon from single localities.
Diversity was not calculated for localities with five or less individuals. Table continues on consecutive pages.
Haplotype: ONER___ π Sample Location Abbrev. N 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 h (x100) 1. Abira ASA-A 19 17 ------2 -- 0.20 0.034 2. Etamink ASA-E 16 9 ------3 -- 3 1 0.63 0.174 3. Gavrushka ASA-G 5 2 ------1 2 ------na na 4. Kamchatka ASA-K 1 ------1 ------na na 5. Vichenkiya ASA-V 10 5 ------2 1 1 1 0.71 0.232 6. Goodnews AES-G 16 16 ------0.00 0.000 222 7. Kenai AES-K 16 16 ------0.00 0.000 8. Togiak AES-T 16 13 ------1 ------2 ------0.34 0.101 9. Gingut Ck. SAN-G 9 7 ------1 ------1 ------0.22 0.038 10. Iskut-Stikine SAN-I 15 12 -- 2 ------1 -- 0.36 0.065 11. Situk SAN-S 11 7 ------1 ------3 ------0.44 0.148 12. Taku SAN-T 16 11 1 ------1 ------2 -- 1 ------0.44 0.163 13. Adams FBC-A 10 10 ------0.20 0.068 14. Cultus FBC-C 7 2 ------5 ------0.48 0.081 15. Horsefly FBC-H 10 10 ------0.00 0.000
Haplotype: ONER___ π Sample Location Abbrev. N 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 h (x100) 16. Baker SSR-B 10 4 3 ------1 ------2 -- 0.51 0.087 17. Cedar River SSR-C 9 8 ------1 ------0.00 0.000 18. Issaquah Ck. SSR-I 8 8 ------0.00 0.000 19. Nooksack SSR-N 7 4 ------1 -- -- 2 ------0.67 0.178 20. Whatcom SSR-W 47 40 ------4 -- -- 1 2 ------0.27 0.048 21. Big Creek OPN-B 10 6 4 ------0.00 0.000 22. Ozette OPN-O 10 2 ------8 ------0.36 0.060 23. Pleasant OPN-P 15 14 ------1 -- 0.13 0.023 24. Cassimer CBU-C 7 5 -- -- 1 ------1 ------0.48 0.242 25. Osoyoos CBU-O 25 23 -- -- 2 ------0.15 0.078 26. Skaha CBU-S 32 26 -- -- 6 ------0.32 0.160 223 27. Wenatchee CBU-W 29 15 -- -- 9 -- 1 -- 1 ------1 -- 1 ------1 -- 0.63 0.276 28. Link Creek CBL-L 12 7 ------1 3 1 ------0.64 0.126 29. Redfish CBL-R 37 30 -- 7 ------0.41 0.076 30. Suttle CBL-S 33 26 1 ------2 2 -- 1 ------1 ------0.28 0.050
Table 5. Summary of comparative statistics for sockeye salmon, Chinook salmon, and rainbow trout showing range (low to high) by
locality in geographic collective, overall values for collectives considered as a whole, and for all locations pooled together.
Haplotype diversity (h) North Central South All Range Overall Range Overall Range Overall locations Sockeye salmon 0-0.71 0.38 0-0.67 0.36 0.15-0.64 0.40 0.38 Chinook salmon 0.17-0.68 0.41 0.56-0.75 0.69 0.49-0.68 0.74 0.82 Rainbow trout 0-0.67 0.56 0-0.75 0.66 0-0.87 0.86 0.77 Nucleotide diversity (π) x100 North Central South All Range Overall Range Overall Range Overall locations
224 Sockeye salmon 0-0.232 0.117 0-0.178 0.084 0.050-0.276 0.141 0.117
Chinook salmon 0.031-0.159 0.122 0.201-0.479 0.342 0.110-0.164 0.136 0.382 Rainbow trout 0-0.313 0.273 0-0.297 0.230 0-0.288 0.385 0.319
Modified exact tests: % significantly differentiated North Central South Overall Sockeye salmon 41% 45% 60% 44% Chinook salmon 83% 100% 0% 95% Rainbow trout 20% 38% 81% 61%
APPENDIX
APPENDIX A. Supplemental tables for all chapters
Table S1. Chapter One. Summary of sample ages, extraction data, species identification, and DLoop haplotype results for all ancient
samples described in the study. PCR method indicates the method that generated amplifiable DNA. Table continues on consecutive
pages
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45WT41 3070-B 2450 - 4950 1 73 silica none Snake R.: 45WT41 3151-C1 6650 - 7950 1 24 silica none Snake R.: 45WT41 3419-C1-2Q8 6650 - 7950 1 7 silica none 226 Snake R.: 45WT41 3782-C1-2H6 6650 - 7950 1 37 silica none
Snake R.: 45WT41 3785-C1-2O5 6650 - 7950 1 43 silica none Snake R.: 45WT41 3789-C1-2Q2 250 - 9950 1 46 silica none Snake R.: 45WT41 3789-C1-2Q2 250 - 9950 2 152 p:c none Snake R.: 45WT41 3805-C1-2M3 4950 - 6950 1 53 p:c none Snake R.: 45WT41 3805-C1-2M3 4950 - 6950 2 75 silica none Snake R.: 45WT41 4177-3 250 - 9950 1 288 p:c rescue Catostomus spp. Snake R.: 45WT41 4177-3 250 - 9950 2 42 silica rescue Snake R.: 45WT41 4177-4 250 - 9950 1 30 silica rescue Catostomus spp. Snake R.: 45WT41 4177-6 250 - 9950 1 32 silica rescue Catostomus spp. Snake R.: 45WT41 4177-6 250 - 9950 2 195 p:c standard Snake R.: 45WT41 4548-B-C2 1450 - 2950 1 198 p:c standard Chinook Type 1 TSA10 Snake R.: 45WT41 4582-C-O3 6650 - 7950 1 11 silica none Snake R.: 45WT41 4585-C-W2 6650 - 7950 1 14 silica none Snake R.: 45WT41 4619-B-C10 2450 - 4950 1 78 p:c standard Chinook Type 1 TSA01a Snake R.: 45WT41 6360-B 6650 - 7950 1 78 silica none Snake R.: 45WT41 6793-C-W2 250 - 9950 1 unk silica none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45WT41 6844-C-P9 250 - 9950 1 63 p:c standard Chinook Type 1 Incomplete (SE) Snake R.: 45WT41 6844-C-P9 250 - 9950 2 24 silica none Snake R.: 45WT41 6925-C-C2 250 - 9950 1 83 silica none Snake R.: 45WT41 6933-B-C6-1 250 - 9950 1 125 p:c none Snake R.: 45WT41 6933-B-C6-1 250 - 9950 2 11 silica none Snake R.: 45WT41 6933-B-C6-2 250 - 9950 1 75 p:c standard Catostomus spp. Snake R.: 45WT41 6933-B-C6-2 250 - 9950 2 33 silica none Snake R.: 45WT41 6933-B-C6-4 250 - 9950 1 18 silica rescue Catostomus spp. Snake R.: 45WT41 6933-B-C6-5 250 - 9950 1 87 p:c standard Catostomus spp. Snake R.: 45WT41 6940-B-C5 250 - 9950 1 87 p:c none Snake R.: 45WT41 6940-B-C5 250 - 9950 2 24 silica none Snake R.: 45WT41 7005-2K4 250 - 9950 1 8 silica none
227 Snake R.: 45WT41 577 1450 1 52 silica none Snake R.: 45FR40 833 1450 1 56 silica none
Snake R.: 45FR40 1202-1 1450 1 11 silica none Snake R.: 45FR40 1202-2 1450 1 62 p:c standard Chinook Type 1 TSA17 Snake R.: 45FR40 1202-2 1450 2 172 p:c standard Snake R.: 45FR40 1202-2 1450 3 11 silica standard Snake R.: 45FR40 1202-3 1450 1 329 p:c none Snake R.: 45FR40 1202-3 1450 2 46 p:c none Snake R.: 45FR40 1202-3 1450 3 37 silica none Snake R.: 45FR40 1202-4 1450 1 19 silica none Snake R.: 45FR40 1202-5 1450 1 7 silica none Snake R.: 45FR40 1202-6 1450 1 32 silica none Snake R.: 45FR40 1202-7 1450 1 37 silica standard Chinook Type 1 TSA23 Snake R.: 45FR40 1202-7 1450 2 78 p:c rescue Snake R.: 45FR40 1202-7 1450 3 309 p:c none Snake R.: 45FR40 1202-8 1450 1 84 silica standard Chinook Type 1 TSA23
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45FR40 1202-8 1450 2 280 p:c none Snake R.: 45FR40 1202-8 1450 3 132 p:c none Snake R.: 45FR40 1212-F1 1450 1 7 silica none Snake R.: 45FR40 2227-1559 450 - 3950 1 74 silica none Snake R.: 45WT134 2273-1412 450 - 3950 1 63 silica none Snake R.: 45WT134 2305-1578 450 - 3950 1 23 silica none Snake R.: 45WT134 2354-1605 450 - 3950 1 50 silica none Snake R.: 45WT134 2368-810 450 - 3950 1 54 silica none Snake R.: 45WT134 2374-1430 450 - 3950 1 43 silica none Snake R.: 45WT134 3399-781 450 - 3950 1 35 silica none Snake R.: 45WT134 3401-350 450 - 3950 1 37 silica none Snake R.: 45WT134 3402-422 450 - 3950 1 158 p:c rescue Chinook Type 1 TSA26 228 Snake R.: 45WT134 3402-422 450 - 3950 2 41 silica rescue
Snake R.: 45WT134 3403-498 450 - 3950 1 15 silica none Snake R.: 45WT134 3404-584 450 - 3950 1 280 p:c standard Chinook Type 1 TSA17 Snake R.: 45WT134 3404-584 450 - 3950 2 41 silica rescue Snake R.: 45WT134 3405-687 450 - 3950 1 242 p:c standard Oncorhynchus kisutch Snake R.: 45WT134 3405-687 450 - 3950 2 22 silica rescue Snake R.: 45WT134 3406-919 450 - 3950 1 52 silica none Snake R.: 45WT134 3408-1219 450 - 3950 1 45 silica none Snake R.: 45WT134 3435-638 450 - 3950 1 80 silica none Snake R.: 45FR50 3592-68.2 250 - 8950 1 11 silica none Snake R.: 45FR50 5182-68.2 250 - 8950 1 33 silica none Snake R.: 45FR50 5198-68.6 250 - 8950 1 73 silica none Snake R.: 45FR50 5215-68.2 250 - 8950 1 21 silica none Snake R.: 45FR50 6068-68.6 250 - 8950 1 58 silica none Snake R.: 45FR50 6571-68.0 250 - 8950 1 87 silica none Snake R.: 45FR50 6599-68.0 250 - 8950 1 20 silica none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45FR50 9199-68.3-1 250 - 8950 1 24 silica none Snake R.: 45FR50 9199-68.3-3 250 - 8950 1 46 p:c none Snake R.: 45FR50 9199-68.3-3 250 - 8950 2 84 silica none Snake R.: 45FR50 9394-68.1 250 - 8950 1 34 silica none Snake R.: 45FR50 9395-68.4 250 - 8950 1 31 p:c none Snake R.: 45FR50 9395-68.4 250 - 8950 2 22 silica none Snake R.: 45FR50 9401-64.0 250 - 8950 1 143 silica none Snake R.: 45FR50 9423-63.0 250 - 8950 1 191 p:c none Snake R.: 45FR50 9423-63.0 250 - 8950 2 54 silica none Snake R.: 45FR50 9423-63.0 250 - 8950 3 144 p:c none Snake R.: 45FR50 9423-63.0 250 - 8950 4 108 p:c none Snake R.: 45FR50 9434-62.22 250 - 8950 1 21 silica none 229 Snake R.: 45FR50 9440-64.184-1 250 - 8950 1 23 silica none
Snake R.: 45FR50 9440-64.184-2 250 - 8950 1 29 silica none Snake R.: 45FR50 9444-62.16 250 - 8950 1 52 silica none Snake R.: 45FR50 9494-62.10 250 - 8950 1 31 silica none Snake R.: 45FR50 9520-62.19-1 250 - 8950 1 37 silica none Snake R.: 45FR50 9520-62.19-2 250 - 8950 1 54 silica none Snake R.: 45FR39 53-535 250 - 2950 1 167 silica none Snake R.: 45FR39 57-516 250 - 2950 1 96 p:c standard Chinook Type 4 TSA23 Snake R.: 45FR39 57-516 250 - 2950 2 89 p:c none Snake R.: 45FR39 57-516 250 - 2950 3 52 silica none Snake R.: 45FR39 334-515 250 - 2950 1 139 p:c standard Ptychocheilus spp. Snake R.: 45FR39 334-515 250 - 2950 2 66 silica none Snake R.: 45FR39 334-515 250 - 2950 3 79 p:c none Snake R.: 45FR39 341-438 250 - 2950 1 85 p:c standard Oncorhynchus kisutch Snake R.: 45FR39 341-438 250 - 2950 2 135 silica none Snake R.: 45FR39 536-770-1 250 - 2950 1 42 silica standard Chinook Type 1 TSA17
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45FR39 536-770-2 250 - 2950 1 48 silica standard Chinook Type 1 TSA17 Snake R.: 45FR39 704-602 250 - 2950 1 32 silica none Snake R.: 45FR39 813-742 250 - 2950 1 52 silica none Snake R.: 45FR39 813-742 250 - 2950 2 62 p:c none Snake R.: 45FR39 903-235 250 - 2950 1 31 silica none Snake R.: 45FR39 1253-15 250 - 2950 1 87 silica none Snake R.: 45FR39 1253-15 250 - 2950 2 280 p:c none Snake R.: 45FR39 1344-449 250 - 2950 1 28 silica none Snake R.: 45FR39 1387-273 250 - 2950 1 48 silica none Snake R.: 45FR39 1513-658 250 - 2950 1 75 silica none Snake R.: 45FR39 1513-658 250 - 2950 2 58 p:c none Snake R.: 45FR39 1634-796 250 - 2950 1 46 p:c none 230 Snake R.: 45FR39 1634-796 250 - 2950 2 23 silica none
Snake R.: 45FR39 1634-796 250 - 2950 3 27 p:c none Snake R.: 45FR39 1731-450-1 250 - 2950 1 57 silica none Snake R.: 45FR39 2181-637 250 - 2950 1 27 silica rescue Catostomus spp. Snake R.: 45FR39 2379-368 250 - 2950 1 38 silica rescue Ptychocheilus spp. Snake R.: 45FR39 2602-679-1 250 - 2950 1 85 silica none Snake R.: 45FR39 2602-679-2 250 - 2950 1 23 silica none Snake R.: 45FR39 2730-316 250 - 2950 1 31 silica none Snake R.: 45FR39 2764-466-2 250 - 2950 1 54 silica none Snake R.: 45FR39 2834-540 250 - 2950 1 70 p:c none Snake R.: 45FR39 2834-540 250 - 2950 2 110 silica none Snake R.: 45FR39 2834-540 250 - 2950 3 415 p:c none Snake R.: 45FR39 3068-474-1 250 - 2950 1 30 silica standard Ptychocheilus spp. Snake R.: 45FR39 3327-350 250 - 2950 1 144 p:c standard Chinook Type 1 TSA17 Snake R.: 45FR39 3327-350 250 - 2950 2 35 silica standard Snake R.: 45FR39 3355-95 250 - 2950 1 140 p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45FR39 3355-95 250 - 2950 2 57 silica none Snake R.: 45FR39 3362-93 250 - 2950 1 44 silica none Snake R.: 45FR39 3364-134-1 250 - 2950 1 27 silica none Snake R.: 45FR39 3371-37 250 - 2950 1 33 silica standard Chinook Type 1 TSA10 Snake R.: 45FR39 3371-37 250 - 2950 2 168 p:c standard Snake R.: 45FR39 3389-61 250 - 2950 1 26 silica standard Ptychocheilus spp. Snake R.: 45FR39 3408-80 250 - 2950 1 44 silica none Snake R.: 45FR39 3417-66 250 - 2950 1 201 p:c none Snake R.: 45FR39 3417-66 250 - 2950 2 46 silica none Snake R.: 45FR39 3440-11 250 - 2950 1 85 p:c none Snake R.: 45FR39 3440-11 250 - 2950 2 60 silica none Snake R.: 45FR39 3444-153 250 - 2950 1 83 p:c standard Ptychocheilus spp. 231 Snake R.: 45FR39 3444-153 250 - 2950 2 37 silica none
Snake R.: 45FR39 4554-246 250 - 2950 1 49 silica none Snake R.: 45FR39 4556-485-1 250 - 2950 1 30 silica standard Catostomus spp. Snake R.: 45FR39 4556-485-2 250 - 2950 1 32 silica standard Catostomus spp. Snake R.: 45GA61 6071-0 650 - 950 1 73 p:c standard Chinook Type 1 TSA01a Snake R.: 45GA61 9015-0 650 - 950 1 233 p:c standard Chinook Type 1 TSA17 Snake R.: 45GA61 9800-0 450 - 7950 1 265 p:c standard Catostomus spp. Snake R.: 45GA61 9800-0 450 - 7950 2 19 silica none Snake R.: 45GA61 11134-54 650 - 950 1 241 p:c standard Chinook Type 1 TSA10 Snake R.: 45GA61 11135-59 650 - 950 1 412 p:c rescue Chinook Type 1 TSA17 Snake R.: 45GA61 11175-30 650 - 950 1 64 p:c standard Human (contamination) Snake R.: 45GA61 11175-30 650 - 950 2 34 silica none Snake R.: 45GA61 11197-68 450 - 7950 1 23 silica none Snake R.: 45GA61 11920-54 650 - 950 1 315 p:c standard Chinook Type 1 TSA17 Snake R.: 45GA61 11920-54 650 - 950 2 25 silica standard Snake R.: 45GA61 11924-51 450 - 7950 1 31 silica standard Catostomus spp.
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45GA61 11957-123 450 - 7950 1 44 p:c standard Catostomus spp. Snake R.: 45GA61 11957-123 450 - 7950 2 32 silica none Snake R.: 45GA61 11957-123 450 - 7950 3 85 p:c none Snake R.: 45GA61 12206-68 650 - 950 1 84 p:c standard Chinook Type 1 TSA17 Snake R.: 45GA61 12224-30 650 - 950 1 329 p:c standard Chinook Type 1 TSA17 Snake R.: 45GA61 12227-0 450 - 7950 1 38 silica none Snake R.: 45GA61 12248-30 650 - 950 1 37 p:c standard Catostomus spp. Snake R.: 45GA61 12248-30 650 - 950 2 37 silica none Snake R.: 45GA61 12248-30 650 - 950 3 134 p:c none Snake R.: 45GA61 12250-68 650 - 950 1 11 silica none Snake R.: 45GA61 12252-55 650 - 950 1 49 p:c none Snake R.: 45GA61 12252-55 650 - 950 2 32 silica none 232 Snake R.: 45GA61 12257-30 650 - 950 1 21 silica standard Chinook Type 1 TSA17
Snake R.: 45GA61 12270-59 650 - 950 1 437 p:c standard Chinook Type 1 TSA01a Snake R.: 45FR46 79-0 5450 - 7450 1 7 silica none Snake R.: 45FR46 117-0-1 250 - 4450 1 14 silica none Snake R.: 45FR46 117-0-2 250 - 4450 1 25.4 silica standard Chinook Type 4 TSA01a Snake R.: 45FR46 117-0-3 250 - 4450 1 34 silica none Snake R.: 45FR46 220-0 250 - 8950 1 unk silica none Snake R.: 45FR46 220-0-2 250 - 8950 1 28 p:c none Snake R.: 45FR46 220-0-2 250 - 8950 2 36 silica none Snake R.: 45FR46 539-0 250 - 8950 1 60 p:c none Snake R.: 45FR46 539-0 250 - 8950 2 15 silica none Snake R.: 45FR46 818-0 250 - 4450 1 28 silica none Snake R.: 45FR46 979-0-1 250 - 4450 1 20 silica none Snake R.: 45FR46 979-0-2 250 - 4450 1 22 silica none Snake R.: 45FR46 979-0-3 250 - 4450 1 25 silica none Snake R.: 45FR46 979-0-4 250 - 4450 1 26 silica none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45FR46 979-0-5 250 - 4450 1 35 silica none Snake R.: 45FR46 979-0-6 250 - 4450 1 40 silica none Snake R.: 45FR46 979-0-7 250 - 4450 1 253 p:c none Snake R.: 45FR46 1088-0 250 - 4450 1 47 silica none Snake R.: 45FR46 1120-0 250 - 8950 1 57 p:c none Snake R.: 45FR46 1120-0 250 - 8950 2 49 silica none Snake R.: 45FR46 1130-0-1 250 - 4450 1 201 p:c standard Ptychocheilus spp. Snake R.: 45FR46 1130-0-2 250 - 4450 1 15 silica none Snake R.: 45FR46 1130-0-3 250 - 4450 1 18 silica none Snake R.: 45FR46 1130-0-4 250 - 4450 1 24 silica none Snake R.: 45FR46 1130-0-5 250 - 4450 1 26 silica none Snake R.: 45FR46 1152-0-1 250 - 4450 1 46 p:c none 233 Snake R.: 45FR46 1152-0-1 250 - 4450 2 31 p:c none
Snake R.: 45FR46 1152-0-2 250 - 4450 1 52 p:c standard Catostomus spp. Snake R.: 45FR46 1152-0-2 250 - 4450 2 8 silica none Snake R.: 45FR46 1152-0-2 250 - 4450 3 60 p:c none Snake R.: 45FR46 1152-0-3 250 - 4450 1 22 silica none Snake R.: 45FR46 1152-0-3 250 - 4450 2 57 p:c none Snake R.: 45FR46 1185-0 250 - 4450 1 348 p:c standard Chinook Type 1 TSA17 Snake R.: 45FR46 1185-0 250 - 4450 2 37 silica standard Snake R.: 45FR46 2901-0 250 - 4450 1 37 silica none Snake R.: 45FR46 2932-0 250 - 4450 1 125 p:c none Snake R.: 45FR46 2932-0 250 - 4450 2 64 silica none Snake R.: 45FR46 3095-01 250 - 4450 1 60 p:c standard Ptychocheilus spp. Snake R.: 45FR46 3095-02 250 - 4450 1 20 silica none Snake R.: 45FR46 3095-03 250 - 4450 1 31 silica none Snake R.: 45FR46 3095-03 250 - 4450 2 89 p:c none Snake R.: 45FR46 3095-0-1 250 - 4450 1 21 silica none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Snake R.: 45FR46 3095-0-1 250 - 4450 2 46 p:c none Snake R.: 45FR46 3095-0-9 250 - 4450 1 18 silica rescue Ptychocheilus spp. Snake R.: 45FR46 3095-0-10 250 - 4450 1 19 silica none Snake R.: 45FR46 3095-0-11 250 - 4450 1 20 silica none Snake R.: 45FR46 3095-0-11 250 - 4450 3 10 silica none Snake R.: 45FR46 3095-0-2 250 - 4450 1 27 silica rescue Catostomus spp. Snake R.: 45FR46 3095-0-2b 250 - 4450 1 12 silica none Snake R.: 45FR46 3095-0-3 250 - 4450 1 12 silica none Snake R.: 45FR46 3095-0-4 250 - 4450 1 13 silica none Snake R.: 45FR46 3095-0-5 250 - 4450 1 15 silica none Snake R.: 45FR46 3095-0-6 250 - 4450 1 15 silica none Snake R.: 45FR46 3095-0-7 250 - 4450 1 16 silica none 234 Snake R.: 45FR46 3095-0-8 250 - 4450 1 17 silica none
Snake R.: 45FR46 3095-0-12 250 - 4450 1 142 p:c standard Catostomus spp. Snake R.: 45FR46 3207-0-1 250 - 4450 1 57 p:c standard Human (contamination) Snake R.: 45FR46 3207-0-1 250 - 4450 2 126 silica none Snake R.: 45FR46 3207-0-2 250 - 4450 1 45 silica none Snake R.: 45FR46 3207-0-3 250 - 4450 1 26 silica none Snake R.: 45FR46 3207-0-4 250 - 4450 1 38 silica none Spokane R.: 45SP266 328-32-1 3250 1 387 p:c none Spokane R.: 45SP266 328-32-2 3250 1 383 p:c rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 406-40-1 7200 1 18 silica none Spokane R.: 45SP266 406-40-2 7200 1 56 p:c none Spokane R.: 45SP266 458-48 2500 1 21 silica none Spokane R.: 45SP266 458-48 2500 2 267 p:c none Spokane R.: 45SP266 670-69 3250 1 16 silica none Spokane R.: 45SP266 711-74 3250 1 unk silica none Spokane R.: 45SP266 711-74 3250 2 250 p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Spokane R.: 45SP266 722-76-1 7200 1 84.0 p:c standard Chinook Type 1 TSA01a Spokane R.: 45SP266 722-76-1 7200 2 255 p:c rescue Spokane R.: 45SP266 722-76-2 7200 1 17 silica none Spokane R.: 45SP266 722-76-3 7200 1 182 p:c none Spokane R.: 45SP266 722-76-4 7200 1 134.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 722-76-5 7200 1 347.0 p:c rescue Chinook Type 1 TSA27 Spokane R.: 45SP266 766-80-1 7200 1 144 p:c none Spokane R.: 45SP266 766-80-2 7200 1 241.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 766-80-3 7200 1 18 p:c none Spokane R.: 45SP266 766-80-4 7200 1 12 silica none Spokane R.: 45SP266 766-80-5 7200 1 109 p:c none Spokane R.: 45SP266 766-80-6 7200 1 22 p:c none 235 Spokane R.: 45SP266 766-80-7 7200 1 72 p:c none
Spokane R.: 45SP266 766-80-8 7200 1 415 p:c none Spokane R.: 45SP266 766-80-9 7200 1 405.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 816-88 2500 1 47 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 816-88 2500 2 29 p:c none Spokane R.: 45SP266 816-88 2500 3 13 silica none Spokane R.: 45SP266 900-96-1 undetermined 1 30 p:c none Spokane R.: 45SP266 900-96-2 undetermined 1 24.0 silica rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 900-96-3 undetermined 1 233.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 909-97-1 7200 1 37 p:c standard Chinook Type 3 Incomplete (SE) Spokane R.: 45SP266 909-97-1 7200 2 32 p:c none Spokane R.: 45SP266 909-97-2 7200 1 19.0 silica rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 909-97-3 7200 1 29 p:c none Spokane R.: 45SP266 1362-148 3250 1 19 silica none Spokane R.: 45SP266 1362-148 3250 2 30 p:c none Spokane R.: 45SP266 1371-587-1 3250 or 7200 1 19 p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Spokane R.: 45SP266 1371-587-2 3250 or 7200 1 132.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 1371-587-3 3250 or 7200 1 58.0 p:c rescue Chinook Type 1 TSA17 Spokane R.: 45SP266 1371-587-4 3250 or 7200 1 30 p:c none Spokane R.: 45SP266 1371-587-5 3250 or 7200 1 22 silica none Spokane R.: 45SP266 1371-587-6 3250 or 7200 1 23 p:c none Spokane R.: 45SP266 1371-587-7 3250 or 7200 1 63.0 p:c rescue Chinook Type 1 TSA25 Spokane R.: 45SP266 1371-587-7 3250 or 7200 2 71.0 p:c rescue Spokane R.: 45SP266 1371-587-8 3250 or 7200 1 170.0 p:c rescue Chinook Type 1 TSA28 Spokane R.: 45SP266 1371-587-9 3250 or 7200 1 23 p:c none Spokane R.: 45SP266 1375-149-2 7200 1 18 silica none Spokane R.: 45SP266 1375-149-3 7200 1 26 p:c none Spokane R.: 45SP266 1405-153-1 7200 1 38 p:c none Spokane R.: 45SP266 1405-153-2 7200 1 20 silica none Spokane R.: 45SP266 1405-153-3 7200 1 26 p:c none
236 Spokane R.: 45SP266 1405-153-4 7200 1 26 p:c none
Spokane R.: 45SP266 1405-153-5 7200 1 36 p:c none Spokane R.: 45SP266 1405-153-6 7200 1 41 p:c none Spokane R.: 45SP266 1405-153-7 7200 1 26 p:c none Spokane R.: 45SP266 1405-153-8 7200 1 30 p:c none Spokane R.: 45SP266 1421-154-1 7200 1 32 silica none Spokane R.: 45SP266 1421-154-2 7200 1 33 p:c none Spokane R.: 45SP266 1531-165 3250 1 167.0 p:c rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 1544-167 3250 1 34 silica none Spokane R.: 45SP266 1544-167 3250 2 41 p:c none Spokane R.: 45SP266 1548-168 3250 1 26.0 silica rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 1548-168 3250 2 63.0 p:c rescue Spokane R.: 45SP266 1562-170-1 3250 or 7200 1 14 silica rescue Chinook Type 3 TSA17 Spokane R.: 45SP266 1562-170-1 3250 or 7200 2 26 silica none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Spokane R.: 45SP266 1562-170-1 3250 or 7200 3 42 p:c none Spokane R.: 45SP266 1562-170-1 3250 or 7200 4 143.0 p:c standard Spokane R.: 45SP266 1562-170-2 3250 or 7200 1 18.0 silica rescue Chinook Type 1 TSA17 Spokane R.: 45SP266 1562-170-2 3250 or 7200 2 unk p:c rescue Spokane R.: 45SP266 1562-170-2 3250 or 7200 3 28 p:c none Spokane R.: 45SP266 1566-171-1 3250 or 7200 1 66.0 p:c rescue Chinook Type 1 TSA17 Spokane R.: 45SP266 1566-171-1 3250 or 7200 2 17 silica none Spokane R.: 45SP266 1566-171-2 3250 or 7200 1 21 silica standard Chinook Type 1 TSA17 Spokane R.: 45SP266 1566-171-2 3250 or 7200 2 123.9 p:c rescue Spokane R.: 45SP266 1572-173 7200 1 16.7 silica rescue Chinook Type 1 TSA17 Spokane R.: 45SP266 1572-173 7200 2 60.0 p:c standard Spokane R.: 45SP266 1757-189 3250 1 39 p:c none Spokane R.: 45SP266 1781-191-1 7200 1 74.0 p:c standard Chinook Type 1 TSA17 Spokane R.: 45SP266 1781-191-2 7200 1 50 p:c rescue Chinook Type 1 Incomplete (SE)
237 Spokane R.: 45SP266 1781-191-2 7200 2 19 silica none
Spokane R.: 45SP266 1957-209 3250 1 27 p:c none Spokane R.: 45SP266 1964-210 undetermined 1 21 silica none Spokane R.: 45SP266 1964-210 undetermined 2 49 p:c none Spokane R.: 45SP266 2014-218 undetermined 1 75.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 2014-218 undetermined 2 24 silica none Spokane R.: 45SP266 2082-588-1 3250 1 25 silica none Spokane R.: 45SP266 2082-588-2 3250 1 27 p:c none Spokane R.: 45SP266 2082-588-3 3250 1 29 p:c none Spokane R.: 45SP266 2138-231-1 7200 1 44 p:c none Spokane R.: 45SP266 2138-231-2 7200 1 26 p:c none Spokane R.: 45SP266 2179-236 2500 1 30 silica none Spokane R.: 45SP266 2179-236 2500 2 28 p:c none Spokane R.: 45SP266 2211-239 2500 1 140.0 p:c rescue Chinook Type 1 TSA10
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Spokane R.: 45SP266 2249-241 3250 1 46 silica none Spokane R.: 45SP266 2249-241 3250 2 34 p:c none Spokane R.: 45SP266 2260-242 3250 1 29 p:c rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 2260-242 3250 2 46.0 silica none Spokane R.: 45SP266 2296-248-1 undetermined 1 38 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 2296-248-1 undetermined 2 40.0 p:c none Spokane R.: 45SP266 2296-248-2 undetermined 1 24 p:c none Spokane R.: 45SP266 2296-248-2 undetermined 2 49 p:c none Spokane R.: 45SP266 2296-248-3 undetermined 1 32 p:c none Spokane R.: 45SP266 2296-248-3 undetermined 2 28 p:c none Spokane R.: 45SP266 2296-248-5 undetermined 1 39 silica none Spokane R.: 45SP266 2296-248-5 undetermined 2 34 p:c none Spokane R.: 45SP266 2296-248-5 undetermined 3 40 p:c none Spokane R.: 45SP266 2296-248-6 undetermined 1 37 silica none
238 Spokane R.: 45SP266 2296-248-6 undetermined 2 33 p:c none
Spokane R.: 45SP266 2491-264 2500 1 18 silica none Spokane R.: 45SP266 2491-264 2500 2 32 p:c none Spokane R.: 45SP266 2515-269 undetermined 1 22 silica none Spokane R.: 45SP266 2515-269 undetermined 2 41 p:c none Spokane R.: 45SP266 2523-270-1 7200 1 34.5 silica rescue Chinook Type 2 TSA17 Spokane R.: 45SP266 2523-270-1 7200 2 87.0 p:c standard Spokane R.: 45SP266 2523-270-2 7200 1 53.0 p:c rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 2523-270-3 7200 1 82.0 p:c standard Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 2582-276 2500 1 15 silica none Spokane R.: 45SP266 2667-285 3250 1 126.0 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 2667-285 3250 1 17 silica none Spokane R.: 45SP266 2693-289 2500 1 144.0 p:c rescue Chinook Type 1 TSA01a Spokane R.: 45SP266 2742-295-1 3250 1 44 p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Spokane R.: 45SP266 2742-295-2 3250 1 32 p:c none Spokane R.: 45SP266 2742-295-3 3250 1 64.0 p:c none Spokane R.: 45SP266 2742-295-4 3250 1 9 silica none Spokane R.: 45SP266 2742-295-5 3250 1 41 p:c none Spokane R.: 45SP266 2742-295-6 3250 1 39 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 2748-296-1 3250 or 7200 1 40.0 p:c rescue Chinook Type 1 TSA25 Spokane R.: 45SP266 2748-296-1 3250 or 7200 2 unk silica none Spokane R.: 45SP266 2748-296-1 3250 or 7200 3 24.0 p:c none Spokane R.: 45SP266 2748-296-2 3250 or 7200 1 228 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 2748-296-3 3250 or 7200 1 23.1 silica standard Chinook Type 2 Incomplete (SE) Spokane R.: 45SP266 2754-297-1 7200 1 28 p:c none Spokane R.: 45SP266 2754-297-1 7200 2 26 p:c none Spokane R.: 45SP266 2754-297-2 7200 1 17 silica none Spokane R.: 45SP266 2765-298-1 7200 1 34 silica none
239 Spokane R.: 45SP266 2765-298-1 7200 2 27 p:c none
Spokane R.: 45SP266 2765-298-2 7200 1 41 p:c none Spokane R.: 45SP266 2867-307 2500 1 42 p:c none Spokane R.: 45SP266 2916-313-2 3250 or 7200 1 121.4 p:c standard Chinook Type 1 TSA25 Spokane R.: 45SP266 2916-313-3 3250 or 7200 1 58.0 silica standard Chinook Type 1 TSA10 Spokane R.: 45SP266 2916-313-4 3250 or 7200 1 31 p:c rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 3007-324 2500 1 19 silica none Spokane R.: 45SP266 3007-324 2500 2 43 p:c none Spokane R.: 45SP266 3202-349 7200 1 42 p:c rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 3240-354 3250 1 35 silica none Spokane R.: 45SP266 3240-354 3250 2 48 p:c none Spokane R.: 45SP266 3240-354 3250 3 41 p:c none Spokane R.: 45SP266 3272-359 7200 1 64.0 p:c rescue Chinook Type 1 TSA17 Spokane R.: 45SP266 3272-359 7200 2 121.0 p:c rescue
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Spokane R.: 45SP266 4119-443-1 2500 1 17 silica none Spokane R.: 45SP266 4119-443-1 2500 2 42 p:c none Spokane R.: 45SP266 4119-443-2 2500 1 38 p:c none Spokane R.: 45SP266 4119-443-2 2500 2 27 p:c none Spokane R.: 45SP266 5615-598 2500 1 23 silica none Spokane R.: 45SP266 5658-614-1 undetermined 1 unk silica none Spokane R.: 45SP266 5658-614-2 undetermined 1 31.4 silica rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 5658-614-2 undetermined 2 31 p:c standard Spokane R.: 45SP266 5658-614-2 undetermined 3 38.0 p:c none Spokane R.: 45SP266 5658-614-3 undetermined 1 24 p:c none Spokane R.: 45SP266 5658-614-4 undetermined 1 43 p:c none Spokane R.: 45SP266 5662-616-1 3250 or 7200 1 42 p:c rescue Chinook Type 1 TSA10 Spokane R.: 45SP266 5662-616-2 3250 or 7200 1 42.1 silica rescue Chinook Type 1 Incomplete (SE) Spokane R.: 45SP266 5662-616-2 3250 or 7200 2 107.0 p:c standard
240 Spokane R.: 45SP266 5662-616-2 3250 or 7200 3 117.0 p:c standard
Columbia R.: 45ST97 183273-18019 100 1 330 p:c none Columbia R.: 45ST97 183367-18297 100 1 530 p:c none Columbia R.: 45ST97 183587-15148 100 1 19 p:c none Columbia R.: 45ST97 183591-15260 100 1 46 p:c standard Chinook Type 1 TSA01a Columbia R.: 45DO189 3808 3127.5 +/- 132.5 1 103 p:c standard Chinook Type 4 TSA01a Columbia R.: 45DO189 3841 3127.5 +/- 132.5 1 unk p:c none Columbia R.: 45DO189 3746-1 3127.5 +/- 132.5 1 131 p:c standard Chinook Type 1 TSA10 Columbia R.: 45DO189 3746-2 3127.5 +/- 132.5 1 137 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3746-3 3127.5 +/- 132.5 1 138 p:c standard Chinook Type 1 TSA10 Columbia R.: 45DO189 3746-4 3127.5 +/- 132.5 1 117 p:c standard Chinook Type 1 TSA01a Columbia R.: 45DO189 3746-5 3127.5 +/- 132.5 1 94 p:c standard Chinook Type 1 TSA24 Columbia R.: 45DO189 3746-6 3127.5 +/- 132.5 1 105 p:c none Columbia R.: 45DO189 3746-7 3127.5 +/- 132.5 1 unk p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Columbia R.: 45DO189 3746-8 3127.5 +/- 132.5 1 unk p:c none Columbia R.: 45DO189 3758-1 3127.5 +/- 132.5 1 92 p:c standard Chinook Type 1 TSA10 Columbia R.: 45DO189 3758-2 3127.5 +/- 132.5 1 112 p:c standard Chinook Type 1 TSA10 Columbia R.: 45DO189 3758-3 3127.5 +/- 132.5 1 unk p:c none Columbia R.: 45DO189 3770-1 3127.5 +/- 132.5 1 113 p:c standard Chinook Type 1 TSA10 Columbia R.: 45DO189 3770-2 3127.5 +/- 132.5 1 141 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-3 3127.5 +/- 132.5 1 109 p:c standard Chinook Type 1 TSA01b Columbia R.: 45DO189 3770-4 3127.5 +/- 132.5 1 121 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-4 3127.5 +/- 132.5 2 101 p:c standard Columbia R.: 45DO189 3770-5 3127.5 +/- 132.5 1 163 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-6 3127.5 +/- 132.5 1 144 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-7 3127.5 +/- 132.5 1 121 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-8 3127.5 +/- 132.5 1 159 p:c standard Chinook Type 1 TSA01b Columbia R.: 45DO189 3770-9 3127.5 +/- 132.5 1 119 p:c standard Chinook Type 1 TSA17
241 Columbia R.: 45DO189 3770-10 3127.5 +/- 132.5 1 131 p:c standard Chinook Type 1 TSA01b
Columbia R.: 45DO189 3770-11 3127.5 +/- 132.5 1 101 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-12 3127.5 +/- 132.5 1 153 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-13 3127.5 +/- 132.5 1 152 p:c none Columbia R.: 45DO189 3770-14 3127.5 +/- 132.5 1 141 p:c standard Chinook Type 1 TSA23 Columbia R.: 45DO189 3770-15 3127.5 +/- 132.5 1 151 p:c standard Chinook Type 1 TSA01b Columbia R.: 45DO189 3770-16 3127.5 +/- 132.5 1 118 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-17 3127.5 +/- 132.5 1 106 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-18 3127.5 +/- 132.5 1 115 p:c standard Chinook Type 1 TSA01b Columbia R.: 45DO189 3770-19 3127.5 +/- 132.5 1 36 p:c none Columbia R.: 45DO189 3770-20 3127.5 +/- 132.5 1 38 p:c standard Chinook Type 1 TSA10 Columbia R.: 45DO189 3770-21 3127.5 +/- 132.5 1 43 p:c standard Chinook Type 1 TSA17 Columbia R.: 45DO189 3770-22 3127.5 +/- 132.5 1 105 p:c standard Chinook Type 3 TSA01b Columbia R.: 45FE45 7220 1150 +/- 50 1 127 p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Columbia R.: 45FE45 7221 1150 +/- 50 1 141 p:c standard Chinook Type 1 TSA01b Columbia R.: 45FE45 7221 1150 +/- 50 2 13 p:c standard Columbia R.: 45FE45 7274 1150 +/- 50 1 132 p:c none Columbia R.: 45FE45 7356 1150 +/- 50 1 125 p:c standard Chinook Type 1 TSA01b Columbia R.: 45FE45 7356 1150 +/- 50 2 106 p:c standard Columbia R.: 45FE45 7357 1150 +/- 50 1 166 p:c standard Chinook Type 1 TSA01b Columbia R.: 45FE45 7357 1150 +/- 50 2 120 p:c standard Columbia R.: 45FE45 7358 1150 +/- 50 1 101 p:c standard Chinook Type 1 TSA23 Columbia R.: 45FE45 7359 1150 +/- 50 1 unk p:c none Columbia R.: 45FE45 8747 1150 +/- 50 1 131 p:c none Columbia R.: 45FE44 1855 7627 +/- 100 1 140 p:c standard Chinook Type 1 TSA17 Columbia R.: 45FE44 1922 7627 +/- 100 1 98 p:c rescue Chinook Type 1 Incomplete (SE) Columbia R.: 45FE44 2014 7627 +/- 100 1 unk p:c none Columbia R.: 45FE44 2016 7627 +/- 100 1 unk p:c none
242 Columbia R.: 45FE44 2079 7627 +/- 100 1 135 p:c none
Columbia R.: 45FE44 2314 7627 +/- 100 1 90 p:c none Columbia R.: 45FE44 2314 7627 +/- 100 2 85 p:c none Columbia R.: 45FE44 2577 7627 +/- 100 1 107 p:c none Columbia R.: 45FE44 2577 7627 +/- 100 2 92 p:c none Columbia R.: 45FE44 2746 7627 +/- 100 1 unk p:c none Columbia R.: 45FE44 2750 7627 +/- 100 1 unk p:c none CRITFC: Revelstoke, BC CC-R-01 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-02 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-03 58 (year 1892) 1 unk p:c standard Human (contamination) CRITFC: Revelstoke, BC CC-R-04 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-05 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-06 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-06 58 (year 1892) 2 unk other none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype CRITFC: Revelstoke, BC CC-R-07 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-07 58 (year 1892) 2 unk other none CRITFC: Revelstoke, BC CC-R-08 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-08 58 (year 1892) 2 unk other none CRITFC: Revelstoke, BC CC-R-09 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-09 58 (year 1892) 2 unk other none CRITFC: Revelstoke, BC CC-R-10 58 (year 1892) 1 unk p:c none CRITFC: Revelstoke, BC CC-R-10 58 (year 1892) 2 unk other none CRITFC: Golden, BC CC-G-01 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-02 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-03 58 (year 1892) 1 unk p:c standard O. mykiss CRITFC: Golden, BC CC-G-04 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-05 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-05 58 (year 1892) 2 unk other none
243 CRITFC: Golden, BC CC-G-06 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-06 58 (year 1892) 2 unk other none
CRITFC: Golden, BC CC-G-07 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-07 58 (year 1892) 2 unk other none CRITFC: Golden, BC CC-G-08 58 (year 1892) 1 unk p:c none CRITFC: Golden, BC CC-G-08 58 (year 1892) 2 unk other none Sentinel Gap: 45KT1362 667-1 10200 1 2 p:c none Sentinel Gap: 45KT1362 667-2 10200 1 3 p:c none Sentinel Gap: 45KT1362 690-1 10200 1 2 p:c none Sentinel Gap: 45KT1362 690-2 10200 1 5 p:c none Sentinel Gap: 45KT1362 690-3 10200 1 7 p:c none Sentinel Gap: 45KT1362 858-1 10200 1 4 p:c none Sentinel Gap: 45KT1362 858-2 10200 1 4 p:c none
DNA extraction PCR and sequencing results Weight (mg) PCR Group: Site Sample ID Age (YBP) # and method method 12S species & 12S haplotype DLoop haplotype Sentinel Gap: 45KT1362 858-3 10200 1 4 p:c none Sentinel Gap: 45KT1362 858-4 10200 1 4 p:c none Sentinel Gap: 45KT1362 858-5 10200 1 7 p:c none Sentinel Gap: 45KT1362 858-6 10200 1 5 p:c none Sentinel Gap: 45KT1362 1104-1 10200 1 5 p:c none Sentinel Gap: 45KT1362 1104-2 10200 1 2 p:c none Sentinel Gap: 45KT1362 1104-3 10200 1 2 p:c none Sentinel Gap: 45KT1362 1200 10200 1 2 p:c none Sentinel Gap: 45KT1362 1204 10200 1 2 p:c none Sentinel Gap: 45KT1362 1208-1 10200 1 5 p:c none Sentinel Gap: 45KT1362 1208-2 10200 1 6 p:c none Sentinel Gap: 45KT1362 1350 10200 1 2 p:c none Nason Ck KS 01 1 11 (year 1939) 1 13 p:c none Nason Ck KS 01 1 11 (year 1939) 2 22 other none
244 Nason Ck KS 01 002 11 (year 1939) 1 10 p:c none Nason Ck KS 01 002 11 (year 1939) 2 18 other none
Nason Ck KS 01 003 11 (year 1939) 1 9 p:c none Nason Ck KS 01 003 11 (year 1939) 2 32 other none Nason Ck KS 01 004 11 (year 1939) 1 12 p:c none Nason Ck KS 01 004 11 (year 1939) 2 17 other none Nason Ck KS 01 005 11 (year 1939) 1 8 p:c none Nason Ck KS 01 005 11 (year 1939) 2 15 other none
Table S2. Chapter One. Location, run timing, and collection year(s) for contemporary Chinook Salmon in the study. The Columbia
River group is organized by tributary and the Snake River group by genetic stock identification reporting group. Table continues on
consecutive pages
Haplotype: TSA___ Location Run Collection year(s) 1A 1B 4A 10 12 17 18 19 22 25 Columbia River Group Carson and Leavenworth Hatchery Complex Carson Fish Hatchery (1) 1995 ------13 ------Entiat National Fish Hatchery (2,3) Spring 2002 - 2005 ------3 -- 13 ------Leavenworth National Fish Hatchery (2) Spring 2008 ------4 ------Winthrop National Fish Hatchery (2) Spring 2002 - 2010 -- 1 -- 1 -- 20 ------Entiat Rotary Screw Trap (2) Spring 2005 ------2 -- 1 ------
245 Spawning Ground Carcass Recovery (2) Spring & Summer 2003 - 2005 ------2 -- 13 ------Lower River Rotary Screw Trap (2) Spring 2009 -- 1 -- 4 -- 11 ------
Entiat River natural (3) Spring 2011 ------2 -- 8 ------Entiat River smolt (1) Spring 2002 ------9 ------Methow Spawning grounds (3) Spring <2011 ------21 -- -- 3 -- Twisp weir, spawning grounds (3) Spring <2011 ------3 -- 13 -- -- 2 -- Wenatchee White River (2) Spring <2011 ------1 -- 15 ------Dryden trap (2) Summer <2011 ------1 10 ------Chiwawa (1) Spring 2006 -- 1 ------9 ------Nason Creek (1) 2009 -- 1 ------Icicle Icicle Creek (2) Spring 2001 - 2010 -- 1 -- 2 -- 48 -- -- 1 --
Haplotype: TSA___ Location Run Collection year(s) 1A 1B 4A 10 12 17 18 19 22 25 Snake River Group Chamberlain Chamberlain Creek (4) Spring/Summer 2003- 2009 ------6 ------Chamberlain Creek, west fork (4) Spring/Summer 2003, 2009 ------4 ------Hells Canyon Catherine Creek (4) Spring/Summer 2008 - 2010 2 ------7 ------1 Upper Grande Ronde (4) Spring/Summer 2003 - 2009 1 ------9 ------Imnaha River (4) Spring/Summer 1998, 2008 1 1 -- 2 -- 5 ------Lemhi Hayden Creek (4) Spring/Summer 2003, 2009 ------3 ------Lemhi River, L3A Trap (4) Spring/Summer 2009, 2010 1 ------4 ------Middle Fork Salmon River Bear Valley Creek (4) Spring/Summer 2006, 2009 ------2 ------Big Creek (4) Spring/Summer 1999 - 2005 ------3 ------Camas Creek (4) Spring/Summer 2003- 2009 ------3 ------Cape Horn Creek (4) Spring/Summer 2005 - 2009 1 ------2 ------246 Elk Creek (4) Spring/Summer 2003 - 2006 ------3 ------
Marsh Creek (4) Spring/Summer 1989 ------1 ------Sulphur Creek (4) Spring/Summer 2003, 2007 ------2 ------South Fork Salmon River Lake Creek (4) Spring/Summer 2003 - 2010 1 ------6 ------Secesh River (4) Spring/Summer 1989 - 2009 ------5 ------Upper Salmon East Fork Salmon River (4) Spring/Summer 2004 - 2008 ------3 ------Pahsimeroi River (4) Spring/Summer 2002 - 2012 ------5 ------Sawtooth weir, upper Salmon (4) Spring/Summer 2003 - 2011 ------3 ------Upper Lemhi River (4) Spring/Summer 2000 - 2010 ------3 ------Valley Creek (4) Spring/Summer 1989 - 2003 ------2 ------1 Yankee Fork, West fork (4) Spring/Summer 2000 - 2007 1 ------2 ------Tucannon Tucannon River (3)* Spring/Summer <2010 ------21 ------Other Lyons Ferry Hatchery (3)* Fall <2010 3 11 1 4 -- 1 1 1 -- -- Source: (1) National Oceanic and Atmospheric Administration, (2) U.S. Fish and Wildlife Service, (3) Washington Department *Data from Martin et al. 2010
Table S3. Chapter One. Summary statistics for simulations of Wright-Fisher process on a subset of ancient samples for each group.
NeF: 100 500 1000 2500 5000 10K 25K 35K 45K 55K Columbia River group Mean 0.000 0.148 0.320 0.518 0.610 0.663 0.693 0.701 0.705 0.706 Median 0.000 0.000 0.379 0.553 0.633 0.677 0.701 0.706 0.708 0.710 Standard Error 0.000 0.003 0.003 0.002 0.002 0.001 0.001 0.001 0.001 0.000 Standard Deviation 0.011 0.199 0.218 0.163 0.114 0.078 0.050 0.042 0.036 0.033 Minimum 0.000 0.000 0.000 0.000 0.000 0.203 0.450 0.492 0.503 0.544 Maximum 0.502 0.667 0.764 0.821 0.828 0.817 0.814 0.799 0.800 0.791 Snake River group Mean 0.049 0.321 0.394 0.450 0.473 0.483 0.488 0.490 0.491 0.492 Median 0.000 0.376 0.434 0.472 0.482 0.488 0.490 0.491 0.492 0.493 Standard Error 0.002 0.003 0.002 0.002 0.001 0.001 0.001 0.001 0.000 0.000 Standard Deviation 0.131 0.206 0.175 0.127 0.092 0.067 0.043 0.037 0.033 0.029 Minimum 0.000 0.000 0.000 0.000 0.087 0.205 0.276 0.350 0.367 0.370 Maximum 0.666 0.668 0.667 0.667 0.665 0.647 0.612 0.602 0.593 0.584 Spokane River group Mean 0.000 0.018 0.098 0.295 0.424 0.511 0.564 0.580 0.584 0.589 Median 0.000 0.000 0.000 0.349 0.473 0.533 0.577 0.591 0.593 0.597 Standard Error 0.000 0.001 0.002 0.003 0.003 0.002 0.001 0.001 0.001 0.001 Standard Deviation 0.000 0.083 0.175 0.216 0.184 0.142 0.098 0.080 0.074 0.067 Minimum 0.000 0.000 0.000 0.000 0.000 0.000 0.063 0.182 0.199 0.291 Maximum 0.000 0.501 0.667 0.744 0.748 0.748 0.747 0.749 0.742 0.749
247
Table S4. Chapter One. Summary of all ancient samples that were successfully haplotyped.
Table continues on consecutive pages.
Group Site ID Sample ID 12S DLoop Columbia R. Fort Colville 45ST97 183591-15260 1 TSA01A Columbia R. GCDPA Tail Race 45DO189 3746-4 1 TSA01A Columbia R. GCDPA Tail Race 45DO189 3808 4 TSA01A Columbia R. GCDPA Tail Race 45DO189 3770-10 1 TSA01B Columbia R. GCDPA Tail Race 45DO189 3770-15 1 TSA01B Columbia R. GCDPA Tail Race 45DO189 3770-18 1 TSA01B Columbia R. GCDPA Tail Race 45DO189 3770-3 1 TSA01B Columbia R. GCDPA Tail Race 45DO189 3770-8 1 TSA01B Columbia R. GCDPA Tail Race 45DO189 3770-22 3 TSA01B Columbia R. GCDPA Tail Race 45DO189 3746-1 1 TSA10 Columbia R. GCDPA Tail Race 45DO189 3746-3 1 TSA10 Columbia R. GCDPA Tail Race 45DO189 3758-1 1 TSA10 Columbia R. GCDPA Tail Race 45DO189 3758-2 1 TSA10 Columbia R. GCDPA Tail Race 45DO189 3770-1 1 TSA10 Columbia R. GCDPA Tail Race 45DO189 3770-20 1 TSA10 Columbia R. GCDPA Tail Race 45DO189 3746-2 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-11 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-12 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-16 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-17 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-2 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-21 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-4 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-5 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-6 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-7 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-9 1 TSA17 Columbia R. GCDPA Tail Race 45DO189 3770-14 1 TSA23 Columbia R. GCDPA Tail Race 45DO189 3746-5 1 TSA24 Columbia R. Ksunku (Kettle Falls) 45FE45 7221 1 TSA01B Columbia R. Ksunku (Kettle Falls) 45FE45 7356 1 TSA01B Columbia R. Ksunku (Kettle Falls) 45FE45 7357 1 TSA01B Columbia R. Ksunku (Kettle Falls) 45FE45 7358 1 TSA23 Columbia R. Shonitkwu (Kettle Falls) 45FE44 1855 1 TSA17
248
Group Site ID Sample ID 12S DLoop Snake R. Three Springs Bar 45FR39 3371-37 1 TSA10 Snake R. Three Springs Bar 45FR39 3327-350 1 TSA17 Snake R. Three Springs Bar 45FR39 536-770-1 1 TSA17 Snake R. Three Springs Bar 45FR39 536-770-2 1 TSA17 Snake R. Three Springs Bar 45FR39 57-516 4 TSA23 Snake R. Granite Point 45WT41 4619-B-C10 1 TSA01A Snake R. Granite Point 45WT41 4548-B-C2 1 TSA10 Snake R. Harder site 45FR40 1202-2 1 TSA17 Snake R. Harder site 45FR40 1202-7 1 TSA23 Snake R. Harder site 45FR40 1202-8 1 TSA23 Snake R. Hatiuhpuh 45WT134 3404-584 1 TSA17 Snake R. Hatiuhpuh 45WT134 3402-422 1 TSA26 Snake R. Wexpusnime 45GA61 12270-59 1 TSA01A Snake R. Wexpusnime 45GA61 6071-0 1 TSA01A Snake R. Wexpusnime 45GA61 11134-54 1 TSA10 Snake R. Wexpusnime 45GA61 11135-59 1 TSA17 Snake R. Wexpusnime 45GA61 11920-54 1 TSA17 Snake R. Wexpusnime 45GA61 12206-68 1 TSA17 Snake R. Wexpusnime 45GA61 12224-30 1 TSA17 Snake R. Wexpusnime 45GA61 12257-30 1 TSA17 Snake R. Wexpusnime 45GA61 9015-0 1 TSA17 Snake R. Windust Caves 45FR46 117-0-2 4 TSA01A Snake R. Windust Caves 45FR46 1185-0 1 TSA17
249
Group Site ID Sample ID 12S DLoop Spokane R. Latah Site 45SP266 2693-289 1 TSA01A Spokane R. Latah Site 45SP266 722-76-1 1 TSA01A Spokane R. Latah Site 45SP266 1531-165 1 TSA10 Spokane R. Latah Site 45SP266 1548-168 1 TSA10 Spokane R. Latah Site 45SP266 2211-239 1 TSA10 Spokane R. Latah Site 45SP266 2260-242 1 TSA10 Spokane R. Latah Site 45SP266 2523-270-2 1 TSA10 Spokane R. Latah Site 45SP266 2916-313-3 1 TSA10 Spokane R. Latah Site 45SP266 3202-349 1 TSA10 Spokane R. Latah Site 45SP266 328-32-2 1 TSA10 Spokane R. Latah Site 45SP266 900-96-2 1 TSA10 Spokane R. Latah Site 45SP266 5662-616-1 1 TSA10 Spokane R. Latah Site 45SP266 1371-587-3 1 TSA17 Spokane R. Latah Site 45SP266 1562-170-2 1 TSA17 Spokane R. Latah Site 45SP266 1566-171-1 1 TSA17 Spokane R. Latah Site 45SP266 1566-171-2 1 TSA17 Spokane R. Latah Site 45SP266 1572-173 1 TSA17 Spokane R. Latah Site 45SP266 1781-191-1 1 TSA17 Spokane R. Latah Site 45SP266 3272-359 1 TSA17 Spokane R. Latah Site 45SP266 2523-270-1 2 TSA17 Spokane R. Latah Site 45SP266 1562-170-1 3 TSA17 Spokane R. Latah Site 45SP266 1371-587-7 1 TSA25 Spokane R. Latah Site 45SP266 2748-296-1 1 TSA25 Spokane R. Latah Site 45SP266 2916-313-2 1 TSA25 Spokane R. Latah Site 45SP266 722-76-5 1 TSA27 Spokane R. Latah Site 45SP266 1371-587-8 1 TSA28
250
Table S5. Chapter Two. Sample identification, origin (location), extraction method, extraction sub-sample weight (mg), and PCR outcomes for all samples in the study. Listing of "- #" following the extract identification indicates multiple extractions from a single vertebra. PCR outcomes are indicated by "+" for successful amplification and "-" when no amplification occurred. Table continues on consecutive pages.
Extraction PCR outcomes method & Extract Location weight (mg) Standard Rescue SR002 -1 Snake - Strawberry Island EM 1 7 - - + SR002 -2 Snake - Strawberry Island EM 2 53 - - + SR002 -3 Snake - Strawberry Island EM 2 171 - - - + SR009b -1 Snake - Harder site EM 1 11 - + + SR009b -2 Snake - Harder site EM 2 62 - + + SR009b -3 Snake - Harder site EM 2 172 + + SR009g -1 Snake - Harder site EM 1 37 - + + SR009g -2 Snake - Harder site EM 2 309 - - + SR014 -1 Snake - Strawberry Island EM 1 7 - - + SR017a -1 Snake - Strawberry Island EM 1 27 - + - - + SR017b -1 Snake - Strawberry Island EM 1 23 - + + SR017b -2 Snake - Strawberry Island EM 2 86 + + SR017b -3 Snake - Strawberry Island EM 2 91 + + SR017b -4 Snake - Strawberry Island EM 2 86 - - + SR017d -1 Snake - Strawberry Island EM 1 32 - + + SR025 -1 Snake - Strawberry Island EM 1 46 - - - + + SR025 -2 Snake - Strawberry Island EM 2 91 - - + SR025 -3 Snake - Strawberry Island EM 2 133 - - + SR027a -1 Snake - Strawberry Island EM 1 46 - - + SR027a -2 Snake - Strawberry Island EM 2 172 + - + SR027a -3 Snake - Strawberry Island EM 2 199 - - + SR027a -4 Snake - Strawberry Island EM 2 274 - + + SR027b -1 Snake - Strawberry Island EM 1 31 - - + SR027d -1 Snake - Strawberry Island EM 1 10 - + + SR029 -1 Snake - Strawberry Island EM 1 46 - - + + SR029 -2 Snake - Strawberry Island EM 2 203 + + SR029 -3 Snake - Strawberry Island EM 2 233 + +
251
Extraction PCR outcomes method & Extract Location weight (mg) Standard Rescue SR029 -4 Snake - Strawberry Island EM 2 58 - - + + SR029a -1 Snake - Strawberry Island EM 2 97 - + + SR029c -1 Snake - Strawberry Island EM 2 281 - - + - + sr029d -1 Snake - Strawberry Island EM 1 29 - - - + SR039a -1 Snake - Three Springs Bar EM 1 42 + + SR039b -1 Snake - Three Springs Bar EM 1 48 - + - + SR071 -1 Snake - Three Springs Bar EM 1 38 - - + SR076a -1 Snake - Three Springs Bar EM 1 30 + + + + SR076b -1 Snake - Three Springs Bar EM 1 32 + + + + SR086a -1 Snake - Three Springs Bar EM 1 30 + + SR087 -1 Snake - Three Springs Bar EM 1 33 - + + SR087 -2 Snake - Three Springs Bar EM 2 168 + - + SR089 -1 Snake - Three Springs Bar EM 1 27 - - - + SR090 -1 Snake - Three Springs Bar EM 2 83 + + SR094 -1 Snake - Three Springs Bar EM 1 35 - + + SR094 -2 Snake - Three Springs Bar EM 2 144 + + SR097 -1 Snake - Three Springs Bar EM 1 26 - + + SR098c -1 Snake - Granite Point EM 1 42 - - - + SR098c -2 Snake - Granite Point EM 2 288 - - - + SR098d -1 Snake - Granite Point EM 1 30 - - - + SR098f -1 Snake - Granite Point EM 1 32 - - - + SR098f -2 Snake - Granite Point EM 2 195 + + + + SR110d -1 Snake - Granite Point EM 1 18 - - - + + SR115 -1 Snake - Hatiuhpuh EM 1 41 - - - + SR115 -2 Snake - Hatiuhpuh EM 2 280 + + SR118 -1 Snake - Hatiuhpuh EM 1 41 - - + SR118 -2 Snake - Hatiuhpuh EM 2 158 - - - + SR119 -1 Snake - Hatiuhpuh EM 1 22 - - - + SR119 -2 Snake - Hatiuhpuh EM 2 242 + + SR122 -1 Snake - Windust Caves EM 1 37 - + - + SR122 -2 Snake - Windust Caves EM 2 348 + + SR131 -1 Snake - Wexpusnime EM 2 265 + + SR139 -1 Snake - Wexpusnime EM 1 25 - - + + SR139 -2 Snake - Wexpusnime EM 2 315 + + SR140 -1 Snake - Wexpusnime EM 1 21 + + SR141 -1 Snake - Wexpusnime EM 1 31 - + - + SR142 -1 Snake - Wexpusnime EM 2 412 - - + SR158 -1 Snake - Windust Caves EM 1 25 + + SR170 -1 Snake - Windust Caves EM 1 18 - - - + SR175 -1 Snake - Windust Caves EM 2 142 + + SR176b -1 Snake - Windust Caves EM 1 27 - - - +
252
Extraction PCR outcomes method & Extract Location weight (mg) Standard Rescue u0722a -1 Spokane EM 2 84 + + u0722d -1 Spokane EM 2 134 - - - + u0722e -1 Spokane EM 2 347 - - + u0766b -1 Spokane EM 2 241 - - + u0766i -1 Spokane EM 2 405 - - - + u0900b -1 Spokane EM 1 24 - - + u0900c -1 Spokane EM 2 233 - + + u0909b -1 Spokane EM 1 19 - - - + u1371b -1 Spokane EM 2 132 - - - + u1371c -1 Spokane EM 2 58 - - + u1371g -1 Spokane EM 2 63 - - - + u1371g -2 Spokane EM 2 71 - - + u1371h -1 Spokane EM 2 170 - - + u1531 -1 Spokane EM 2 167 - - + u1548 -1 Spokane EM 2 63 - - - - + u1548 -2 Spokane EM 1 26 - - + u1562a -1 Spokane EM 2 143 - + + u1562b -1 Spokane EM 1 18 - - + u1566a -1 Spokane EM 2 66 - - + u1566b -1 Spokane EM 2 124 - - + + u1572 -1 Spokane EM 2 60 - - + u1572 -2 Spokane EM 1 17 - + + u1781a -1 Spokane EM 2 74 - + + u2014 -1 Spokane EM 2 75 - - + u2211 -1 Spokane EM 2 140 - - + u2260 -1 Spokane EM 1 46 - - + u2296a -1 Spokane EM 1 40 - - + u2523a -1 Spokane EM 2 87 - - - + u2523a -2 Spokane EM 1 35 - + + u2523b -1 Spokane EM 2 53 - - + u2523c -1 Spokane EM 2 82 - + + u2667 -1 Spokane EM 2 126 - - + u2693 -1 Spokane EM 2 144 - - + u2742c -1 Spokane EM 2 64 - - + u2748b -1 Spokane EM 2 228 - - + u2748c -1 Spokane EM 1 23 - - + + u2916b -1 Spokane EM 2 121 - - + + u2916c -1 Spokane EM 2 58 - + + u3272 -1 Spokane EM 2 64 - - + u3272 -2 Spokane EM 2 121 - - + u5658b -1 Spokane EM 1 31 - - + +
253
Extraction PCR outcomes method & Extract Location weight (mg) Standard Rescue u5662 -1 Spokane EM 1 42 - - + u5662 -2 Spokane EM 2 107 - + + u5662 -3 Spokane EM 2 117 - + +
254
Table S6. Chapter Four. Results of exact test of population differentiation for sockeye salmon sample locations and geographical sub-groupings. Non-differentiation exact P-values are listed with significant values (P<0.05) highlighted in grey. Population and group abbreviations are indicated in Chapter Four, Table 1. Table continues on consecutive pages.
ASAA ASAE ASAV AESG AESK AEST SANG SANI SANS SANT FBCA FBCC FBCH ASAA 0.06 0.04 0.49 0.49 0.09 0.19 0.38 0.01 0.04 0.43 0.00 0.53 ASAE 0.06 0.84 0.01 0.01 0.19 0.10 0.10 0.43 0.13 0.21 0.00 0.12 ASAV 0.04 0.84 0.00 0.00 0.23 0.33 0.09 0.73 0.29 0.32 0.02 0.03 AESG 0.49 0.01 0.00 1.00 0.23 0.12 0.10 0.02 0.04 0.38 0.00 1.00 AESK 0.49 0.01 0.00 1.00 0.23 0.12 0.10 0.02 0.04 0.38 0.00 1.00 AEST 0.09 0.19 0.23 0.23 0.23 0.38 0.24 0.38 0.49 0.45 0.00 0.69 SANG 0.19 0.10 0.33 0.12 0.12 0.38 0.28 0.32 0.49 0.58 0.01 0.21 SANI 0.38 0.10 0.09 0.10 0.10 0.24 0.28 0.04 0.24 0.46 0.00 0.70 SANS 0.01 0.43 0.73 0.02 0.02 0.38 0.32 0.04 0.32 0.17 0.01 0.14 SANT 0.04 0.13 0.29 0.04 0.04 0.49 0.49 0.24 0.32 0.70 0.00 0.66 FBCA 0.43 0.21 0.32 0.38 0.38 0.45 0.58 0.46 0.17 0.70 0.00 1.00 FBCC 0.00 0.00 0.02 0.00 0.00 0.00 0.01 0.00 0.01 0.00 0.00 0.00 FBCH 0.53 0.12 0.03 1.00 1.00 0.69 0.21 0.70 0.14 0.66 1.00 0.00 SSRB 0.01 0.09 0.25 0.00 0.00 0.01 0.06 0.02 0.02 0.10 0.02 0.01 0.01 SSRC 0.38 0.18 0.19 0.36 0.36 0.43 1.00 0.44 0.22 0.69 1.00 0.00 0.47 SSRI 1.00 0.27 0.20 1.00 1.00 0.69 1.00 0.69 0.22 0.86 1.00 0.01 1.00 SSRN 0.03 0.09 0.30 0.02 0.02 0.08 0.22 0.09 0.10 0.26 0.09 0.03 0.05 SSRW 0.13 0.00 0.00 0.69 0.70 0.06 0.20 0.06 0.01 0.01 0.37 0.00 1.00 OPNB 0.01 0.02 0.06 0.01 0.01 0.02 0.09 0.02 0.03 0.22 0.09 0.00 0.09 OPNO 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 OPNP 1.00 0.08 0.02 0.48 0.48 0.49 0.32 0.73 0.04 0.22 0.65 0.00 1.00 CBUC 0.13 0.14 0.63 0.08 0.08 0.28 0.67 0.31 0.25 0.54 0.36 0.02 0.16 CBUO 0.16 0.00 0.01 0.51 0.51 0.06 0.12 0.10 0.01 0.02 0.34 0.00 1.00 CBUS 0.01 0.00 0.00 0.16 0.16 0.01 0.04 0.01 0.00 0.00 0.10 0.00 0.31 CBUW 0.01 0.00 0.02 0.03 0.02 0.01 0.14 0.02 0.01 0.01 0.17 0.00 0.18 CBLL 0.01 0.02 0.11 0.01 0.01 0.04 0.29 0.03 0.07 0.07 0.16 0.20 0.11 CBLR 0.02 0.00 0.00 0.09 0.09 0.01 0.03 0.36 0.00 0.00 0.09 0.00 0.32 CBLS 0.34 0.00 0.02 0.82 0.82 0.32 0.66 0.28 0.08 0.15 0.71 0.00 1.00
255
SSRB SSRC SSRI SSRN SSRW OPNB OPNO OPNP CBUC CBUO CBUS CBUW CBLL CBLR CBLS ASAA 0.01 0.38 1.00 0.03 0.13 0.01 0.00 1.00 0.13 0.16 0.01 0.01 0.01 0.02 0.34 ASAE 0.09 0.18 0.27 0.09 0.00 0.02 0.00 0.08 0.14 0.00 0.00 0.00 0.02 0.00 0.00 ASAV 0.25 0.19 0.20 0.30 0.00 0.06 0.00 0.02 0.63 0.01 0.00 0.02 0.11 0.00 0.02 AESG 0.00 0.36 1.00 0.02 0.69 0.01 0.00 0.48 0.08 0.51 0.16 0.03 0.01 0.09 0.82 AESK 0.00 0.36 1.00 0.02 0.70 0.01 0.00 0.48 0.08 0.51 0.16 0.02 0.01 0.09 0.82 AEST 0.01 0.43 0.69 0.08 0.06 0.02 0.00 0.49 0.28 0.06 0.01 0.01 0.04 0.01 0.32 SANG 0.06 1.00 1.00 0.22 0.20 0.09 0.00 0.32 0.67 0.12 0.04 0.14 0.29 0.03 0.66 SANI 0.02 0.44 0.69 0.09 0.06 0.02 0.00 0.73 0.31 0.10 0.01 0.02 0.03 0.36 0.28 SANS 0.02 0.22 0.22 0.10 0.01 0.03 0.00 0.04 0.25 0.01 0.00 0.01 0.07 0.00 0.08 SANT 0.10 0.69 0.86 0.26 0.01 0.22 0.00 0.22 0.54 0.02 0.00 0.01 0.07 0.00 0.15 FBCA 0.02 1.00 1.00 0.09 0.37 0.09 0.00 0.65 0.36 0.34 0.10 0.17 0.16 0.09 0.71 FBCC 0.01 0.00 0.01 0.03 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.20 0.00 0.00 FBCH 0.01 0.47 1.00 0.05 1.00 0.09 0.00 1.00 0.16 1.00 0.31 0.18 0.11 0.32 1.00 SSRB 0.03 0.04 0.18 0.00 0.46 0.00 0.01 0.18 0.00 0.00 0.01 0.02 0.00 0.00 SSRC 0.03 1.00 0.10 0.34 0.08 0.00 0.62 0.40 0.32 0.08 0.21 0.24 0.08 0.67 SSRI 0.04 1.00 0.08 1.00 0.09 0.00 1.00 0.20 1.00 0.32 0.39 0.27 0.32 1.00 SSRN 0.18 0.10 0.08 0.02 0.06 0.00 0.04 0.46 0.02 0.01 0.06 0.14 0.01 0.07 SSRW 0.00 0.34 1.00 0.02 0.00 0.00 0.39 0.10 0.09 0.00 0.00 0.01 0.00 0.03 OPNB 0.46 0.08 0.09 0.06 0.00 0.00 0.02 0.11 0.01 0.00 0.01 0.04 0.00 0.07 OPNO 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 OPNP 0.01 0.62 1.00 0.04 0.39 0.02 0.00 0.22 0.29 0.06 0.04 0.02 0.06 0.65 CBUC 0.18 0.40 0.20 0.46 0.10 0.11 0.00 0.22 0.20 0.26 0.76 0.34 0.04 0.27 CBUO 0.00 0.32 1.00 0.02 0.09 0.01 0.00 0.29 0.20 0.45 0.01 0.00 0.01 0.22 CBUS 0.00 0.08 0.32 0.01 0.00 0.00 0.00 0.06 0.26 0.45 0.07 0.00 0.00 0.01 CBUW 0.01 0.21 0.39 0.06 0.00 0.01 0.00 0.04 0.76 0.01 0.07 0.02 0.00 0.00 CBLL 0.02 0.24 0.27 0.14 0.01 0.04 0.00 0.02 0.34 0.00 0.00 0.02 0.00 0.25 CBLR 0.00 0.08 0.32 0.01 0.00 0.00 0.00 0.06 0.04 0.01 0.00 0.00 0.00 0.00 CBLS 0.00 0.67 1.00 0.07 0.03 0.07 0.00 0.65 0.27 0.22 0.01 0.00 0.25 0.00
256
Table S7. Chapter Four. Results of exact test of population differentiation for Chinook salmon sample locations. P-values are listed with significant values (P<0.05) highlighted in grey.
Population numbers match those in Martin et al. (2010) and Chapter four, Table 4.
Pop 1 Pop 2 Pop 3 Pop 4 Pop 5 Pop 6 Pop 7 Pop 8 Pop 9 Pop 10 Pop 11 Pop 1 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00 Pop 2 0.00 0.64 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 3 0.00 0.64 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 4 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 Pop 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 7 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 Pop 8 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 10 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.25 Pop 11 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.25
257
Table S8. Chapter Four. Results of exact test of population differentiation for rainbow trout sample locations. P-values are listed with
significant values (P<0.05) highlighted in grey. Population numbers match those in Brunelli et al. (2010) and Chapter four, Table 5.
Pop 4 Pop 8 Pop 9 Pop 10 Pop 14 Pop 15 Pop 16 Pop 18 Pop 20 Pop 21 Pop 25 Pop 27 Pop 29 Pop 30 Pop 31 Pop 32 Pop 34 Pop 36 Pop 37 Pop 38 Pop 51 Pop 52 Pop 53 Pop 55 Pop 4 0.04 0.78 0.78 0.06 0.04 0.00 0.05 0.01 0.13 0.05 0.26 0.94 0.24 0.01 0.02 0.07 0.00 0.01 0.22 0.02 0.05 0.02 0.00 Pop 8 0.04 0.23 0.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.25 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 Pop 9 0.78 0.23 1.00 0.01 0.01 0.00 0.01 0.00 0.03 0.01 1.00 0.56 0.06 0.00 0.01 0.02 0.00 0.00 0.12 0.00 0.02 0.01 0.00 Pop 10 0.78 0.23 1.00 0.01 0.01 0.00 0.01 0.00 0.03 0.02 1.00 0.56 0.06 0.00 0.01 0.02 0.00 0.00 0.12 0.00 0.02 0.01 0.00 Pop 14 0.06 0.00 0.01 0.01 0.47 0.00 1.00 0.01 0.38 0.33 0.00 0.20 1.00 0.72 0.21 0.18 0.00 0.17 0.04 0.31 1.00 0.06 0.00 Pop 15 0.04 0.00 0.01 0.01 0.47 0.00 0.47 0.01 0.28 0.15 0.00 0.07 0.47 0.06 0.10 0.12 0.00 0.03 0.03 0.14 0.48 0.04 0.00 Pop 16 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Pop 18 0.05 0.00 0.01 0.01 1.00 0.47 0.00 0.01 0.17 0.20 0.00 0.07 0.58 0.23 0.21 0.07 0.00 0.04 0.02 0.31 -1.00 0.05 0.00 Pop 20 0.01 0.00 0.00 0.00 0.01 0.01 0.00 0.01 0.02 0.22 0.00 0.02 0.05 0.00 0.34 0.06 0.00 0.15 0.02 0.00 0.01 0.01 0.00 258 Pop 21 0.13 0.00 0.03 0.03 0.38 0.28 0.00 0.17 0.02 0.49 0.00 0.22 0.75 0.12 0.16 0.54 0.01 0.14 0.31 0.31 0.19 0.18 0.01
Pop 25 0.05 0.00 0.01 0.02 0.33 0.15 0.00 0.20 0.22 0.49 0.00 0.22 1.00 0.12 1.00 0.56 0.00 0.33 0.11 0.37 0.32 0.05 0.00 Pop 27 0.26 0.25 1.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.01 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 Pop 29 0.94 0.01 0.56 0.56 0.20 0.07 0.00 0.07 0.02 0.22 0.22 0.07 0.67 0.06 0.14 0.11 0.00 0.08 0.21 0.01 0.22 0.02 0.00 Pop 30 0.24 0.00 0.06 0.06 1.00 0.47 0.00 0.58 0.05 0.75 1.00 0.01 0.67 0.14 0.58 0.55 0.02 0.18 0.27 0.17 1.00 0.08 0.00 Pop 31 0.01 0.00 0.00 0.00 0.72 0.06 0.00 0.23 0.00 0.12 0.12 0.00 0.06 0.14 0.08 0.14 0.00 0.29 0.02 0.01 0.26 0.01 0.00 Pop 32 0.02 0.00 0.01 0.01 0.21 0.10 0.00 0.21 0.34 0.16 1.00 0.00 0.14 0.58 0.08 0.27 0.00 0.29 0.05 0.05 0.21 0.02 0.00 Pop 34 0.07 0.00 0.02 0.02 0.18 0.12 0.00 0.07 0.06 0.54 0.56 0.00 0.11 0.55 0.14 0.27 0.02 0.76 0.21 0.06 0.18 0.07 0.00 Pop 36 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.02 0.00 0.00 0.02 0.00 0.01 0.00 0.00 0.00 0.00 Pop 37 0.01 0.00 0.00 0.00 0.17 0.03 0.00 0.04 0.15 0.14 0.33 0.00 0.08 0.18 0.29 0.29 0.76 0.00 0.06 0.00 0.11 0.01 0.00 Pop 38 0.22 0.01 0.12 0.12 0.04 0.03 0.00 0.02 0.02 0.31 0.11 0.02 0.21 0.27 0.02 0.05 0.21 0.01 0.06 0.05 0.02 0.05 0.01 Pop 51 0.02 0.00 0.00 0.00 0.31 0.14 0.00 0.31 0.00 0.31 0.37 0.00 0.01 0.17 0.01 0.05 0.06 0.00 0.00 0.05 0.42 0.02 0.00 Pop 52 0.05 0.00 0.02 0.02 1.00 0.48 0.00 -1.00 0.01 0.19 0.32 0.00 0.22 1.00 0.26 0.21 0.18 0.00 0.11 0.02 0.42 0.05 0.00 Pop 53 0.02 0.00 0.01 0.01 0.06 0.04 0.00 0.05 0.01 0.18 0.05 0.00 0.02 0.08 0.01 0.02 0.07 0.00 0.01 0.05 0.02 0.05 0.00 Pop 55 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00
APPENDIX B. Annotated script used to model genetic drift (Chapter One).
################# MODEL SET-UP ################# #!/usr/bin/perl use strict; ##Used to call a "pragma," to tell perl to be strict in interpreting the code use warnings; ##Used to call a "pragma," to tell perl to issue warnings use Getopt::Long; ##Used to call a module, used in gathering information from the command line use Pod::Usage qw(pod2usage); ##Used to produce information if the user call the help flag (i.e. perl Simulation_Wright_Fisher_with_notes.pl -help) ########################################################## ##Default Values ##These are the three variables that can be altered and their default values my $simulations = 5000; ##Number of simulations my $repeats = 795; ##Number of generations
259 my $population = 100; ##Effective population size
########################################################## ##Usage Information and Help, this is what shows up when you use the help flag =head1 SYNOPSIS Simulation_Wright_Fisher.pl -sims 5000 -rep 795 -N 100 =head1 OPTIONS -sims number of simulations to perform -rep number of repeated draws per simulation -N number of populations =cut ########################################################## ##Modified Values and Calling Help Section, modifies the default values based on input from the user GetOptions( q(help) => \my $help, q(verbose) => \my $verbose,
"reps=s" => \$repeats, "sims=s" => \$simulations, "N=s" => \$population ) or pod2usage(q(-verbose) => 1); pod2usage(q(-verbose) => 1) if $help; ########################################################## ##Print settings, prints out the variables used in the analysis (very useful if you are running many analyses) print "Initial Settings\nSimulations: $simulations\nRepeats: $repeats\nPop. Size: $population\n";
################# MODEL ################# ##Create New Populations my $simulation_count = 1; ##Counter to let you know what simulation is being conducted (so you don't stare at a blank screen wondering if it is working
260 print "\t\t\t\tType1\tType2\tType3\tType4\tType5\tType6\n"; ###Formats the output
##The while function is a looping function, and will continue until the part in parentheses is no longer true ##Example1: while (pizza is still hot){eat the pizza} This loop will never end so you can eat pizza forever ##Example2: while (captain Picard has hair){cut hair} This loop ended when he went bald and his hair is no longer cut ##No claim of accuracy, legality, suitability, or humor is made for these examples while ($simulations > 0){ ##This 'while' loop will continue until simulations variable reaches zero (see below), each iteration reduces by 1 print STDERR "Simulation #$simulation_count\n"; ##Prints out the information regarding which simulation is being conducted (optional) my %population = (); ## type => count ##"hash"/"dictionary" where population information is stored (has structure: key => associated value) my $type1 = 0.071; ##The starting haplotype frequencies (can be changed, removed, or additional added) my $type2 = 0.214; my $type3 = 0.311; my $type4 = 0.329;
my $type5 = 0.075; my $pop_count = $repeats; ##New temporary variable is created using number of times a population should be created (generations) ##This is where it starts to get a bit difficult, conceptually ##Here another while loop is started within the old while loop ##Example3: while (while I am at work) { while (the temperature is above 70 outside) { I am sad } } ##In Example3 I will be sad if I am at work and it is above 70 degrees outside while ($pop_count > 0){ ##For each simulation, this while loop will iterate for the number of generations (or repeats) specified my $pop_size = $population; ##Variable for the number of individuals to generate %population = (); ##Population information cleared so previous generation isn't mixed up with current generation ##This is a while loop within the other two while loops while ($pop_size > 0){ ##This while loop will generate each individual in the current generation
261 ##This part is used to first create an individual and then assign it to a type
##Example4: individual -> 0.4 ##0-----0.1999|0.2----0.3999|0.4---- ##Type 1 |Type 2 |Type 3 ##In this example, the individual is Type 3, and would be counted in the dictionary as type three (i.e. type 3 => old count + 1) my $individual = rand(1); ##Random number between 0-1 if ($individual < $type1){ ##Type 1 $population{"1"} ++; }elsif($individual < $type1+$type2){ ##Type 2 $population{"2"} ++; }elsif($individual < $type1+$type2+$type3){ ##Type 3 $population{"3"} ++; }elsif($individual < $type1+$type2+$type3+$type4){ ##Type 4 $population{"4"} ++; }elsif($individual < $type1+$type2+$type3+$type4+$type5){ ##Type 5
$population{"5"} ++; $pop_size --; ##Number of individuals remaining to be counted and is reduced each iteration } $pop_count --; ##Number of generations remaining to be counted and is reduced each iteration ##Replaces the old frequencies of each type with the newly generated one ##Replace current type frequencies of each type with newly generated frequencies if (exists $population{"1"} and $population{"1"} > 0){$type1 = $population{"1"}/$population}else{$type1 = 0}; if (exists $population{"2"} and $population{"2"} > 0){$type2 = $population{"2"}/$population}else{$type2 = 0}; if (exists $population{"3"} and $population{"3"} > 0){$type3 = $population{"3"}/$population}else{$type3 = 0}; if (exists $population{"4"} and $population{"4"} > 0){$type4 = $population{"4"}/$population}else{$type4 = 0}; if (exists $population{"5"} and $population{"5"} > 0){$type5 = $population{"5"}/$population}else{$type5 = 0}; } print "\t\t\t\t$type1\t$type2\t$type3\t$type4\t$type5\t$type6\n"; ##For each simulation, the type frequency is output after the final generation is completed
262 $simulations --; ###Simulation variable is decreased by one
$simulation_count ++; }