INFORMING CONSERVATION ACTIONS: GENETIC APPROACHES TO CHINOOK SALMON MANAGEMENT IN THE SACRAMENTO RIVER OF CALIFORNIA By Sara Hugentobler A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Integrative Biology – Doctor of Philosophy Ecology, Evolutionary Biology, and Behavior – Dual Major 2023 ABSTRACT Biocomplexity provides many benefits to ecosystems, chief among them allowing for species persistence and resilience. Genetic diversity is the fundamental building block that allows for intraspecific diversity. Unfortunately, anthropogenic changes to the environment have led to a sharp decrease in the abundance of many species, therefore reducing genetic diversity within and among populations. This increases the need for ways to monitor this complexity, and genetic methods are a very promising tool. In this dissertation, I explore biocomplexity through the lens of life history diversity in Chinook salmon (Oncorhynchus tshawytscha), a culturally, ecologically, and economically important species. The Sacramento River of California is the only place throughout the entire range where fish return at four different times during the year to spawn (Fall, Late Fall, Spring, and Winter), providing them with important adaptive variation to ensure population persistence and resilience. This phenotype, known as their run type or run timing, can be identified using genetics and is important for monitoring, as the run types are morphologically indistinguishable at most life stages and two of the run types are federally listed as threatened and endangered (Spring and Winter, respectively). My research explores broadly how we can use genetic methods to assess and monitor biocomplexity of imperiled species in highly altered environments using genetic tools and methods. In Chapter 1, “Genetic Assessment of Floodplain Habitat Use by Juvenile Chinook salmon”, I explore how biocomplexity in juvenile Chinook salmon in a managed floodplain in California can buffer the effects of climate change. To do this, I genetically identified juvenile Chinook salmon samples from surveys in the Sacramento River and Yolo Bypass to run type using an 80 loci informative Fluidigm panel. I first analyzed how accurate current management methods are at identifying run-type as compared to genetic methods, finding that current methods are often inaccurate. I further found that drought conditions had negative impacts on imperiled populations of juvenile Chinook salmon. Despite this, I found that even during periods of drought, the Yolo Bypass juvenile Chinook salmon attained larger sizes than the adjacent Sacramento River, suggesting that managed floodplain is critical for maintaining diversity in this system. In Chapter 2, “Remnant salmon life history diversity rediscovered in a highly compressed habitat”, I explore how genetic tools can be utilized to understand run timing in the Yuba River of California. I did this by assigning individuals to early or late migrating groups, based on informative genetic markers from the GREB1L region of the Chinook genome and compared that to date of entry in the system. I found that despite large amounts of anthropogenic alteration, the Yuba River supports life history diversity of Fall and Spring run types, and this diversity is correlated with the genetic markers in the GREB1L region of the Chinook genome. This study highlights the incredible resilience of Chinook salmon populations in the Yuba River, as well as validates exciting new genomic regions that can be used for monitoring populations in the Sacramento River. In Chapter 3, “Genetic divergence of recently introduced populations of Chinook salmon (Oncorhynchus tshawytscha) in New Zealand”, I explore how New Zealand populations of Chinook salmon compare genetically to each other, as well as how they have diverged from their source, Chinook salmon from the Sacramento River of California. This research is the first time New Zealand Chinook salmon have been compared to all four run types in the Sacramento River and is a critical first step in understanding those relationships. To do this, I analyzed genomic data obtained from genotyping by sequencing and restriction site associated DNA sequencing data from Chinook salmon sampled in rivers in New Zealand and all major tributaries and run types from the Sacramento River. I found that there is genetic structure between the different rivers in New Zealand, and that although New Zealand fish have diverged from the Sacramento River fish, they appear more genetically similar to contemporary Fall and Spring run populations than to contemporary Winter run. This research highlights the importance of genomic tools to understand genetic relationships and could inform restoration efforts such as genetic rescue. This body of work highlights the importance of using genetic tools for management of imperiled species, especially identifying and monitoring biocomplexity. It also addresses how anthropogenic activities can impact species and systems, which will be important in informing how to mitigate potential impacts on imperiled species. To all those who wander and are lost: this is the beginning of the journey, not the end. One day you will find the light that was taken from you again. iv ACKNOWLEDGMENTS The PhD is an immense undertaking with many challenges that cannot be completed without the help of others. I am incredibly thankful for all those who have helped and mentored me on this journey. At the top of this list is my advisor, Dr. Mariah Meek, for answering the email of a wandering professional in need of a doctoral home, and for all the guidance and tutelage after that. Mariah built a wonderful community that is a testament to her skill and compassion. My academic home was truly a home thanks to her. I would next like to thank my committee for their mentorship. Kim Scribner, Sarah Fitzpatrick, and Nadya Mamoozadeh provided invaluable feedback to make this dissertation the best it could be, all while suffering through numerous proposal and chapter drafts. I would also like to thank the endless support of my lab mates, who provided commiseration and solace through this difficult journey. In particular, I thank Miranda Wade for taking this journey with me, providing many opportunities for books, coffee, and analysis to help me through these last 6 years. I also want to acknowledge Ben Kline, who brought light and laughter to our lab, and was always there with a baked good and hug when the journey was difficult, pushing me through the last stage of my dissertation when it was darkest. I also thank my other lab members, past and present, for their support including Charlene Tarsa, Sierra Kaszubinski, Shannon O’Leary, Tasha Thompson, Hannah Rothkopf, Erin Collins, Alexandra Zhang, Emily Bardwell, Arianna Troia, and Scott Jackson. This dissertation would not be possible without the support and resources provided to me of many people and institutions. I received financial support from fellowships and grants, including Ecology, Evolution, and Behavior summer fellowships, travel fellowships through the Council of Graduate Students, the MSU Graduate School, and the American Fisheries Society Genetics Section, and the MSU Dissertation Completion Fellowship, just to name a few. I would also like to acknowledge our many authors and collaborators. Special thanks to Ted Sommer, Louise Conrad, Alisha Goodbla, Brian Schreier, and Pascale Goertler for your time and expertise on the Yolo chapter. Special thanks to Malte Willmes, Rachel Johnston, Flora Cordoleani, Anna Sturrock, and Natalie Stauffer-Olsen for your help writing the Yuba Chapter. Thank you to Andrea Schreier and Emily Funk for use of your genomic resources, and particularly Andrea Schreier for all your help with friendly review and genetic analysis. Thank you to Daniel Gomez- Uchida and Rodrigo Marin Nahuelpi at the Universidad de Concepción for your genetic resources on the New Zealand samples. Lastly, thank you to Melanie Cheung and especially the v Winnemem Wintu tribe for your precious resources and for helping the New Zealand chapter become a reality. I also give thanks to the many other friends that helped me along this journey. To all the IBIO and EEB girlies, this community at MSU wouldn’t have been possible or even nice without you. In particular, I thank Emily Liljestrand for indulging me in rants, providing me snacks, and calling me when times got tough. I also thank those the friends I had before graduate school. Eleni Shenk for being the best friend a girl could have, Adriana Hertel-Wulff for our solace together as scientists at a young age, and Bryce Williams and Ryan Price for your friendship and laughter over the years. Getting into graduate school and pursuing my dreams would not be possible without the help of countless mentors throughout my educational career. Thanks to the many biology teachers who encouraged my love of all natural things, but in particular, Herond Hoyt, who pushed me to be the best biologist I could be. My thanks go out to Carl Safina, who gave a young high school student career advice and then later provided mentorship and letters of recommendation to get me here. To Wayne Potts and Jimmy Ruff, who first gave me my first research experience and opened the door to graduate school, one doorknob I would not have even seen. Special thanks to Jimmy who taught me invaluable genetics skills and mentored me for hours to interview for PhD programs. To Mark Elvin, who reminded me again of my value when the degree ground my confidence down. And lastly, I am thankful to my family. Grey, Drew, but especially Mom and Dad for your support throughout my entire life to helping me pursue crazy dream. To my parents for supporting me through endless standardized tests, trips to destinations to nurture the budding biologist, helping me move more than once without a sun visor, and supporting me through the process of applying to graduate school not once or twice but five times. I will be forever grateful. And most thanks to my small family, Fluff and Zac, who have perhaps suffered most on my behalf in this doctoral journey, having encountered a very different woman six years ago when I started this journey. Thank you Fluff for the near constant snuggles and diligent supervision when writing this dissertation. And most of all Zac. I couldn’t imagine a better companion on this dissertation journey, and you have been there since the start. I am grateful for all the tears you wiped away when I truly believed I couldn’t do it. This dissertation is here because of you. vi TABLE OF CONTENTS CHAPTER 1: GENETIC ASSESSMENT OF FLOODPLAIN HABITAT USE BY JUVENILE CHINOOK SALMON .....................................................................................................................1 BIBLIOGRAPHY ..............................................................................................................22 APPENDIX 1A: SUPPLEMENTAL TABLE AND FIGURES .......................................29 CHAPTER 2: REMNANT SALMON LIFE HISTORY DIVERSITY REDISCOVERED IN A HIGHLY COMPRESSED HABITAT ..........................................................................................33 BIBLIOGRAPHY ..............................................................................................................49 APPENDIX 2A: GREB1L FLUIDIGM SEQUENCES ....................................................54 APPENDIX 2B: GENOTYPES BY SAMPLING METHOD ..........................................55 CHAPTER 3: GENETIC DIVERGENCE OF RECENTLY INTRODUCED POPULATIONS OF CHINOOK SALMON IN NEW ZEALAND ..........................................................................56 BIBLIOGRAPHY ..............................................................................................................79 vii CHAPTER 1: GENETIC ASSESSMENT OF FLOODPLAIN HABITAT USE BY JUVENILE CHINOOK SALMON Chapter 1: This chapter has been submitted for publication to San Franscisco Estuary and Watershed Sciences and is currently under peer-review. Other contributing authors: J. Louise Conrad, Alisha Goodbla, Ted Sommer, Mariah Meek ABSTRACT Climate change is having widespread negative effects on freshwater environments, including an increasing frequency and severity of droughts. Drought conditions present unique challenges for the federally listed Central Valley Chinook Salmon (Oncorhynchus tshawytscha), which use the already limited floodplain in the Central Valley as rearing habitat. In this study, we examined how differing hydrologic conditions influence the run composition of juvenile Chinook Salmon in the floodplain (Yolo Bypass) versus the mainstem of the Sacramento River. Juvenile Chinook Salmon from the Yolo Bypass and areas along the Sacramento River were identified to the genetically distinct runs (fall, late fall, winter, and spring) from 2013-2019. We found overwhelmingly that Length at Date methods are misclassifying fish, particularly late fall and spring run fish, and winter-run fish in the bypass. Using this genetic run-timing, we found that the abundances of endangered runs (spring and winter) are reduced during low flow periods in both the bypass and Sacramento River. Even during drought conditions, juvenile Chinook Salmon rearing in the Yolo Bypass attained significantly larger sizes than those in the Sacramento River. When comparing fish growth across time, during wet years fish in the bypass start smaller and get significantly larger over the course of the year as compared to drought years, while during both wet and dry years fish in the Sacramento River largely attain the same size. This suggests that floodplain habitat is critical to maintaining diversity in juvenile Chinook Salmon. KEYWORDS Chinook Salmon, genetics, monitoring, drought, floodplain, life history diversity, Yolo Bypass, Sacramento River INTRODUCTION Climate change presents a distinct threat to freshwater systems, as these systems often have a lack of connectivity between habitats making it often impossible for species to migrate to more favorable environments. The rise in temperature in freshwater basins is likely to lead to 1 changes in habitat quality and quantity, and conditions are predicted to worsen (Woodward et al. 2010; Ficklin et al. 2014). Overall changing water conditions (such as increase in temperature or habitat fragmentation) have already greatly reduced some freshwater population sizes, likely altering the overall amount of biodiversity and biocomplexity in these systems (Ficke et al. 2007; Brucet et al. 2012). Maintaining genetic and phenotypic diversity is necessary for population resiliency in the face of these fluctuations, and loss of biocomplexity further reduces any given population’s ability to respond to climatic change and drought (Crozier et al. 2008). The Central Valley of California is predicted to become one of the most water scarce areas in the world due to climate change and increasing water use (Famiglietti et al. 2011). Recently, the Central Valley experienced one of the longest and most severe droughts in California history, spanning the years 2012-2016 (Xiao et al. 2017). During this time, water in the largest watershed in California, the Sacramento River watershed, was at an all-time low, with the worst period of drought occurring in 2015. This led to vastly reduced connectivity between Sacramento River and its adjacent floodplain habitats, which have already been negatively impacted by extensive development and channelization (James and Singer 2008). Reduced access to floodplain habitat is particularly troubling because seasonally flooded habitats in the Sacramento River are critical for native freshwater species, providing important spawning, rearing, and feeding opportunities (Sommer et al. 2001a; van Dyke and Wasson 2005). For example, the Yolo Bypass, one of the few remaining large scale seasonal floodplain habitats in the upper San Francisco estuary, provides habitat for 45 different animal species and flood protection for the lower Sacramento Valley (Salcido 2012). One species that utilizes the bypass is Chinook Salmon (Oncorhynchus tshawytscha), which includes two federally listed ESUs (Sommer et al. 2001a). Many juvenile Chinook within the Sacramento River basin use the bypass as feeding and rearing habitat as they make outward migrations to the Pacific Ocean. For fish migrating from the Sacramento Valley, the primary alternative route is the mainstem of the Sacramento River, which is extremely channelized and has high water velocities (Sommer et al. 2001c). This mainstem habitat is often suboptimal for Chinook Salmon rearing and is correlated with high mortality (Michel et al. 2015). In contrast, off-channel habitats often provide more favorable conditions in the form of increased food resources and shelter from predators (Jeffres et al. 2008; Limm and Marchetti 2009). 2 Within the bypass, more favorable habitat conditions are correlated with an increase in the overall abundance and size of juvenile Chinook (Katz et al. 2013; Hellmair et al. 2018). Furthermore, evidence suggests that the bypass facilitates increased biocomplexity in the form of variation in juvenile size at out-migration, which can have significant impacts on ocean survival (MacArthur 1955; Schindler et al. 2010; Woodson et al. 2013; Goertler et al. 2018). Evidence from other off-channel habitats in the Central Valley suggest that areas like the bypass can provide a “shifting habitat mosaic” which leads to differing growth rates during differing hydrological conditions (Cordoleani et al. 2022). This diversity of habitats across space and time is important for maintaining biocomplexity, leading to an overall portfolio effect (Greene et al. 2009). This can lead to some life history traits performing better under different conditions, providing population buffering and overall stability of the species (Sturrock et al. 2015). In addition to the portfolio effect provided by variation in size, the Central Valley is the only location within the Chinook Salmon range that has four distinct Chinook Salmon spawning life-histories (runs), named for the time they return from the ocean to freshwater rivers to spawn (Meek et al. 2014). It is widely accepted that this life history diversity provides important biocomplexity necessary to mitigate adverse effects of changing conditions in the environment on any one population (Hilborn et al. 2003; Carlson and Satterthwaite 2011). However, the spring and winter-run are experiencing population declines in excess of 90%, and the U.S. Fish and Wildlife Service currently lists them as threatened and endangered, respectively, under the US Endangered Species Act (National Marine Fisheries Service 2014). The loss of either or both the spring and winter-run would represent an extreme loss of the biocomplexity of the region. Intensifying drought conditions in the Central Valley have led to extremely low water conditions that may reduce its ability to provide adequate habitat for all runs. Currently, we do not know to what extent the bypass supports juvenile Chinook of the different runs in terms of abundance or residence time. Many of the natural resource agencies working in the Central Valley have used non-genetic methods to identify juveniles to run type, mainly using a Length at Date (LAD) model (Harvey 2011). These criteria were introduced in the 1970s and incorporate fork length and date of capture to determine a classification. Although this method is expedient for use in the field, there is evidence the classifications are highly inaccurate (Harvey et al. 2014). 3 The purpose of this study was to examine the differences (if any) in run biocomplexity between the two habitat types—floodplain of the bypass (YBY) and mainstem river habitat of the Lower Sacramento River (LSR) by addressing the following questions: 1) Do genetic methods and the LAD model show similar patterns of run compositiacross the YBY and adjacent LSR? And 2) Do genetically determined run and size distributions differ between the YBY and LSR? By answering these questions, we will be able to better understand how Chinook Salmon (and available habitat) can be better managed to promote run biocomplexity. On a more regional level, this study provides insights into the degree to which different run identification methods (LAD vs. genetic) are usable in different habitats (e.g., floodplain vs. channel). METHODS Study Site The bypass is a managed floodplain that provides flood control for the city of Sacramento. The 24,000-ha region is one of the only remaining floodplain habitats within the Sacramento River basin. Habitat in the bypass includes grasslands, managed wildlife areas, agriculture, tidal wetlands and channels and perennial ponds (Sommer et al. 2001b; Sommer et al. 2005). Water enters the bypass from a few sources, creating access points for juvenile Chinook Salmon (Fig 1.1). Downstream migrating Chinook Salmon can most easily enter the bypass when the Sacramento River overtops the Fremont Weir, located at the northern part of the bypass (Sommer et al. 2001a). When water overtops the weir, water fans out across the bypass, creating suitable Chinook rearing habitat (Katz et al. 2017; Takata et al. 2017). During dry periods when the Sacramento River does not spill over the Fremont Weir, there are still substantial tidal river flows in Yolo Bypass from its base near Rio Vista, allowing young salmon to access the region (Goertler et al. 2018). During these periods, there are additional flow inputs from smaller westside tributaries (e.g. Putah and Cache Creeks) that enter a perennial channel called the “Toe Drain.” Consequently, juvenile salmon can access the region in both flood and non-flood years, but connectivity between Yolo Bypass and Sacramento River is greatest in wet years (Sommer et al. 2005). In contrast, the adjacent Sacramento River channel, is a deep and fast-flowing river system with water reaching depths of >5 m and flows as high as ~311 m3/s with little vegetation (Sommer et al. 2001c). This channel is always available for juvenile Chinook Salmon, but provides almost no opportunities for rearing, feeding, and protection from predators (May and Brown 2002; Brown and Bauer 2010). 4 Sampling Morphometric data, DNA samples, and environmental water conditions were obtained from monitoring projects operating within the Sacramento River from both the California Department of Water Resources (CDWR), which operates the bypass Fish Monitoring Program (Pien and Kwan 2022), and the United States Fish and Wildlife Service, which operates the Delta Juvenile Fish Monitoring Program (Mahardja et al. 2019). In the bypass (YBY), sampling occurred during winter and spring by two main methods, a rotary screw trap and beach seines, a program operated by CDWR that started in 1998. The rotary screw trap sits at the base of the toe drain of the bypass and is approximately 2.6 meters in diameter. It was operated and fished 5 to 7 days a week depending on water conditions. Beach seines measuring 7.62 by 1.22 m were towed parallel to the shoreline in 17 spots along the bypass with 10 spots along the toe drain, 3 perennial ponds, and 4 high flow sites. These sampling events occurred once every other week as water conditions allowed (Sommer et al. 2001c; Goertler et al. 2018; Schreier et al. 2018). Samples from the Lower Sacramento River (LSR) were collected by USFWS using three different methods. In the tidal Sacramento River at Sherwood Harbor, a Kodiak trawl was operated from October to March and towed between two boats (Brandes and Mclain 2001; del Rosario et al. 2013). During the months of April and September, a midwater trawl was used and towed with one boat. Sampling by both trawls was at the surface and usually consisted of 10 tows per day, 3 days per week. The second method was beach seines, conducted at two sites downriver from the Fremont Weir entrance to the bypass and adjacent to sampling seines in the bypass (Fig 1.1). 5 Cache Creek Putah Creek Figure 1.1: Map of the sampling region. Black symbols indicate Chinook juvenile sampling locations within the bypass, collected by the DWR. Blue symbols indicate sampling locations collected by the USFWS. Adapted from Goertler et al. 2018a. Juveniles in both regions were then measured for fork length (mm) and assigned to run, across the years of 2013-2019 (Table 1.1). The primary method currently used by many management agencies to assign individuals to run in both systems is “Length at Date” (LAD) method (Fisher 1992). This model uses fork length and date of capture to assign individuals to run. The bypass uses the “Delta” version of the model and the “River model” is employed for identification in the mainstem Sacramento River. The primary difference between these models is based on different algorithms for length-at-date calculation (Fisher 1992; Harvey et al. 2014). 6 Table 1.1: Summary of fish sampled per location and year reported here form all sampling sites mentioned in Figure 1. LSR= Lower Sacramento River and YBY = Yolo Bypass. Year 2013 2014 2015 2016 2017 2018 2019 Sample Numbers YBY n LSR n 60 211 23 199 983 42 453 139 165 67 289 632 249 517 Because we were interested in the ecology of the wild populations in the Central Valley, we excluded all known hatchery fish by excluding fish that lacked an adipose fin. Throughout the Sacramento River System, hatcheries clip the adipose fin of Chinook Salmon juveniles of all runs to signify hatchery origin. Only 25% of fall-run fish raised in hatcheries have their adipose fin clipped, therefore it is possible that some fall-run hatchery origin fish are included in our dataset. However, during the period of our study, hatchery fish were released in the river only during 2013. After that, conditions in the river were so dry that hatchery fish were transported directly to the Delta (a site below our study area) to increase survival, making us confident that no (or very few) hatchery origin fall-run fish were included in our analyses for the other years (Sturrock et al. 2019). Genotyping and run assignment We collected fin clips for genetic analyses from 10 randomly sampled fish per LAD run classification per sampling site per day. Tissue samples were placed in 95% Ethanol and transported back to the lab. We extracted DNA from fins using the DNeasy® Blood and Tissue extraction kit (Qiagen, Valencia, CA). Samples were then genotyped using a Fluidigm Single Nucleotide Polymorphism (SNP) assay of 80 run-type informative markers following the protocol of Meek et al. 2016. This assay was developed using adult Chinook populations of known run throughout the Central Valley. We then assigned samples to run using ONCOR and 7 the baseline described in the previous study (Kalinowski et al. 2008; Meek et al. 2016). We assigned a genetic run to samples with 80% or greater probability of assignment to a particular run, while those below that threshold were assigned as “unknown.” Samples that were “unknown” by genetic methods were not included in the analysis. Statistical analyses To assess the accuracy of LAD identification, we assigned all samples a value of 0 or 1, 1 indicating a match between LAD and genetic run assignment, and 0 indicating mismatch between LAD and genetic run assignment. We then separated samples by run (fall, late fall, spring, and winter) and assessed for mean accuracy employing bootstrap methods using the ‘boot’ function in the R program boot. In this code, means from a random sample of the assigned values were calculated 1000 times. To evaluate if there were differences in proportion of run among individual years in the bypass vs the Sacramento Mainstem, we ran a chi-squared test of independence in R using the “chisq.test” function. We then used the R package “corrplot” to evaluate the residuals in each year and run to determine which values were contributing the most to the overall statistic. Each year is classified to a hydrologic classification based on the Sacramento Valley water year Hydrologic Classification Indices (Whitney 2007; Chronological Reconstructed Sacramento and San Joaquin Valley water year Hydrologic Classification Indices). Within this system there are 5 different classifications: Critical (C), Dry (D), Below Normal (BN), Above Normal (AN), and Wet (W). This metric is determined by taking into account the levels of unimpaired runoff and the previous year’s index (Davis et al. 2000). Next, we analyzed size differences in the bypass vs the Lower Sacramento River Fork Length and Date of fall-run fish by putting these data into a linear regression and using the “glm” function in R. We separated the data by year and location (YBY vs LSR), assuming a normal distribution. We then evaluated these models for statistical differences between years by using a two-sided t-test to compare the difference between the relative slopes. We compared each slope within one year individually and by each location. In both these analyses, we only used fall-run fish to reduce the chance that differences in size were due to life history characteristics present in other runs. Additionally, fall-run was the only population with large enough sample sizes to provide meaningful and statistically sound comparisons. 8 To further evaluate differences between size among juveniles in different hydrological regimes, we compared all mean fork lengths of fall-run fish by water year. We evaluated these means by two statistical methods. To compare between the Yolo Bypass and the Sacramento River, we completed a t-test between each location in each water year. To compare all water years, we ran all samples within a specific location through an ANOVA. To compare what years were contributing the ANOVA statistic, we did further analysis by running a post-hoc Tukey test. RESULTS Concordance between LAD and genetic methods for inferring run type varied by run type and habitat. We found higher concordance between LAD and genetic run assignment in fall-run in both habitats and winter-run in the Lower Sacramento River (Fig. 1.2). Concordance between assignment methods was very low for spring and late fall-runs in both habitats and winter-run in the bypass. During all years of sampling, we classified no juvenile Chinook as late fall by the LAD method in the bypass. In the Sacramento River, we classified a very small number as late fall. However, our genetic assignments show in both systems, there was a non-negligible amount of genetically late fall fish. 9 Figure 1.2: Results from a bootstrap analysis of Length at Date Classification Mismatches, organized by genetic run classification and location. A value closer to 1.0 indicates higher concordance between genetic and length at date classification. LSR= Lower Sacramento River and YBY = Yolo Bypass. 10 Table 1.2: Summary of fish sampled per location and run, and results from the bootstrap analysis depicted in Figure 2 comparing concordance between genetic and LAD classification methods. A value closer to 1.0 in the bootstrap mean column indicates higher concordance between genetic and length at date classification. LSR= Lower Sacramento River and YBY = Yolo Bypass. Sample numbers of fish classified to run in each system (LSR vs YBY) and each method of classification are also reported. Run Fall YBY LSR Late Fall YBY LSR Spring YBY LSR Winter YBY LSR Unknown Sample numbers LAD n Genetic n Boot mean 95% CI 1827 1573 0 16 453 476 36 178 0 1890 1811 96 97 58 120 29 81 530 0.76 0.76 0 0 0.2 0.2 0.31 0.85 - ±0.02 ±0.02 ±0.1 ±0.1 ±0.2 ±0.04 - To ascertain which misclassifications were contributing the most to the lack of concordance between genetic and LAD in these statistics, we compared both methods of classification across all years by plotting fork length versus date of capture (Fig. 1.3). Strikingly, we found that the majority of spring-run misclassifications were genetically fall-run individuals. Most of the genetic fall-run fish above a certain size are reflected in the LAD classifications in the spring graph, leading to spring-run juveniles to be massively overestimated. Alternatively, many genetic late fall-run fish were classified as fall, leading to those juveniles to be largely underestimated. 11 Lower Sacramento River Yolo Bypass Figure 1.3: Comparison of the distribution of run-timing based on classification method across the years 2013-2019. In each plot, the blue and grey colored dots represent the genetic identification, while the orange points show the LAD misidentified individuals. In order, starting with the sites on the Lower Sacramento River we show the a) fall, b) spring, c) late fall, and winter-runs; followed by the e) fall, f) spring, g) late fall, and h) winter-runs in the bypass. 12 We found a significant difference in run proportion between years in both the bypass and Lower Sacramento River sites (Yolo Bypass: X-squared = 128.58, df = 18, p-value < 2.2e-16, Sacramento River: X-squared = 103.86, df = 18, p-value = 4.316e-14). When we calculated the residuals, it was clear that some proportions were contributing more to the chi-square statistic than others (Figure 4). In the years 2013 and 2014, our results showed more spring and winter- run fish in the bypass than expected when compared the proportion of other runs as well as proportions of winter and spring across the years, contributing positively to the chi square statistic. These results were similar to those in the Lower Sacramento River, where only winter- run proportion was higher than expected in 2013 and spring was similarly higher in 2014. Particularly in the later years of the drought (2015-2016), there was a dearth of ESA listed runs (spring and winter) when compared to fall-run. Interestingly, we saw an increase in late fall proportions in the years 2018 in the Sacramento River and 2019 in the bypass (Figure 5). Yolo Bypass a) c) Lower Sacramento River b) d) Figure 1.4: Length at Date and Genetic run proportions within the bypass and Sacramento River. Length at Date proportion of each run within year in the (a) Yolo Bypass and (b) the Lower Sacramento mainstem. Genetic proportion of each run within year in the (c) Yolo Bypass and (d) the Lower Sacramento mainstem. Residuals from a chi-square test for independence in the (e) Yolo Bypass and (f) Lower Sacramento mainstem indicating significance in that particular cell is compared to all other cells. Larger dots indicate a higher contribution to the chi-square statistic, while blue dots indicate a positive contribution and red dots indicate a negative contribution. 13 Figure 1.4 (cont’d) e) f) Figure 1.5: Mean fork length (mm) of genetically assigned fall-run juvenile Chinook in the Lower Sacramento River (LSR) vs the bypass (YBY), organized by water year. Comparisons for all years were statistically significant, as indicated by the p-values. 14 Table 1.3: Results from the posthoc Tukey test, comparing Fork Length of fall-run juveniles within each year in the a) Yolo Bypass and the b) Lower Sacramento River. Each Tukey group represents a significantly different mean fork length as it contributes to the overall significant difference in the ANOVA. Groups with the same letter are similar to each other, while groups with different letters are significantly different from each other. a) b) Water year Mean Fork Length Tukey group Water year Type Fall-run – Yolo Bypass 2013 2014 2015 2016 2017 2018 2019 67.92857 59.62445 63.8333 51.78302 51.33795 62.82222 77.94915 b c bc d d bc a Drought Critical Critical Below Normal Wet Below Normal Wet Fall-run – Lower Sacramento River Water year Mean Fork Length Tukey group Water year Type 2013 2014 2015 2016 2017 2018 2019 49.19048 49.60440 46.02941 63.64217 65.16715 54.96350 62.6407 bc bc c a a b a Drought Critical Critical Below Normal Wet Below Normal Wet We found different fork lengths in fall-run between habitat types across all years, with the bypass having significantly larger fall-run fish in every year except 2016 and 2017, where the means were significantly smaller (Table 1.3). We also found a significant difference among years within habitat type (Table 1.4). We found that in both habitats in 2019, both habitats had a 15 larger mean fork length as compared with other years. In the Lower Sacramento River in years classified as Wet or Below Normal, we found fish attained greater size. This same pattern is not reflected in the bypass, where the smallest mean fork lengths were in the years of 2016 and 2017. To further explore the effect habitat has on the size distribution of juvenile fall-run Chinook in the bypass, we evaluated the fork lengths of fish over time in the bypass as compared to the Lower Sacramento River. When we used the linear regression based on year and location, the slope of fall-run fish fork lengths over time from the bypass were significantly different from that in the Sacramento mainstem in the years 2016, 2017, and 2019 (Table 4, Fig. 1.6), with sizes increasing more over time in the bypass. We found that the slopes of fork lengths over time were not significantly different between habitat types for 2013-2015, the three Drought and Critical years (Table 1.4, Table S1.1, Figure S1.1-7). Table 1.4: Changes in fork length over time between habitat types as input in a linear regression. Here we show the results from comparison of slopes of both sampling regions as input in a glm model. The coefficient indicates the level of interaction between the location and sample date, where the bypass is the point of reference. An asterisk indicates a significantly larger slope between locations. Water year Coefficient 2013 2014 2015 2016 2017 2018 2019 0.1072 0.0125 0.0871 0.1644 0.0452 0.1082 0.2982 Fall-run P value 0.0892 0.787 0.634 0.0012* 0.0188* 0.496 2.48-11 * Water year Type Drought Critical Critical Below Normal Wet Below Normal Wet 16 Figure 1.6: Changes in size of fall-run fish over time in all years classified as Wet (blue) and Drought (orange)with 95% confidence intervals (gray). Here we show the comparison between wet and dry years in the bypass (a) versus wet and dry years in the Lower Sacramento River (b). In both locations you can see there is a significant difference between wet and dry years, with the bypass experiencing a larger change in slope across time. 17 DISCUSSION Habitat diversity is essential for supporting diverse juvenile salmon populations, particularly as climate change creates large fluctuations in environmental conditions (Beechie et al. 2013; Herbold et al. 2018). The diversity of migration timings, natal homing strategies, and outmigration tactics provides population buffering under differing environmental conditions, such as differing water levels and temperatures (Hilborn et al. 2003; Greene et al. 2009; Schindler et al. 2010). It is therefore imperative to monitor and manage life history diversity accurately and how that diversity is impacted by varying habitat conditions. Our study found that genetics methods would be most effective, and that floodplain habitat in the Central Valley is vital for supporting diverse Chinook Salmon populations. In particular, the diverse habitat provided by the bypass supports all runs as well as a diversity in fish size. LAD versus Genetic Methods Because monitoring life history diversity requires accurate classification of this diversity, we first compared classification methods in the system. LAD methods are still widely being used as the main method of classification, so we sought to understand how that compared with genetic methods. We found a large mismatch between the length at date and genetic run assignments, indicating a lack of accuracy in the LAD model. Our work shows that overwhelmingly, length at date metrics are overestimating the occurrence of spring-run and underestimating fall, late fall, and winter-run (Fig. 1.3). This is an important issue because it means that the LAD approach does not present an accurate picture of how many threatened and endangered juvenile fish are in the system. For example, this has contributed to a knowledge gap around the abundance of threatened spring run in the system (Nelson et al. 2022). For endangered winter-run, the export of water by state and federal pumping operations is directly tied to how many fish are in the system (NOAA NMFS 2009; Harvey et al. 2014). These management decisions rely heavily on quantifying the exact number of winter-run juvenile fish in the system to mitigate negative impacts. Therefore, accurate classification methods are vital to enabling the protection of the listed run and sustainable use of the water resource (Brown et al. 2009; Brown and Bauer 2010; Stewardship Council Delta Science Program 2019). 18 Chinook Salmon Life History Diversity Our data indicate that all four runs and a diversity of sizes are present in the bypass. The is likely because the bypass provides several benefits to juvenile Chinook, including increased food resources and protection from predators (Jeffres et al. 2008; Limm and Marchetti 2009). Unfortunately, during low water periods, especially years that experience drought conditions, our work shows the proportion of spring, late fall, and winter-run decreased over time, suggesting that life history diversity is compromised when the bypass is more difficult to access (Fig. 1.4). Spring and winter-run did not begin to increase proportionally until 2019, when flooding occurred. Proportions of runs other than fall were similarly negatively affected in the Lower Sacramento River. These data indicate that maintaining higher flows help support life history diversity in juvenile Chinook Salmon, regardless of habitat. This is consistent with studies showing that spring-run phenotypes lose habitat with increased drought periods and have decreased survival rates (Cordoleani et al. 2021; Notch et al.). Previous work has shown the importance of habitat diversity at all life stages for growth and fitness of many Pacific salmon species (Healey 1991). In the Sacramento River, there is evidence that wetland and managed floodplain habitats provide important opportunities for juvenile salmon growth, and different habitats have different food resources that impact growth (Jeffres et al. 2008; Cordoleani et al. 2022). Size and growth in the early life stages is imperative for success in the marine environment and entering the ocean at a larger size can make the difference between survival or death, so it is important to maintain habitats that provide opportunities for growth (Beamish and Mahnken 2001; Woodson et al. 2013). We found that there is a significant difference in juvenile fall-run Chinook fork length every year between the bypass and Sacramento River, as well as across years in both systems during different hydrological conditions (Fig. 1.5). In particular, fall-run juveniles were significantly bigger in the bypass almost every year. This suggests that the bypass is an important habitat for juvenile Chinook that allows growth and refuge, even during drought conditions. Because the floodplain conditions have shown to be important for growth, particularly periods of wet conditions or inundation, managing the bypass with an aim of promoting this life history and run diversity will be imperative for the persistence of these populations. Although fish tended to be larger in Yolo Bypass than the Sacramento River although there are two notable exceptions. For example, fish were relatively smaller in two divergent 19 water year types, 2015 (extremely dry) and 2017 (extremely wet). This is likely because fish size is an imperfect metric of growth since the target habitats are open to immigration and emigration. This is because fish size in either Yolo Bypass or Sacramento River is a complex function of not only regional growth, but also influx of new individuals from upstream areas. Fish tagging or otolith measurements would have been a better tool to better characterize growth patterns but was not within the scope of this study. There is ample evidence from tagging methods that growth is consistently faster in Yolo Bypass (Sommer et al. 2001c; Katz et al. 2017; Takata et al. 2017) In addition, our research indicates during wet conditions the bypass supports smaller fish earlier in the season while sustaining higher growth rates than the adjacent Sacramento River later in the season. This suggests that the bypass proffers benefits for size diversity during wet conditions (Table 1.4 and Fig. 1.6). The greatest fold difference slope of fork length over time between the two locations was during 2019, which had many weir-overtopping events (51 days) as compared to years that had less (in comparison 2014 had 0 days and 2015 had 3 days) (California Department of Water Resources). When all data points are combined from wet and dry years, there is still a significant pattern of increased growth in the bypass over time (Fig. 1.6). This aligns with other research that has shown increased rearing opportunities in the bypass leads to increased variation in size and growth rates in juvenile Chinook Salmon (Goertler et al. 2016; Goertler et al. 2018) MANAGEMENT IMPLICATIONS The Central Valley is predicted to be one of the areas most impacted by drought in the world (Famiglietti 2014). As the climate warms and intensifies, this is predicted to cause longer, more intense, and frequent droughts in the Central Valley, and more intense flood years (Gershunov et al. 2013; Trenberth et al. 2013; Swain et al. 2016). During more intense drought periods, young juvenile Chinook Salmon may have reduced access to off-channel habitat. Our work shows the diversity of habitats is essential to preserving a diversity of run-types and size distributions in juvenile Chinook Salmon. Consequently, reduced habitat variation could lead to a loss diversity of juvenile Chinook Salmon, which could weaken the portfolio effect and long- term stability and resilience. For this reason, managing connectivity between the bypass and Sacramento River represents a potentially valuable tool to sustain Chinook Salmon populations. For example, a major habitat restoration project is underway that will allow Yolo Bypass to be 20 inundated at lower flows (USBR and CDWR 2019). Similarly, several large-scale tidal wetlands projects are being constructed in lower Yolo Bypass, which could improve access to the floodplain and habitat quality during dry years and seasons (CDWR 2021; CDWR). Another important finding from our study is that the LAD method is relatively inaccurate at identifying the full range of salmon runs. For example, the LAD models overestimate spring- run fish because large fall-run fish in the system are being misclassified as spring-run. This issue is already relatively well-recognized (Harvey et al. 2014; Brandes et al. 2021; Nelson et al. 2022), so genetics is increasingly being added as a monitoring and management tool in the Sacramento watershed and downstream estuary. Hence, we strongly support expanded use of genetic methods to monitor and manage Chinook Salmon in the system more accurately. 21 BIBLIOGRAPHY Beamish RJ, Mahnken C. 2001. A critical size and period hypothesis to explain natural regulation of salmon abundance and the linkage to climate and climate change. Prog Oceanogr. [accessed 2022 May 12];49:423–437. https://doi.org/10.1016/S0079- 6611(01)00034-9 Beechie T, Imaki H, Greene J, Wade A, Wu H, Pess G, Roni P, Kimball J, Stanford J, Kiffney P, Mantua N. 2013. RESTORING SALMON HABITAT FOR A CHANGING CLIMATE. River Res Appl. [accessed 2022 October 19];29:939–960. https://doi.org/10.1002/RRA.2590 Brandes PL, Mclain JS. 2001. Juvenile Chinook Salmon Abundance, Distribution, and Survival in the Sacramento-San Joaquin Estuary. Contributions to the Biology of Central Valley Salmonids. 2:39–136. Brandes PL, Pyper B, Banks M, Jacobson D, Garrison T, Cramer S. 2021. Comparison of Length-at-Date Criteria and Genetic Run Assignments for Juvenile Chinook Salmon Caught at Sacramento and Chipps Island in the Sacramento-San Joaquin Delta of California. San Francisco Estuary and Watershed Science. [accessed 2023 September 19];19:1–15. https://doi.org/10.15447/SFEWS.2021V19ISS3ART2 Brown LR, Bauer ML. 2010. Effects of hydrologic infrastructure on flow regimes of California’s Central Valley rivers: Implications for fish populations. River Res Appl. [accessed 2023 February 19];26:751–765. https://doi.org/10.1002/RRA.1293 Brown LR, Kimmerer W, Brown R. 2009. Managing water to protect fish: A review of California’s environmental water account, 2001-2005. Environ Manage. [accessed 2023 February 19];43:357–368. https://doi.org/10.1007/S00267-008-9213-4/FULLTEXT.HTML Brucet S, Boix D, Nathansen LW, Quintana XD, Jensen E, Balayla D, Meerhoff M, Jeppesen E. 2012. Effects of temperature, salinity and fish in structuring the macroinvertebrate community in shallow lakes: Implications for effects of climate change. PLoS One. https://doi.org/10.1371/journal.pone.0030877 California Department of Water Resources. California Data Exchange Center. In: Query Tools. Available from: https://cdec.water.ca.gov/dynamicapp/staMeta?station_id=FRE. [accessed 2022 Mar 30]. Carlson SM, Satterthwaite WH. 2011. Weakened portfolio effect in a collapsed salmon population complex. Canadian Journal of Fisheries and Aquatic Sciences. 68:1579–1589. https://doi.org/10.1139/f2011-084 CDWR. March 5, 2021. Multi-Agency Collaboration Restores Critical Habitat for Endangered Delta Smelt, Other Native Species. Available from: https://water.ca.gov/News/Blog/2021/March/Lower-Yolo-Ranch-Tidal-Restoration-Project. [accessed 2023 May 14]. 22 CDWR. DWR Certifies Final EIR for Delta’s Largest Tidal Habitat Restoration Project. Available from: https://water.ca.gov/News/Blog/2020/Nov-2020/DWR-Certifies-Final-EIR- for-Largest-Tidal-Habitat-Restoration-Project. [accessed 2023 May 14]. Chronological Reconstructed Sacramento and San Joaquin Valley Water Year Hydrologic Classification Indices. Available from: https://cdec.water.ca.gov/reportapp/javareports?name=WSIHIST. [accessed 2023 Feb 19]. Cordoleani F, Holmes E, Bell-Tilcock M, Johnson RC, Jeffres C. 2022. Variability in foodscapes and fish growth across a habitat mosaic: Implications for management and ecosystem restoration. Ecol Indic. [accessed 2022 May 21];136:108681. https://doi.org/10.1016/J.ECOLIND.2022.108681 Cordoleani F, Phillis CC, Sturrock AM, FitzGerald AM, Malkassian A, Whitman GE, Weber PK, Johnson RC. 2021. Threatened salmon rely on a rare life history strategy in a warming landscape. Nat Clim Chang. 11:982–988. https://doi.org/10.1038/s41558-021-01186-4 Crozier LG, Hendry AP, Lawson PW, Quinn TP, Mantua NJ, Battin J, Shaw RG, Huey RB. 2008. PERSPECTIVE: Potential responses to climate change in organisms with complex life histories: evolution and plasticity in Pacific salmon. Evol Appl. https://doi.org/10.1111/j.1752-4571.2008.00033.x Davis G, Stubchaer J, Forster MJ, Chair V, Brown J, Baggett AG, Pettit W. 2000. Water Right Decision 1641. 1–225 p. Available from: http://www.swrcb.ca.gov/ del Rosario RB, Redler YJ, Newman K, Brandes PL, Sommer T, Reece K, Vincik R. 2013. Migration patterns of Juvenile Winter-run-sized Chinook salmon (Oncorhynchus tshawytscha) through the Sacramento-San Joaquin delta. San Francisco Estuary and Watershed Science. 11:. https://doi.org/10.15447/sfews.2013v11iss1art3 Famiglietti JS. 2014. The global groundwater crisis. Nat Clim Chang. 4:945–948. https://doi.org/10.1038/nclimate2425 Famiglietti JS, Lo M, Ho SL, Bethune J, Anderson KJ, Syed TH, Swenson SC, de Linage CR, Rodell M. 2011. Satellites measure recent rates of groundwater depletion in California’s Central Valley. Geophys Res Lett. 38:. https://doi.org/10.1029/2010GL046442 Ficke AD, Myrick CA, Hansen LJ. 2007. Potential impacts of global climate change on freshwater fisheries. Rev Fish Biol Fish. https://doi.org/10.1007/s11160-007-9059-5 Ficklin DL, Barnhart BL, Knouft JH, Stewart IT, Maurer EP, Letsinger SL, Whittaker GW. 2014. Climate change and stream temperature projections in the Columbia River basin: Habitat implications of spatial variation in hydrologic drivers. Hydrol Earth Syst Sci. https://doi.org/10.5194/hess-18-4897-2014 Fisher FW. 1992. Chinook salmon, Oncorhynchus tshawytscha, growth and occurrence in the Sacramento-San Joaquin river system. California Department of Fish and Game, Inland Fisheries Division, Red Bluff, California. 23 Gershunov A, Rajagopalan B, Overpeck J, Guirguis K, Cayan D, Hughes M, Dettinger M, Castro C, Schwartz RE, Anderson M, Ray AJ, Barsugli J, Cavazos T, Alexander M, Dominguez F. 2013. Future climate: Projected extremes. Assessment of Climate Change in the Southwest United States: A Report Prepared for the National Climate Assessment. [accessed 2022 December 11];126–147. https://doi.org/10.5822/978-1-61091-484- 0_7/FIGURES/5 Goertler PAL, Simenstad CA, Bottom DL, Hinton S, Stamatiou L. 2016. Estuarine Habitat and Demographic Factors Affect Juvenile Chinook (Oncorhynchus tshawytscha) Growth Variability in a Large Freshwater Tidal Estuary. Estuaries and Coasts. [accessed 2021 July 21];39:542–559. https://doi.org/10.1007/s12237-015-0002-z Goertler PAL, Sommer TR, Satterthwaite WH, Schreier BM. 2018. Seasonal floodplain-tidal slough complex supports size variation for juvenile Chinook salmon (Oncorhynchus tshawytscha). Ecol Freshw Fish. 27:580–593. https://doi.org/10.1111/eff.12372 Greene CM, Hall JE, Guilbault KR, Quinn TP. 2009. Improved viability of populations with diverse life-history portfolios. Biol Lett. 6:382–386. Harvey B. 2011. Length-at-date criteria to classify juvenile Chinook Salmon in the California Central Valley: development and implementation history. In: Interagency Ecological Program Newsletter. Available from: https://cadwr.app.box.com/v/InteragencyEcologicalProgram/file/571036622021. [accessed 2023 Feb 19]. Harvey BN, Jacobson DP, Banks MA. 2014. Quantifying the Uncertainty of a Juvenile Chinook Salmon Race Identification Method for a Mixed-Race Stock. N Am J Fish Manag. https://doi.org/10.1080/02755947.2014.951804 Healey MC. 1991. Life history of Chinook Salmon (Oncorhynchus tshawytscha). Hellmair M, Peterson M, Mulvey B, Young K, Montgomery J, Fuller A. 2018. Physical Characteristics Influencing Nearshore Habitat Use by Juvenile Chinook Salmon in the Sacramento River, California. N Am J Fish Manag. 38:959–970. https://doi.org/10.1002/nafm.10201 Herbold B, Carlson SM, Henery R, Johnson RC, Mantua N, McClure M, Moyle P, Sommer T. 2018. Managing for salmon resilience in California’s variable and changing climate. San Francisco Estuary and Watershed Science. [accessed 2022 October 19];16:. https://doi.org/10.15447/SFEWS.2018V16ISS2ART3 Hilborn R, Quinn TP, Schindler DE, Rogers DE. 2003. Biocomplexity and fisheries sustainability. Proceedings of the National Academy of Sciences. 100:6564–6568. https://doi.org/10.1073/pnas.1037274100 James LA, Singer MB. 2008. Development of the lower sacramento valley flood-control system: Historical perspective. Nat Hazards Rev. https://doi.org/10.1061/(ASCE)1527- 6988(2008)9:3(125) 24 Jeffres CA, Opperman JJ, Moyle PB. 2008. Ephemeral floodplain habitats provide best growth conditions for juvenile Chinook salmon in a California river. Environmental Biology of Fishes 2008 83:4. [accessed 2022 May 12];83:449–458. https://doi.org/10.1007/S10641- 008-9367-1 Kalinowski S, Manlove K, Taper M. 2008. ONCOR: Software for genetic stock identification. Katz J, Jeffres C, Conrad L, Sommer T, Corline N, Martinez J, Brumbaugh S, Takata L, Ikemiyagi, Naoaki; Kiernan J, Moyle P. 2013. The Experimental Agricultural Floodplain Habitat Investigation at Knaggs Ranch on Yolo Bypass 2012-2013. 78 p. Katz JVE, Jeffres C, Conrad JL, Sommer TR, Martinez J, Brumbaugh S, Corline N, Moyle PB. 2017. Floodplain farm fields provide novel rearing habitat for Chinook salmon. PLoS One. [accessed 2023 February 19];12:e0177409. https://doi.org/10.1371/JOURNAL.PONE.0177409 Limm MP, Marchetti MP. 2009. Juvenile Chinook salmon (Oncorhynchus tshawytscha) growth in off-channel and main-channel habitats on the Sacramento River, CA using otolith increment widths. Environ Biol Fishes. [accessed 2022 May 14];85:141–151. https://doi.org/10.1007/S10641-009-9473-8 MacArthur R. 1955. Fluctuations of Animal Populations and a Measure of Community Stability. Ecology. https://doi.org/10.2307/1929601 Mahardja B, Speegle J, Nanninga A, Barnard D. June 7, 2019. Interagency Ecological Program: Over four decades of juvenile fish monitoring data from the San Francisco Estuary, collected by the Delta Juvenile Fish Monitoring Program, 1976-2018. In: Interagency Ecological Program (IEP). Available from: https://portal.edirepository.org/nis/mapbrowse?scope=edi&identifier=244&revision=3. [accessed 2023 May 14]. May JT, Brown LR. 2002. Fish communities of the Sacramento River Basin: implications for conservation of native fishes in the Central Valley, California. 373–388 p. Meek MH, Baerwald MR, Stephens MR, Goodbla A, Miller MR, Tomalty KMH, May B. 2016. Sequencing improves our ability to study threatened migratory species: Genetic population assignment in California’s Central Valley Chinook salmon. Ecol Evol. 6:7706–7716. https://doi.org/10.1002/ece3.2493 Meek MH, Stephens MR, Wong AK, Tomalty KM, May B, R. BM. 2014. Genetic characterization of California’s Central Valley chinook salmon. Ecology. 95:1431. Michel CJ, Ammann AJ, Lindley ST, Sandstrom PT, Eric D, Thomas MJ, Singer GP, Klimley AP, Macfarlane RB, Michel CJ, Division FE, Fisheries S, Cruz S, Cruz S, Marine N, Service F, Oceanic N, Cruz S, Fisheries S, Marine N, Service F, Oceanic N, Cruz S. 2015. Chinook salmon outmigration survival in wet and dry years in California’ s Sacramento River. Canadian Journal of Fisheries and Aquatic Sciences. 1759:1749–1759. https://doi.org/10.1139/cjfas-2014-0528 25 National Marine Fisheries Service. 2014. Recovery Plan for the Evolutionarily Significant Units of Sacramento River Winter-Run Chinook Salmon and Central Valley Spring-Run Chinook Salmon and the Distinct Population Segment of Central Valley Steelhead. Nelson PA, Baerwald M, Burgess O, Bush E, Collins A, Cordoleani F, DeBey H, Gille D, Goertler PAL, Harvey B, Johnson RC, Kindopp J, Meyers E, Notch J, Phillis CC, Singer G, Sommer T. 2022. Considerations for the Development of a Juvenile Production Estimate for Central Valley Spring-Run Chinook Salmon. San Francisco Estuary and Watershed Science. [accessed 2022 October 27];20:. https://doi.org/10.15447/SFEWS.2022V20ISS2ART2 NOAA NMFS. 2009. Biological opinion and conference opinion on the long-term operations of the Central Valley Project and State Water Project. Notch JJ, Mchuron AS, Michel CJ, Cordoleani F, Johnson M, Henderson MJ, Ammann AJ. Outmigration survival of wild Chinook salmon smolts through the Sacramento River during historic drought and high water conditions. [accessed 2022 December 08]. https://doi.org/10.1007/s10641-020-00952-1 Pien C, Kwan N. 2022. Interagency Ecological Program: Fish catch and water quality data from the Sacramento River floodplain and tidal slough, collected by the Yolo Bypass Fish Monitoring Program, 1998-2018. - Datasets - California Open Data. In: Interagency Ecological Program . Available from: https://data.ca.gov/dataset/interagency-ecological- program-fish-catch-and-water-quality-data-from-the-sacramento-1998-2018. [accessed 2023 May 14]. Salcido RE. 2012. The Success and Continued Challenges of the Yolo Bypass Wildlife Area: A Grassroots Restoration. Ecol Law Q. 39:1085–1134. https://doi.org/10.15779/Z38B541 Schindler DE, Hilborn R, Chasco B, Boatright CP, Quinn TP, Rogers LA, Webster MS. 2010. Population diversity and the portfolio effect in an exploited species. Nature. 465:609–612. https://doi.org/10.1038/nature09060 Schreier B, Davis B, Ikemiyagi N, Sommer T, Conrad L, Takata L, Aha N, Bedwell M, Goertler P. August 14, 2018. Interagency Ecological Program: Fish catch and water quality data from the Sacramento River floodplain and tidal slough, collected by the Yolo Bypass Fish Monitoring Program, 1998-2018. Available from: https://portal.edirepository.org/nis/metadataviewer?packageid=edi.233.1. https://doi.org/10.6073/pasta/0ab359bec7b752c1f68621f5e1768eb0 Sommer T, Harrell B, Nobriga M, Brown R, Moyle P, Kimmerer W, Schemel L. 2001a. California’s Yolo Bypass: Evidence that flood control Can Be compatible with fisheries, wetlands, wildlife, and agriculture. Fisheries (Bethesda). 26:6–16. https://doi.org/10.1577/1548-8446(2001)026<0006:CYB>2.0.CO;2 Sommer T, Harrell B, Nobriga M, Brown R, Moyle P, Kimmerer W, Schemel L. 2001b. California’s Yolo Bypass: Evidence that flood control Can Be compatible with fisheries, 26 wetlands, wildlife, and agriculture. Fisheries (Bethesda). 26:6–16. https://doi.org/10.1577/1548-8446(2001)026<0006:CYB>2.0.CO;2 Sommer TR, Harrell WC, Nobriga ML. 2005. Habitat Use and Stranding Risk of Juvenile Chinook Salmon on a Seasonal Floodplain. N Am J Fish Manag. 25:1493–1504. https://doi.org/10.1577/M04-208.1 Sommer TR, Nobriga ML, Harrell WC, Batham W, Kimmerer WJ. 2001c. Floodplain rearing of juvenile chinook salmon: evidence of enhanced growth and survival. Canadian Journal of Fisheries & Aquatic Sciences. 58:325–333. https://doi.org/10.1139/cjfas-58-2-325 Stewardship Council Delta Science Program D. 2019. Delta Science Plan, June 2019 Report. [accessed 2023 February 19]. Sturrock AM, Satterthwaite WH, Cervantes-Yoshida KM, Huber ER, Sturrock HJW, Nusslé S, Carlson SM. 2019. Eight Decades of Hatchery Salmon Releases in the California Central Valley: Factors Influencing Straying and Resilience. [accessed 2021 July 21]. https://doi.org/10.1002/fsh.10267 Sturrock AM, Wikert JD, Heyne T, Mesick C, Hubbard AE, Hinkelman TM, Weber PK, Whitman GE, Glessner JJ, Johnson RC. 2015. Reconstructing the migratory behavior and long-term survivorship of juvenile Chinook salmon under contrasting hydrologic regimes. PLoS One. 10:1–23. https://doi.org/10.1371/journal.pone.0122380 Swain DL, Horton DE, Singh D, Diffenbaugh NS. 2016. Trends in atmospheric patterns conducive to seasonal precipitation and temperature extremes in California. Sci Adv. [accessed 2022 December 11];2:. https://doi.org/10.1126/SCIADV.1501344/SUPPL_FILE/1501344_SM.PDF Takata L, Sommer TR, Louise Conrad J, Schreier BM. 2017. Rearing and migration of juvenile Chinook salmon (Oncorhynchus tshawytscha) in a large river floodplain. Environ Biol Fishes. [accessed 2023 February 19];100:1105–1120. https://doi.org/10.1007/S10641-017- 0631-0/TABLES/4 Trenberth KE, Dai A, Van Der Schrier G, Jones PD, Barichivich J, Briffa KR, Sheffield J. 2013. Global warming and changes in drought. Nature Climate Change 2014 4:1. [accessed 2022 December 11];4:17–22. https://doi.org/10.1038/nclimate2067 USBR, CDWR. 2019. Yolo Bypass Salmonid Habitat Restoration and Fish Passage Project. Available from: https://www.usbr.gov/mp/nepa/nepa_project_details.php?Project_ID=30484. [accessed 2023 May 14]. van Dyke E, Wasson K. 2005. Historical ecology of a central california estuary: 150 Years of habitat change. Estuaries. 28:173–189. https://doi.org/10.1007/BF02732853 Whitney V. 2007. Sacramento Valley Water Year Hydrologic Classification. [accessed 2023 Feb 19] 27 Woodson LE, Wells BK, Weber PK, MacFarlane RB, Whitman GE, Johnson RC. 2013. Size, growth, and origin-dependent mortality of juvenile Chinook salmon Oncorhynchus tshawytscha during early ocean residence. Mar Ecol Prog Ser. [accessed 2022 May 12];487:163–175. https://doi.org/10.3354/MEPS10353 Woodward G, Perkins DM, Brown LE. 2010. Climate change and freshwater ecosystems: Impacts across multiple levels of organization. Philosophical Transactions of the Royal Society B: Biological Sciences 365. https://doi.org/10.1098/rstb.2010.0055 Xiao M, Koppa A, Mekonnen Z, Pagán BR, Zhan S, Cao Q, Aierken A, Lee H, Lettenmaier DP. 2017. How much groundwater did California’s Central Valley lose during the 2012–2016 drought? Geophys Res Lett. 44:4872–4879. https://doi.org/10.1002/2017GL073333 28 APPENDIX 1A: SUPPLEMENTAL TABLE AND FIGURES Table S1.1: Results from comparison of slopes of both sampling regions as input in a glm model. The coefficient indicates the level of interaction between the location and sample date, where the Yolo Bypass is the point of reference. An asterisk indicates a significantly larger slope between locations. Fall Run Water Year Coefficient 2013 2014 2015 2016 2017 2018 2019 0.1072 0.0125 0.0871 0.1644 0.0452 0.1082 0.2982 P value 0.0892 0.787 0.634 0.0012* 0.0188* 0.496 2.48-11 * Water Year Type Drought Critical Critical Below Normal Wet Below Normal Wet Figure S1.1: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2013 with 95% confidence intervals (gray). Significance is indicated by a start symbol. 29 Figure S1.2: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2014 with 95% confidence intervals (gray). Significance is indicated by a start symbol. Figure S1.3: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2015 with 95% confidence intervals (gray). Significance is indicated by a star symbol. 30 Figure S1.4: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2016 with 95% confidence intervals (gray). Significance is indicated by a star symbol. * Figure S1.5: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2017 with 95% confidence intervals (gray). Significance is indicated by a start symbol. * 31 Figure S1.6: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2018 with 95% confidence intervals (gray). Significance is indicated by a start symbol. Figure S1.7: Graph comparing fork lengths of fall run fish between the Yolo Bypass (blue) and Lower Sacramento River mainstem (red) with a linear model generated for the year 2019 with 95% confidence intervals (gray). Significance is indicated by a start symbol. * 32 CHAPTER 2: REMNANT SALMON LIFE HISTORY DIVERSITY REDISCOVERED IN A HIGHLY COMPRESSED HABITAT Chapter 2: This chapter has been submitted for publication to Evolutionary Applications and is currently under peer-review. Other contributing authors: Anna Sturrock, Malte Willmes, Tasha Thompson, Rachel Johnson, Flora Cordoleani, Natalie Stauffer-Olsen, George Whitman, Mariah Meek ABSTRACT Chinook salmon (Oncorhynchus tshawystcha) display remarkable life history diversity underpinning their ability to adapt to environmental change. Maintaining life history diversity is vital to the resilience and stability of Chinook metapopulations, particularly under rapidly changing climates. However, the conditions that promote life history diversity are rapidly disappearing, as anthropogenic forces promote homogenization of habitats and genetic lineages. In this study, we use the highly modified Yuba River in the Central Valley of California to understand if distinct genetic lineages and life history still exist, despite reductions in spawning habitat and hatchery practices that have promoted introgression. There currently is a concerted effort to protect federally listed spring run populations, given that few wild populations still exist. Despite this, we lack a comprehensive understanding of the genetic and life history diversity of Chinook salmon present in the Yuba River system. To understand if this diversity still exists, we collected migration timing data from acoustic tagging and carcass surveys and GREB1L genotypes from Chinook salmon in the Yuba River between 2009-2011. Variation in the GREB1L region of the genome is tightly linked with run timing in Chinook salmon but the relationship between this variation and entry on spawning grounds is little explored in the Central Valley. We found that the date Chinook salmon crossed the lowest barrier to spawning habitat (Daguerre Point Dam) was tightly correlated with their GREB1L genotype. Importantly, our study confirms that ESA-listed spring run Chinook salmon are spawning in the Yuba River, promoting a portfolio of life history and genetic diversity, despite spawning in a compressed habitat. This work highlights the need to identify and protect this life history diversity in heavily impacted systems to conserve and promote diverse and healthy Chinook salmon metapopulations. Without this, we run the risk of losing the last vestiges of important variation. KEYWORDS Life history diversity, GREB1L, acoustic tagging 33 INTRODUCTION Life history diversity is critical for species to respond to environmental variability (Beechie et al., 2006; Moore et al., 2014). This diversity often includes differences in morphology, size, and age at maturity and is often influenced both by environmental and genetic factors (Healey, 1991; Thibaut and Connolly, 2013). In particular, genetic diversity is important as it often harbors the adaptive potential for populations to respond to future or changing conditions (Brooks et al., 2006; Chapin et al., 2000). Additionally, genetic diversity within a species or population can result in the expression of diverse life history strategies that spread survival risk across time and space, stabilizing populations and ecosystem services. This phenomenon is referred to as biocomplexity (Hilborn et al., 2003) and can help buffer the effects of natural and anthropogenic change (Narum et al., 2018). Unfortunately, biocomplexity, and in turn genetic diversity, is being lost at alarming rates due to anthropogenic change, particularly in freshwater ecosystems (Allendorf et al., 2014; Des Roches et al., 2021; Heino et al., 2009; Sih et al., 2000). To protect biocomplexity and promote life history diversity, it is vital to identify, monitor, and protect unique phenotypic and genetic traits present within and among populations. In general, salmonids in the United States have been losing biocomplexity over the last century due to anthropogenic stressors (Dittman and Quinn, 1996; Finney et al., 2002; Malick and Cox, 2016). For example, Chinook salmon (Oncorhynchus tshawytscha) have faced declines in excess of 99% of their original population sizes in their native range due to overfishing, damming, mining, and climate change (Mahnken et al., 1998; National Marine Fisheries Service, 2014). This is particularly troubling because Chinook salmon are a keystone species of high cultural, economic, and ecological value (Bottom et al., 2009; Colombi, 2012; Layman et al., 2006). With large population losses, many Chinook salmon populations have also experienced marked reduction in genetic diversity (Johnson et al., 2018; Thompson et al., 2019; Weeder et al., 2005). These significant losses in genetic diversity have had negative consequences in terms of reduction in phenotypic diversity and adaptive capacity (Carlson and Satterthwaite, 2011; Griffiths et al., 2014). Thus, it is vital that we identify and protect the remaining biocomplexity found in Chinook salmon populations to promote population persistence and resilience in an anthropogenically influenced system. The California Central Valley (CCV) is the southernmost portion of the native Chinook salmon range and populations are greatly imperiled due to the negative impact of anthropogenic 34 stressors such as dams, historic mining operations, and extensive urbanization (Herbold et al., 2018; Moyle et al., 2017). Due to its southern location, Chinook salmon populations in the CCV are also highly vulnerable to climate change (Crozier et al., 2019). Despite these threats, the Sacramento River is the only part of the entire species’ range that contains four distinct spawning life history timings while all other systems have only two distinct run timings. This makes the Chinook salmon in the CCV a uniquely diverse population complex (Williams, 2006),. These life history phenotypes are referred to as “run-types” and are named after the season by which they migrate upriver to spawn (fall, late fall, spring, and winter). Historical temporal and spatial separation have resulted in limited gene flow among CCV runs within the same river system, leading to these populations becoming genetically distinct (Meek et al., 2020). This genetic variation provides the adaptive capacity necessary to result in phenotypically diverse populations. This biocomplexity in run-types is essential in maintaining Chinook salmon stock abundance across years, facilitating a “portfolio effect” that allows the species to withstand environmental heterogeneity and perturbations (Schindler et al., 2010). Although we know much about the biology of Chinook salmon, much is still unknown about the heritability or genetic basis of life history traits of Central Valley populations (Cordoleani et al., 2020). Spring run Chinook salmon were once the most abundant run in the CCV, existing in the hundreds of thousands prior to the construction of impassable dams, extensive levees that converted floodplain and marsh habitat to agriculture, and overfishing (Lindley et al., 2004; Yoshiyama et al., 1998). Spring run fish display a unique spawning strategy of migrating into the system early when water temperatures are low from high spring flows and oversummering in cool headwaters before spawning in the fall (Quinn et al., 2016). Unfortunately, dam construction in the CCV, which began in the early 1900s, cut off access to historical spring run spawning habitat for most populations throughout the CCV. This forced spring run to face the double threat of both having to oversummer in much warmer downstream waters while also spawning in the same habitat as fall run, which enter the system after the heat of the summer and spawn immediately in downstream reaches (Healey, 1991). Consequently, spring run numbers have decreased precipitously, with most populations going entirely extinct in the CCV. As a result, they are now listed as threatened under the Endangered Species Act (National Marine Fisheries Service, 2014). 35 The Yuba River, a tributary of the Feather River within the Sacramento River watershed, once supported an independent spring run population, but like much of the rest of the CCV, due to extensive damming, historic spring run spawning grounds are no longer accessible, making it an excellent system for identifying and understanding if and how various life history forms co- exist in a heavily impacted system (James, 2005). The Yuba River Chinook population is currently managed as one independent fall-run population even though it is unknown how much life history variation within the system exists and is assumed to be largely influenced by strays from the nearby Feather River Hatchery (Lindley et al., 2004). It is unknown if there is an independently spawning spring run population in the Yuba River. If a Yuba River spring run population still exists, it will be critical to manage this watershed appropriately to protect the ESA listed population and, in turn, promote the spring run portfolio. In recent years, notable progress has been made towards understanding the genetic underpinnings of run timing diversity in Chinook salmon. Research in other systems has shown that variation in return timing of fall and spring run Chinook salmon is tightly correlated with variation in the GREB1L to ROCK1 region of the genome, hence referred to in this paper as GREB1L (Prince et al., 2017; Thompson et al., 2019). Chinook salmon homozygous for the early returning variant exhibit an early run timing distribution in the spring while individuals homozygous for the late returning variant exhibit a later distribution in the fall. Heterozygotes in other systems exhibit an intermediate return timing that overlaps to some extent with homozygotes of both alleles. Although this correlation has been well studied and documented in other river systems (such as the Rogue River, Oregon and Klamath River, California) using well- phenotyped samples from migrating adults, studies in the CCV to date have relied on phenotypic proxies for run timing, such as carcass collection date or entry time into a hatchery (Thompson et al., 2020, 2019). While these studies were sufficient to demonstrate the strong correlation of the GREB1L region with run timing in the CCV, the information from live individuals in the midst of their migration provides much more precise information about the timing distributions of each genotype. More precise timing distributions in the CCV could prove to be an invaluable monitoring tool for the conservation of Chinook salmon populations throughout the Central Valley, given the rarity of spring run. In this study, we seek to both identify how many wild- produced migration phenotypes are present in the Yuba River and to explore the relationship between GREB1L genotypes and return time of Chinook salmon in the CCV. Understanding this 36 in the highly impacted Yuba River system will be invaluable for not only the management of the Yuba River, but also will be important for understanding how life history diversity is maintained in highly impacted systems and how we should identify, monitor, and protect this life history diversity to promote salmonid recovery. METHODS Study Site The Yuba River is a tributary of the Feather River, which flows into the Sacramento River. The Yuba has 3 main tributaries, the north, middle, and south forks, which were once historic Chinook salmon spawning habitat but are now inaccessible due to dams on the river. The Yuba River has two main dams that serve as barriers to Chinook salmon migration: the Daguerre Point Dam (DPD), which is located at river mile 11 and passable by salmon via two fish ladders on either side, and the Englebright Dam, which is located at river mile 24 and impassable by salmon (Fig. 2.1). In addition to these complications, upstream from the lower Yuba River there is a large hatchery located on the Feather River that produces both spring and fall run that are thought to potentially stray into the Yuba River during spawning migrations. A key management objective in this system is the Yuba River Accord which is an agreement between all agencies in the Lower Yuba River Management Team (RMT) to manage for improved salmon and steelhead habitat. Within the Yuba River Accord Fisheries Agreement, is a stated purpose to evaluate the presence and viability of spring run Chinook salmon in the lower Yuba River (Yuba County Water Agency et al., 2007). 37 Figure 2.1: Map of the Yuba River system, a tributary of the Feather River. Black bars indicate dams. Orange highlighted areas indicate sampling locations: 1) spawner survey sampling location, 2) acoustic tagging sampling area, and 3) carcass sampling area. Sample collection Two sampling efforts, an acoustic telemetry project and a carcass survey, were conducted by the RMT between the years 2009-2011 during their annual spawner surveys to characterize Chinook migration up the Yuba River to the spawning reaches. For the acoustic telemetry project, adult fish were caught via hook-and-line sampling, targeting fish in the lowermost reaches from the confluence of the Yuba and Feather Rivers to DPD from May to October (Sampling Area 1, Fig. 2.1). Fin clips were collected from all captured fish (N=122), but only fish that were determined to be in “good condition” (showing no signs of disease or injury) were also acoustically tagged (N=42, we refer to these as the “acoustic tagging samples'' and those that 38 were just fin clipped but not tagged as “spawner survey samples''). The acoustic tagging samples were tagged with VEMCO V13-1L acoustic transmitters via esophageal/gastric insertion and were detected via two ultrasonic receivers located in the north and south side of the top of the fish ladder to detect fish successfully passing DPD from both sides (Sampling Area 2; PSMFC, 2011; VEMCO, 2010). The most upstream area was sampled via carcass surveys that occurred upstream of the DPD on a weekly basis (Sampling Area 3, Fig. 2.1), starting 10-15 days after the first spawning redds were detected each year. Only fresh carcasses (possessing at least one clear eye and gills that are red or pink) were sampled to avoid sampling fish that had degraded DNA and had already been in the system for a long period of time. In 2009 and 2010, tissue samples were taken from carcasses throughout the river reach between the DPD and Englebright Dam (Sampling Area 3). All fin clips, regardless of survey method, were dried and placed into individual envelopes then sent to the Meek genetics lab at Michigan State University for processing. Table 2.1: Analyzed genetic samples. Numbers are presented by year and survey type. Note that Acoustic tagging individuals were first surveyed in the spawner survey and then again when they passed DPD, and as such are a portion of the spawner survey individuals. Sample Year 2009 2010 2011 Survey type Spawner Survey Acoustic Tagging Carcass Survey Total Spawner Survey Acoustic Tagging Carcass Survey Total Spawner Survey Acoustic Tagging Carcass Survey Total Total 39 N 0 0 37 37 92 18 35 127 30 24 0 30 194 Run-type Assignment We first assigned individuals to phenotypic run-timing by the date of their detection in the system. The Yuba River RMT uses two “differentiation days” to classify individuals to either early spring, late spring, or fall run timing category. If an individual fish passes DPD prior to July 15th, they are considered early spring run migrants, while after that but prior to October 1st they are considered late spring run migrants. All fish after October 1st are considered fall run migrants (Poxon and Bratovich, 2020). We used these same metrics to classify individuals to their phenotypic run-timing and compare with their GREB1L genotypes. We used this same method of classification for fish surveyed below DPD. To genotypically assign run-type, we extracted DNA from fin clips using the DNeasy® Blood and Tissue extraction kit (Qiagen, Valencia, CA). We genotyped fish at the GREB1L locus by selecting five Single Nucleotide Polymorphisms (SNPs) across the GREB1L region of the genome that had been identified as strongly associated with run timing in previous analyses (Koch and Narum, 2020; Prince et al., 2017; Thompson et al., 2020, 2019). SNPs were screened from the input design sequences (Suppl. Table 2.1) by cross-checking against a multi-population dataset utilized by Thompson et al. (2019). We developed those SNPs into Fluidigm SNPtype assays. Individuals were genotyped at the five SNPs using the Fluidigm EP1 platform (Fig. 2.2). From those markers we were able to make assignments to either homozygous early, homozygous late, or heterozygous genotypes. Genotypes were only allowed to have a total of two or fewer missing SNP genotypes otherwise they were deemed ambiguous and reported as “not called.” Those samples were not included in the final analyses. 40 Figure 2.2: Diagram of relative SNP positions in the GREB1L region on chromosome 28 of the Chinook genome, Otsh_v2.0 (GCF_018296145.1) used for genotyping analysis (Christensen et al., 2018). Statistical analysis We calculated the mean return date for each run using the day of year converted to Julian date of detection in the system by each of the three methods: spawner surveys, acoustic tagging, and carcass surveys. To test if there was a significant difference in mean detection date for each of the three genotypes within each survey method, we used a Kruskal-Wallis test due to the unequal variance among sampling dates. After determining whether the differences between mean detection dates for the genotypes were significant, we then ran a Dunn test of significance to see which genotype detection dates specifically were significantly different from each other within each method, with a full pairwise comparison: homozygous early vs heterozygous, heterozygous vs homozygous late, and homozygous late vs homozygous early. RESULTS Within the Yuba River, genetic assignments show there are genetically spring run (GREB1L homozygous early), fall run (GREB1L homozygous late) and GREB1L heterozygous individuals in the system. In total, we found 125 homozygous early, 25 heterozygous, and 44 41 homozygous late individuals. When compared with survey data, we found that genetic versus date assigned run type were not in perfect agreement. We found homozygous early individuals in both spring early and spring late migrant phenotypic categories, while homozygous late individuals show up in the fall phenotypic category (Fig. 2.3). Interestingly heterozygous individuals appear below DPD at the same time as homozygous early individuals and were categorized as spring early and spring late based on sample date (Fig. 2.3A), however all heterozygous fish with acoustic tags crossed DPD later in the season. This caused them to be categorized as spring late and fall based on sample date (Fig. 2.3B). We found that this was likely because although homozygous early and heterozygous individuals arrive at the dam at the same time (as early as May 25th, Fig 4A), they crossed the dam at different time periods with homozygous early fish crossing the dam earliest (as early as June 30th). We did not see the heterozygous individuals crossing the dam until later (at the earliest by August 28th, Fig 4B). For the post-spawning carcass surveys, we saw a similar, albeit less protracted pattern, with homozygous early being detected at earlier dates, homozygous late being detected at later dates, and heterozygous individuals being detected at intermediate times (Fig 4C). Our results clearly show that homozygous early individuals cross the dam earlier while homozygous late individuals cross the dam later in the season, with the mean return date being statistically significantly different (p = 0.0004). The same pattern was statistically significant across all sampling methods, with homozygous late mean return dates being later than homozygous early (spawner survey: p = 0.0067, carcass survey: p = 5.58 e -11). Across all sampling methods, heterozygous mean migration dates were not significantly different from homozygous early, despite slight differences in the mean migration date (Table 2.2). 42 A B Figure 2.3: Stacked bar graphs of GREB1L genotyped proportions of individuals sorted into phenotypes classified by when they entered the system as Spring early (before July 15th), Spring late (after July 15th but before October 1st) or Fall (after October 1st) using A) Fish surveyed when they first arrived in the system below DPD, B) Fish in A that were acoustically tagged by the date they passed DPD. A Figure 2.4: Genotypic assignments plotted against date of entry into the Yuba River system colored by GREB1L genotype and median return date using A) Fish surveyed as they entered the Yuba River below DPD, B) acoustically tagged fish in Panel A that passed DPD, and C) Fish detected in carcass surveys, post-spawn. Sample Date is in Julian days, with the equivalent calendar days as follows: 150 = May 30th and day 350 = December 16th. 43 Figure 2.4 (cont’d) B C 44 Table 2.2: Statistical results for Kruskal-Wallis comparisons and Dunn test of detection date for each of the three collection methods, comparing within each method for each of the three genotype classifications, where * indicates a significant value. Survey Type Genotype Mean Return Date Kruskal Wallis 𝛘2 Kruskal Wallis p Comparison Spawner Survey Homozygous early 198.41 10.23 0.01* Heterozygous 183.13 Homozygous late 298.33 Acoustic Tagging Homozygous early 227.03 13.45 0.0011* Heterozygous 256.00 Homozygous late 298.60 Carcass Survey Homozygous early 280.57 46.93 6.24e-11* Heterozygous 289.08 Homozygous late 312.75 45 Homozygous early/ heterozygous Heterozygous/ homozygous late Homozygous late/ homozygous early Homozygous early/ heterozygous Heterozygous/ homozygous late Homozygous late/ homozygous early Homozygous early/ heterozygous Heterozygous/ homozygous late Homozygous late/ homozygous early Dunn test p 0.30 0.002* 0.006* 0.777 0.067 0.0004* 0.2883 0.0001* 5.58e-11* DISCUSSION This study provides direct evidence of spring run Chinook salmon in the Yuba River, and further validation that the GREB1L run timing genotypes are correlated with early or late river sample date. Our data show that individuals entering the system early in the season are genetically homozygous for the early migrating allele or heterozygous, while individuals that enter the system late are homozygous for the late migrating allele. From the acoustic tagging data collected, it appears that heterozygous individuals are passing the dam at a slightly intermediate time point, even though they first appear in the system at the same time as homozygous early running fish. We recognize that our sample numbers for heterozygotes are lower than one would prefer (Fig. 2.4, Suppl. Table 2.2) and additional acoustic tagging would assist in further elucidating the strength of these relationships, however, given the extremely threatened nature of these fish and their very low population sizes, we think the information provided by these samples is incredibly valuable. Additionally, the fact that we didn’t find more heterozygotes in this system also points to the maintenance of these distinct life histories and genotypes, despite homogenizing anthropogenic influence. We show there is clearly a pattern of homozygous early genotypes entering the system early through all survey methods. In addition, we see a clear significant difference in spawning time between homozygous early and homozygous late that maintains their temporal segregation in spawning time despite the elimination of spatial separation. Although it is plausible the carcasses were not surveyed until after fish had entered the system, we are certain that surveys were carried out weekly and decomposition rates in this system are fast enough for us to be confident that those samples had not spawned in the system for many additional days beyond when they were sampled. Our validation of the relationship between GREB1L genotypes and migration phenotypes in the Central Valley is exciting because it means GREB1L can be used to detect, monitor, and quantify the presence of different runs in the Central Valley. The advent of SHERLOCK, which allows especially fast, economical, and field deployable genotyping of the GREB1L locus, makes this possibility even more feasible and has the potential to revolutionize our ability to understand and monitor Chinook salmon life history diversity throughout the Central Valley (Baerwald et al., 2020). In addition, the results found in this study and the combination of tagging and carcass surveys could be used to provide spring run spawner abundance estimates each year, which is critical information for managing this spring run separately from fall run. 46 Our study also shows that although the dam has eliminated spatial separation between the runs creating some overlap between the presence of spring and fall returning individuals in the system, it does appear that time of entry in the system can also be used as a proxy to determine run-type in the Yuba River. Our research shows that despite anthropogenic influence and very limited to no historical access to spring run spawning habitat due to dam construction, there are still both spring and fall returning populations that are genetically distinct and temporally separated from each other in the Yuba River. This temporal separation is likely only possible due to cold water pools above the DPD and below the Englebright Dam that allow for spring running fish to survive over summer and spawn (Pasternack et al., 2010). It is encouraging that the Yuba River has maintained a spring run population, indicating that important diversity needed to maintain federally listed populations still exists within this altered landscape. Unfortunately, populations have been excluded from large areas of historic oversummering habitat and the remaining habitats are predicted to disappear with a warming climate, leaving only the north fork of the Yuba as potential habitat for spring-running fish (Cordoleani et al., 2021). To ensure the persistence of spring-running fish, it will be necessary to maintain and manage cold water access for these populations. Discovering that distinct early migrants exist within the Yuba River provides evidence that the system may be able to recover if appropriate conservation efforts and management actions are taken. There is currently an agreement among state, federal, and local officials to reopen large portions of habitat for Yuba River fish. This planned restoration includes the testing and creation of a comprehensive reintroduction plan to reintroduce CCV spring run Chinook salmon into the upper Yuba River habitats as well as habitat restoration design to allow more natural passage around Daguerre Point Dam (California State Government, 2023). This is an important step towards spring run Chinook salmon recovery, however, given impending threats posed by climate change, further actions may be required to ensure that spring run populations recover and persist. Research has shown that intraspecific diversity within spring run Chinook salmon is critical for responding to changing climatic conditions, particularly increases in river and ocean temperatures, helping populations to maintain biocomplexity necessary for resilience and persistence (Cordoleani et al., 2021). More research is needed to fully understand how the amount of migration timing diversity, particularly withing spring run, contributes to an overall portfolio effect, but this will likely be curtailed by lack of available habitat. Because spring run 47 Chinook salmon rely on cool water to hold over during the summer months, this makes them more susceptible to future threats and continued anthropogenic change such as climate change and water diversion (Meyers et al., 1998; National Research Council, 2004; Quinn et al., 2016). It will therefore be important to ensure that any management actions in the Yuba River promote both the genetic and phenotypic diversity in the system, as well as the hydrological conditions needed to support that diversity. The Central Valley is a complex and highly altered system with many historical and contemporary threats to life history diversity in fishes (Fisher, 2016; Williams, 2006). However, our work shows that altered ecosystems can still sustain genetic and life history diversity. Life history diversity in salmon has been especially important to maintain species resiliency and persistence, and will continue to be of high importance as we experience more development and more extreme climate regimes (Beechie et al., 2006; Bourret et al., 2016; Pearson et al., 2014). It is often assumed that systems where subpopulations are extirpated or contain introgressed individuals are lacking or have lost life history diversity and biocomplexity. Without a full understanding of variation in genotypes and phenotypes in degraded systems, it is all but impossible to manage them to maintain this diversity. This study highlights the importance of identifying, monitoring, and protecting diversity, even in highly altered environments. In order to ensure the persistence and resilience of the populations in the face of climate change, it will be necessary to protect the little diversity that is left before it is lost forever. 48 BIBLIOGRAPHY Allendorf, F.W., Berry, O., Ryman, N., 2014. So long to genetic diversity, and thanks for all the fish. Mol. Ecol. https://doi.org/10.1111/mec.12574 Baerwald, M.R., Goodbla, A.M., Nagarajan, R.P., Gootenberg, J.S., Abudayyeh, O.O., Zhang, F., Schreier, A.D., 2020. Rapid and accurate species identification for ecological studies and monitoring using CRISPR-based SHERLOCK. Mol. Ecol. Resour. 20, 961–970. https://doi.org/10.1111/1755-0998.13186 Beechie, T., Buhle, E., Ruckelshaus, M., Fullerton, A., Holsinger, L., 2006. Hydrologic regime and the conservation of salmon life history diversity. Biol. Conserv. 130, 560–572. https://doi.org/10.1016/j.biocon.2006.01.019 Bottom, D.L., Jones, K.K., Simenstad, C.A., Smith, C.L., 2009. Reconnecting social and ecological resilience in salmon ecosystems. Ecol. Soc. 14. https://doi.org/10.5751/ES- 02734-140105 Bourret, S.L., Caudill, C.C., Keefer, M.L., 2016. Diversity of juvenile Chinook salmon life history pathways. Rev. Fish Biol. Fish. 26, 375–403. https://doi.org/10.1007/s11160-016- 9432-3 Brooks, T.M., Mittermeier, R.A., Da Fonseca, G.A.B., Gerlach, J., Hoffmann, M., Lamoreux, J.F., Mittermeier, C.G., Pilgrim, J.D., Rodrigues, A.S.L., 2006. Global biodiversity conservation priorities. Science. https://doi.org/10.1126/science.1127609 California State Government, O. of the G., 2023. Governor Newsom Announces Agreement to Reopen Yuba River to Salmon and Launch River Restoration. Off. Gov. Gavin Newsom. Carlson, S.M., Satterthwaite, W.H., 2011. Weakened portfolio effect in a collapsed salmon population complex. Can. J. Fish. Aquat. Sci. 68, 1579–1589. https://doi.org/10.1139/f2011-084 Chapin, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek, P.M., Reynolds, H.L., Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C., Díaz, S., 2000. Consequences of changing biodiversity. Nature. https://doi.org/10.1038/35012241 Christensen, K.A., Leong, J.S., Sakhrani, D., Biagi, C.A., Minkley, D.R., Withler, R.E., Rondeau, E.B., Koop, B.F., Devlin, R.H., 2018. Chinook salmon (Oncorhynchus tshawytscha) genome and transcriptome. PLoS ONE 13, 1–15. https://doi.org/10.1371/journal.pone.0195461 Colombi, B.J., 2012. Salmon and the Adaptive Capacity of Nimiipuu (Nez Perce) Culture to Cope with Change. Am. Indian Q. 36, 75–97. https://doi.org/10.5250/amerindiquar.36.1.0075 Cordoleani, F., Phillis, C.C., Sturrock, A.M., FitzGerald, A.M., Malkassian, A., Whitman, G.E., Weber, P.K., Johnson, R.C., 2021. Threatened salmon rely on a rare life history strategy in a warming landscape. Nat. Clim. Change 11, 982–988. https://doi.org/10.1038/s41558- 021-01186-4 49 Cordoleani, F., Satterthwaite, W.H., Daniels, M.E., Johnson, M.R., 2020. Using Life-Cycle Models to Identify Monitoring Gaps for Central Valley Spring-Run Chinook Salmon. San Franc. Estuary Watershed Sci. 18. https://doi.org/10.15447/sfews.2020v18iss4art3 Crozier, L.G., McClure, M.M., Beechie, T., Bograd, S.J., Boughton, D.A., Carr, M., Cooney, T.D., Dunham, J.B., Greene, C.M., Haltuch, M.A., Hazen, E.L., Holzer, D.M., Huff, D.D., Johnson, R.C., Jordan, C.E., Kaplan, I.C., Lindley, S.T., Mantua, N.J., Moyle, P.B., Myers, J.M., Nelson, M.W., Spence, B.C., Weitkamp, L.A., Williams, T.H., Willis- Norton, E., 2019. Climate vulnerability assessment for Pacific salmon and steelhead in the California Current Large Marine Ecosystem. PLOS ONE 14, e0217711. https://doi.org/10.1371/journal.pone.0217711 Des Roches, S., Pendleton, L.H., Shapiro, B., Palkovacs, E.P., 2021. Conserving intraspecific variation for nature’s contributions to people. Nat. Ecol. Evol. 5, 574–582. https://doi.org/10.1038/s41559-021-01403-5 Dittman, A., Quinn, T., 1996. Homing in Pacific salmon: mechanisms and ecological basis. J. Exp. Biol. 199, 83–91. https://doi.org/migration comportment olfaction orientation signaux orienteurs Finney, B.P., Gregory-Eaves, I., Douglas, M.S.V., Smol, J.P., 2002. Fisheries productivity in the northeastern Pacific Ocean over the past 2, 200 years. Nature. https://doi.org/10.1038/416729a Fisher, F.W., 2016. Past and Present Status of Central Valley Chinook Salmon 8, 870–873. Griffiths, J.R., Schindler, D.E., Armstrong, J.B., Scheuerell, M.D., Whited, D.C., Clark, R.A., Hilborn, R., Holt, C.A., Lindley, S.T., Stanford, J.A., Volk, E.C., 2014. Performance of salmon fishery portfolios across western North America. J. Appl. Ecol. 51, 1544–1563. https://doi.org/10.1111/1365-2664.12341 Healey, M.C., 1991. Life history of Chinook Salmon (Oncorhynchus tshawytscha), Pacific Salmon Life Histories. Heino, J., Virkkala, R., Toivonen, H., 2009. Climate change and freshwater biodiversity: Detected patterns, future trends and adaptations in northern regions. Biol. Rev. https://doi.org/10.1111/j.1469-185X.2008.00060.x Herbold, B., Carlson, S.M., Henery, R., Johnson, R.C., Mantua, N., McClure, M., Moyle, P., Sommer, T., 2018. Managing for Salmon Resilience in California’s Variable and Changing Climate. San Franc. Estuary Watershed Sci. 16. https://doi.org/10.15447/sfews.2018v16iss2art3 Hilborn, R., Quinn, T.P., Schindler, D.E., Rogers, D.E., 2003. Biocomplexity and fisheries sustainability. Proc. Natl. Acad. Sci. 100, 6564–6568. https://doi.org/10.1073/pnas.1037274100 James, L.A., 2005. Sediment from hydraulic mining detained by Englebright and small dams in the Yuba basin. Geomorphology, Dams in Geomorphology 71, 202–226. https://doi.org/10.1016/j.geomorph.2004.02.016 50 Johnson, B.M., Kemp, B.M., Thorgaard, G.H., 2018. Increased mitochondrial DNA diversity in ancient Columbia River basin Chinook salmon Oncorhynchus tshawytscha. PLoS ONE 13, 1–26. https://doi.org/10.1371/journal.pone.0190059 Koch, I.J., Narum, S.R., 2020. Validation and association of candidate markers for adult migration timing and fitness in Chinook Salmon. Evol. Appl. 13, 2316–2332. https://doi.org/10.1111/eva.13026 Layman, R.C., Boyce, J.R., Criddle, K.R., 2006. Economic Valuation of the Chinook Salmon Sport Fishery of the Gulkana River, Alaska, under Current and Alternate Management Plans. Land Econ. https://doi.org/10.2307/3147161 Lindley, S.T., Schick, R., May, B.P., Anderson, J.J., Greene, S., Hanson, C., Low, A., McEwan, D., MacFarlane, R.B., Swanson, C., Williams, J.G., 2004. Population structure of threatened and endangered Chinook salmon ESUs in California’s Central Valley basin. NOAA Tech. Memo. NOAA-TM-NM, 70 p. Mahnken, C., Ruggerone, G., Waknitzl, W., Flagg, T., 1998. A Historical Perspective on Salmonid Production from Pacific Rim Hatcheries. North Pac. Anadromous Fish Comm. Bull. 1, 38–53. Malick, M.J., Cox, S.P., 2016. Regional-scale declines in productivity of pink and chum salmon stocks in western North America. PLoS ONE. https://doi.org/10.1371/journal.pone.0146009 Meek, M.H., Stephens, M.R., Goodbla, A., May, B., Baerwald, M.R., 2020. Identifying hidden biocomplexity and genomic diversity in chinook salmon, an imperiled species with a history of anthropogenic influence. Can. J. Fish. Aquat. Sci. 77, 534–547. https://doi.org/10.1139/cjfas-2019-0171 Meyers, J.M., Kope, R.G., Gregory, B.J., Teel, D., Lierheimer, L.J., Grant, W.S., Waknitz, F.W., Neely, K., Lindley, S.T., Waples, R.S., 1998. Status Review of Chinook Salmon from Washington, Idaho, Oregon, and California (National Oceanic and Atmospheric Administration). Washington D.C. Moore, J.W., Yeakel, J.D., Peard, D., Lough, J., Beere, M., 2014. Life-history diversity and its importance to population stability and persistence of a migratory fish: Steelhead in two large North American watersheds. J. Anim. Ecol. https://doi.org/10.1111/1365- 2656.12212 Moyle, P.B., Lusardi, R.A., Samuel, P.J., Katz, J.V.E., 2017. State of Salmonids: Status of California’s Emblematic Fishes 2017. San Francisco, CA. Narum, S.R., Genova, A.D., Micheletti, S.J., Maass, A., 2018. Genomic variation underlying complex life-history traits revealed by genome sequencing in Chinook salmon. Proc. R. Soc. B Biol. Sci. 285. https://doi.org/10.1098/rspb.2018.0935 National Marine Fisheries Service, 2014. Recovery Plan for the Evolutionarily Significant Units of Sacramento River Winter-Run Chinook Salmon and Central Valley Spring-Run Chinook Salmon and the Distinct Population Segment of Central Valley Steelhead, Notes. 51 National Research Council, 2004. Endangered and Threatened Fishes in the Klamath River Basin: Causes of Decline and Strategies for Recovery. National Academies Press, Washington, D.C. https://doi.org/10.17226/10838 Pasternack, G., Fulton, A., Morford, S., 2010. Yuba River analysis aims to aid spring-run chinook salmon habitat rehabilitation. Calif. Agric. 64, 69–77. Pearson, R.G., Stanton, J.C., Shoemaker, K.T., Aiello-Lammens, M.E., Ersts, P.J., Horning, N., Fordham, D.A., Raxworthy, C.J., Ryu, H.Y., Mcnees, J., Akçakaya, H.R., 2014. Life history and spatial traits predict extinction risk due to climate change. Nat. Clim. Change. https://doi.org/10.1038/nclimate2113 Poxon, B., Bratovich, P., 2020. 2020 Update Lower Yuba River Vaki Riverwatcher Chinook Salmon Passage and Run Differentiation Analyses. Lower Yuba River Accord River Management Team. Prince, D.J., O’Rourke, S.M., Thompson, T.Q., Ali, O.A., Lyman, H.S., Saglam, I.K., Hotaling, T.J., Spidle, A.P., Miller, M.R., 2017. The evolutionary basis of premature migration in Pacific salmon highlights the utility of genomics for informing conservation. Sci. Adv. 3. https://doi.org/10.1126/sciadv.1603198 PSMFC, 2011. Specific Sampling Protocols and Procedures for Chinook Salmon Escapement Monitoring. Quinn, T.P., McGinnity, P., Reed, T.E., 2016. The paradox of “premature migration” by adult anadromous salmonid fishes: patterns and hypotheses. Can. J. Fish. Aquat. Sci. 73, 1015– 1030. https://doi.org/10.1139/cjfas-2015-0345 Schindler, D.E., Hilborn, R., Chasco, B., Boatright, C.P., Quinn, T.P., Rogers, L.A., Webster, M.S., 2010. Population diversity and the portfolio effect in an exploited species. Nature 465, 609–612. https://doi.org/10.1038/nature09060 Sih, A., Jonsson, B.G., Luikart, G., 2000. Habitat loss: Ecological, evolutionary and genetic consequences. Trends Ecol. Evol. 15, 132–134. https://doi.org/10.1016/S0169- 5347(99)01799-1 Thibaut, L.M., Connolly, S.R., 2013. Understanding diversity–stability relationships: towards a unified model of portfolio effects. Ecol. Lett. 16, 140–150. https://doi.org/10.1111/ele.12019 Thompson, N.F., Anderson, E.C., Clemento, A.J., Campbell, M.A., 2020. A complex phenotype in salmon controlled by a simple change in migratory timing 613, 609–613. Thompson, T.Q., Renee Bellinger, M., O’Rourke, S.M., Prince, D.J., Stevenson, A.E., Rodrigues, A.T., Sloat, M.R., Speller, C.F., Yang, D.Y., Butler, V.L., Banks, M.A., Miller, M.R., 2019. Anthropogenic habitat alteration leads to rapid loss of adaptive variation and restoration potential in wild salmon populations. Proc. Natl. Acad. Sci. U. S. A. https://doi.org/10.1073/pnas.1811559115 VEMCO, 2010. V13 coded transmitter [WWW Document]. Innovasea. URL https://innovasea.com/fish-tracking/ (accessed 7.5.23). 52 Weeder, J.A., Marshall, A.R., Epifanio, J.M., 2005. An Assessment of Population Genetic Variation in Chinook Salmon from Seven Michigan Rivers 30 Years after Introduction. North Am. J. Fish. Manag. 25, 861–875. https://doi.org/10.1577/M03-227.1 Williams, J.G., 2006. Central Valley salmon: a perspective on Chinook and steelhead in the Central Valley of California. San Franc. Estuary Watershed Sci. 4, 1–393. https://doi.org/10.5811/westjem.2011.5.6700 Yoshiyama, R.M., Fisher, F.W., Moyle, P.B., 1998. Historical Abundance and Decline of Chinook Salmon in the Central Valley Region of California. North Am. J. Fish. Manag. 18, 487–521. https://doi.org/10.1577/1548-8675(1998)018<0487:HAADOC>2.0.CO;2 Yuba County Water Agency, California Department of Fish and Game, South Yuba River Citizens League, Friends of the River, Trout Unlimited, The Bay Institute, 2007. Lower Yuba River Fisheries Agreement. 53 Table S2.1: Single Nucleotide Polymorphisms used in Fluidigm type assays from the GREB1L region with their name, genomic positions, original publication, and sequence. APPENDIX 2A: GREB1L FLUIDIGM SEQUENCES SNPtyp e Fluidig m Name Otsh_ v1.0_ NC_s caffol d Otsh_v 1.0_N W_scaf fold Otsh _v2. 0 Original publicatio n identifyin g SNP Sequence used as input for Fluidigm assay design chr28_ NW_0 201285 28.1|:2 194538 NC_ 0564 56.1: 1345 7880 NC_0 37124 .1:122 73002 GREB1l _pos219 4538 Thompso n et al. (2020) GATAAGGGGATAAGGGAGGTCATGCAAATTCCATACCATCCAGGTCAGACAGTGCTAGAACTTTAACCGGAACGCTGC ATGAGTTTAGGGAACATTCTCTTTAGTA[T/C]CAGACTGAACATCCAAATCTTCCTTCACTTCTAGATACACGCTTTAAGG GCCCTCTAGGCAGCTAACTCTGCATCCACAGTAATATAACCCATTCTAGGAGACATTCTTATAACACTGGCCTAGACTAC AAATCACTCTTAACATAACCCTGTAGCTGTGTCCATGATCACAGGGTCACTATCAA chr28_ NW_0 201285 28.1|:2 198644 NC_ 0564 56.1: 1346 1994 NC_0 37124 .1:122 77108 Koch & Narum (2020) GREB1l _pos219 8644 chr28_ NW_0 201285 28.1|:2 199210 NC_ 0564 56.1: 1346 2560 NC_0 37124 .1:122 77674 Koch & Narum (2020) GREB1l _pos219 9210 TTTGTCTTCCATTGATATTTGACCTCATGTGGATGTGCCAATGACAACATTATTATTCTCACTCTTAAATCCAACATTAGG GAGACTTAAAACAACCTCAAAAGAGCTACACAATATATTCACGATAACACCATATGTCGYTTGTYTCCTTCACCTGCAA CCTTCTATTCAACAGTCCATTCTTAGAAAAATGACAAGCCYGAGTAAGCCAGTCGGTGAGCCATTCATAACAATCTTAA CATTACTTT[T/A]CAAAAATATTGGATTCGGAATATGGATTCATAACATAATTATGTTATCCTGGATCATTCAAGAGAAAT GAACAGACGGATGAAACATTAAGTCAGAGGATGTTGATCATGACCATATTGTTTAACTGTAATTCTTTCATTTTCATCTT TGCATAGCCAGGCAAGCCGTGTGACTGACTGACTGCCTTAGTCTTCAGTTCATTACAGCAGATCTAGTCAACAGTTGGTT TAATCTGTCCGTATAACTCTTCTCACCTCCT CTCCACACCACTCATTCATCATACACACATCGCGCATTCTATGCTGAACSTGGCGGTTCGTGTCCATTGCATTATTATACG ACACAGCGTCTGTCTSTCTGWATGGACTCTRTAGGCTCCCGGGGGTAGTCCATTTGAAACAGTTGGAGTAAAGAATGAA AGAAAGAGATGACTTGTKCCCTAAGAGGAGACGAGCATTACAGTTAGTAAACATTACAGTTTCC[T/A]GTCTGAGGTAA ATCAACATATGACCACTCGAAAACTCCCCAAATAAGCTCATTTGGTACAGACCAGCACTAGCAGCAAGTTCAACCTGGG AAGAGGAGTCTCACGGKGTGATTAATCTCCCCCAGCTCCCAGCAGTAGCTCCCTCCCTCCCYGACTTTGACACAGCAGC CACGTTTAAATAGACCCGTTTGAAATGAAGATAATGAGTAAACCCAGCGGTTTCTTTGGCCTCAAAMGAGCCCTGTGTG GGAAAACAAAAGAGCCT chr28_ NW_0 201285 28.1|:2 200828 NC_ 0564 56.1: 1346 4173 chr28_ NW_0 201285 28.1|:2 202893 NC_ 0564 56.1: 1346 6238 NC_0 37124 .1:122 79292 NC_0 37124 .1:122 81357 GREB1l _pos220 0828 GREB1l _pos220 2893 TAAGGGTTGTGGGTGGTGGGGTGGATTAGCCAGTGGGGACTATAAAGGGGAGTGAACTAGGGTTTAAGGCCTGTTGTGA CAGAGGAGCTGGGGAAGGGCTGATGGGGGGGCKGGGGGGAGGCGGACAAAAGGAGCATTTGGGCAGATGAAGAAGTC ATCATCATTAAGCCACTGGAAGTTTACTGTCCAGTTATAAAAGTCATTTCAAAATTAGGRGGTTAGGGGGTGCGTGTGA AAGG[G/A]GAGAAGGGCTCAGAGTGCCTGAGAAGGCCTGGGGGYGGGGCAGATGAGAGCTGTGGCCTGTGGTTGTGAG GGACTCTGTGGGACTGGGGGGCCAATTCATTAGGGGCACAGCCCAGCCTTTGTGTTTGCACCAGGTTGATTGGAGTGCT GACCTTGCCTTGCCTCCCAGCCTTCCCTGCACACTCTGCTCTGGCCCAGTGGAGGATGAGTATAAGGGCAAGGCATTTAA CCTTCAACTAAATCCCAGCCTCAACCACAGCAGACAAAGG Koch & Narum (2020) Thompso n et al. (2020) ATTTACCTCCCTGCCCCAGACAATTCTTGAATCACATGGCTGCTGCATTTCATAATGAAAAACAAGGCCA[A/T]ATCAGG AAGTTCAGCCCTCTTTAAATGTGGAAAAMAAAATACAKAGAACATTTTCACTTAGTGTTGTTCTTTTTAAATTTAATTTG AGGCCTGGAGGACAAACTCAATCAATGTGCGGAATTACTGATAATTGACCATGCTCGCTGAGAAGGCCRAATAAAATTG AAGCCCTGAKTGAACCCGCTCTGCATTTTACAACACTGC 54 APPENDIX 2B: GENOTYPES BY SAMPLING METHOD Table S2.2: Number of samples in each genotype category organized by each type of survey. Note that samples from individuals in the that were in the acoustic tagging survey were also included in the spawning survey, since that is where they were first detected. Survey Type Spawner Survey Acoustic Tagging Carcass Survey Total Homozygous early 102 33 23 92 Heterozygous 12 4 13 21 Homozygous Late 8 5 36 41 55 CHAPTER 3: GENETIC DIVERGENCE OF RECENTLY INTRODUCED POPULATIONS OF CHINOOK SALMON IN NEW ZEALAND INTRODUCTION Over the past century, anthropogenic activities have resulted in a loss of biodiversity so severe that scientists have classified it as earth’s sixth mass extinction event (Tilman 2009; Pyšek and Richardson 2010; Bellard et al. 2012; Intergovernmental Panel on Climate Change 2014; Ceballos et al. 2015). Overfishing, habitat loss and fragmentation, introduced species, and changes in environmental conditions due to climate change have impacted aquatic systems disproportionately (Beddington et al. 2007; Worm et al. 2009; Sadovy de Mitcheson et al. 2013). One of the key challenges facing conservation and management practitioners is how to address this widespread loss of biodiversity. Biodiversity encompasses ecosystem diversity, species diversity, and genetic diversity. In particular, genetic diversity is important as it provides the fundamental building blocks for speciation, as well as harboring the adaptive potential for populations (Chapin et al. 2000; Brooks et al. 2006). Additionally, genetic diversity within a species or population can result in a diversity of phenotypes that stabilizes populations over time, often referred to as biocomplexity (Hilborn et al. 2003). This type of diversity buffers the effects of natural and anthropogenic change (Narum et al. 2018). Imperiled populations often have reduced genetic variation due to the consequences of large population declines in abundance (Nei et al. 1975; Vrijenhoek 1994). This reduced genetic variation as well as inbreeding due to low population sizes can affect populations’ ability to persist (Frankham 2005; Frankham 2015). One way to potentially combat the ill effects of inbreeding and recover these populations is to use genetic rescue, translocating genetically diverse individuals from another region to provide an influx of genetic diversity (Whiteley et al. 2015; Fitzpatrick et al. 2020). Although there are some potential risks to introducing individuals to a population, such as outbreeding depression, increasing empirical evidence also shows that a small amount of gene flow to genetically depauperate populations can result in increased abundance and reproductive success (Frankham 2016; Fitzpatrick et al. 2020). Chinook salmon is an ecologically, culturally, and economically important species that displays an extraordinary amount of phenotypic diversity that contributes to their overall persistence and resilience (Yoshiyama 1999; Bottom et al. 2009; Raheema et al. 2009; Bourret et al. 2016; Quinn 2018). One life history trait that is particularly important for population 56 resilience is spawn return timing, with most Chinook populations across the range displaying some variation in what time of year they make their spawning migrations (Moore et al. 2014; Bourret et al. 2016). Chinook salmon in the Central Valley of California (CCV) are unique in that they are the only populations that four genetically distinct run-timing phenotypes (Fall, Late Fall, Spring, and Winter run) co-occur (Williams 2006; Meek et al. 2016). Unfortunately, populations in the CCV have been heavily impacted by human activity (Yoshiyama et al. 1998; National Research Council 2004; IUCN 2017). These activities include overfishing and exclusion from historical spawning grounds, leading many populations to be extirpated and others to be numerically depressed and listed under the Endangered Species Act (Spring and Winter runs) (Williams 2006). Declines in abundance and distribution of salmon have negatively impacted native peoples of California, including the Winnemem Wintu people, a state recognized tribe of California(Houck 2019). Chinook are spiritually important to the Winnemem Wintu and healthy Chinook populations are of paramount importance(Dallman et al. 2013). The Winnemem are currently engaging in several initiatives to recover and restore their native lands, which largely fall under the more holistic indigenous feminist paradigm of rematriation and can include efforts like Land Back and Water Back (Gray 2022; Leonard et al. 2023; How the Winnemem Wintu won their ancestral land back and help save Chinook Salmon - Vox). Currently, the Winnemem are seeking to rematriate populations of Chinook salmon in the McCloud River, a tributary of the Sacramento River in the CCV. This habitat previously served as spawning grounds for Chinook salmon, but was blocked with the construction of the Shasta Dam, leading to their extirpation above the dam (Houck 2019). One source of Chinook being considered for the McCloud River rematriation is that of Chinook salmon (Oncorhynchus tshawystcha) found in New Zealand (NZ). Between 1901 and 1907, Chinook salmon from the CCV were introduced into the Waitaki River in NZ (McDowall 1994) (Fig. 3.1). It is currently unknown which tributary in the CCV the NZ Chinook originated from because many of the records were lost in a fire (McDowall 1994). From what records do exist, it is clear that the Chinook in NZ originated from a tributary of the Sacramento River, the largest river in the CCV (McDowall 1994). Previous work used microsatellite data to show divergence of the NZ salmon from Battle Creek Fall run in the CCV, one of the hypothesized sources, while another study showed divergence from the Feather River Spring run in the CCV 57 (Quinn et al. 2001; O’Malley et al. 2007). Unfortunately, these microsatellite data were of low resolution, and to date further comparisons of NZ populations to other populations characterized by different spawning timing from the CCV have not been conducted (Kinnison et al. 2002). Given this context, we sought to understand the genetic structure and diversity of NZ Chinook using other molecular methodology. By comparing all populations in both the CCV and NZ we aimed to explore the genetic diversity and structure of NZ Chinook salmon compared to current day CCV populations. Since the initial introduction, NZ Chinook have populated several other rivers near the Waitaki River by natural processes such as straying. This has potentially allowed for different populations in each new river or tributary in NZ to adapt and thus generate unique genetic diversity. NZ populations of Chinook also exhibit divergence in phenotypic traits, including freshwater growth rate, reproductive output, and run timing (Quinn et al. 2001). Because adaptation has potentially taken place in the NZ populations, it is possible reservoirs of genetic diversity exist in NZ Chinook that could be used to inform rematriation efforts, for example genetic rescue aimed at decreasing the negative effects of low genetic diversity in CCV populations. Empirical evidence suggests that an influx of genetic diversity from an evolutionarily similar population can increase population growth, making the NZ Chinook an excellent system to explore for CCV Chinook recovery efforts (Whiteley et al. 2015). Here, we investigate the spatial patterns of genetic diversity in a novel environments for a newly introduced species using high throughput sequencing, showing divergence within and among chinook populations in NZ and the CCV, and address how that can inform a rematriation effort of NZ Chinook salmon to the CCV. The purpose of this research is to assess the spatial patterns of genetic diversity in NZ Chinook salmon, and compare the diversity patterns found within NZ and compared to the CCV. We examine the following questions: 1. Is there population structure within Chinook salmon populations in NZ, and if so; 2. How do NZ Chinook salmon compare in terms of overall allelic richness, heterozygosity, and levels of inbreeding? 3. Using this marker set, do NZ Chinook salmon appear genetically unique compared to CCV populations? 58 METHODS Sample Collection and Sequencing The data analyzed here were obtained from two separate sources that corresponded with CCV vs NZ Chinook salmon. We obtained NZ adult Chinook salmon DNA extracted from fin tissue collected by the Cawthron Institute as part of regular post spawning surveys from three main river catchments between the years 2017-2018 (Fig. 3.1). DNA was extracted from these samples using the high salt method described in Clarke et al, and dried down 20 uL(Clarke et al. 2014). We then rehydrated these samples using a low TE solution and quantified the DNA using a Qubit 3 fluorometer and the High Sensitivity quantitation kit (Thermofisher Scientific). SNP panel design and selection was completed by Danile Gomez-Uchida and Rodrigo Marin Nahuelpi at the Universidad de Concepción. They selected a set of single nucleotide polymorphisms (SNPs) previously identified from restriction site associated DNA RAD sequencing experiments mentioned in previous Chinook studies (Hecht et al. 2015; McKinney et al. 2016; Narum et al. 2017). Briefly, they pulled down the sequence metadata from Hecht et al 2015, Narum et al 2017, and McKinney et al 2016 to get the sequences associated with the polymorphic SNPs, aligned them to the reference genome available at the time (GCA_002872995.1) (Christensen et al. 2018), and used BLASTn in order to get position information for each polymorphic SNP, keeping only one SNP per rad tag. We then used this raw data for our study. The studies the SNPs were obtained from were all performed single digest RAD sequencing on Chinook utilizing the SbfI restriction enzyme using the methods explained in Miller et al 2007 and Baird et al 2008, which allows for direct comparison to the CCV samples used from Meek et al 2019 (Miller et al. 2007; Baird et al. 2008; Meek et al. 2019a). The research these sites were pulled from found these SNPs to be useful for delineating populations, but some were possible sources of adaptive variation. In total, 17,062 were sent for probe development to LGC Biosearch Technologies (hereafter referred to as LGC) to create a SeqSNP panel based on each individual SNP. LGC uses the SNP locations to design high- specificity probes (no off-targets) to create a genotyping panel for targeted genotyping by sequencing using single primer enrichment technology (LGC Group 2023). DNA samples were sent to LGC to be sequenced with this custom panel using the Illumina NextSeq 500 platform, single-end 1 x 75 bp run. In total we obtained genetic information from 89 samples from 3 locations in NZ (Table 3.1). 59 CCV genetic information was obtained from a previously available dataset as described in Meek et al 2019. Briefly, this was a RADseq paired end dataset with read lengths of 150 bp. This research included individuals from all major runs of Chinook Salmon (Fall, Late Fall, Spring and Winter run) from all major tributaries within the CCV, with sampling spanning the years 2001-2010 (Fig. 3.1) (Meek et al. 2019b). In total, we obtained genetic information from 563 individuals from 11 tributaries in the CCV (Table 3.1). Figure 3.1: Maps of sampling locations. The map on the left is of sampling locations in NZ as adapted from Quinn et al 2001(Quinn et al. 2001) and the right is locations in the CCV as adapted from O’Leary et al 2021(O’Leary et al. 2021). Sampling locations in the major rivers of NZ (Rangitata NZ-RG, Rakaia NZ-RK, and Waitaki NZ-WT) are colored in purple dots. Sampling locations in the Sacramento River are colored by run-timing (Fall = blue, Late Fall = red, Yellow = Spring, and Winter = Green) with squares indicating hatcheries. The tributary abbreviations are as follows: MER = Merced River, TOU = Tuolumne River, STN = Stanislaus River, MKH = Mokelumne River Hatchery, NIM = Nimbus River Hatchery, FRH = Feather River Hatchery, BUT = Butte Creek, DER = Deer Creek, MIL = Mill Creek, COL = Coleman Hatchery, USR = Upper Sacramento River. 60 Table 3.1: Table indicating number of samples collected and the known source population and run timing, if available. Abbreviations for the various sites are the same as Figure 1. LOCATION ABBREVIATION Fall Late Fall Winter Spring Unknown Coleman Hatchery Mill Creek Deer Creek Butte Creek COL MIL DER BUT Feather River Hatchery FRH Nimbus River Hatchery NIM Mokelumne River Hatchery MKH Tuolumne River TOU Merced River Hatchery MRH Merced River MER Upper Sacramento River USR Rangitata River Rakaia River Waitaki River Genotyping NZ_RG NZ_RK NZ_WT 30 20 15 21 27 30 28 23 30 31 - - - - - - - - - - - - - - - - - - - - - - - - 21 26 - - - - - - - 16 27 19 7 - - - - - - - - - - - - - - - - - - - - 29 28 32 We processed and quality filtered the genetic data in preparation for alignment. CCV samples were de-multiplexed using the “process_radtags” program in STACKS (Catchen et al. 2013; Rochette and Catchen 2017). NZ samples were pre-processed for quality by LGC using their standard procedures for their SeqSNP projects (LGC Group 2023). We repeated the LGC procedures on all CCV samples, which we explain here. First, reads were clipped to remove adapter sequence and then quality trimmed. Quality trimming consisted of removing reads containing Ns with trimming at the 3’-end over a window of 10 bases to get a minimum average 61 Phred quality score of over 30. Reads with less than 65 bases were discarded. Because later analyses would require reads to be all of the same length and the CCV sequences were longer than NZ reads, we trimmed the end of the reads in all CCV samples to a length of 75 bp and to allow for comparison to NZ samples, we only used the forward reads from this point on. We then assessed sequence quality using fastqc (Babraham Bioinformatics - FastQC A Quality Control tool for High Throughput Sequence Data). After all samples were processed and quality filtered, we aligned them to the reference genome. We then aligned all reads using the bwamem program (Li and Durbin 2009), aligning to the newest published version of the Chinook reference genome (GCA_018296145.1)(Oncorhynchus tshawytscha genome assembly Otsh_v2.0). We completed SNP discovery using the populations program in gstacks portion of Stacks v2.64, via the creation of a catalog (Rochette and Catchen 2017; Rivera-Colón and Catchen 2021). To investigate the dataset multiple ways, we created two catalog datasets to begin genotype filtering because the NZ samples were such a small proportion of the total samples in the NZ/CCV comparison. In order to meaningfully compare NZ populations and minimize the amount of SNPs from NZ populations that would be lost to stringent filters in the NZ/CCV comparison, we created catalogs for both the NZ samples on their own, and the NZ and CCV samples together. The first dataset contains only individuals from NZ, and we will hereafter refer to that dataset as NZO. The second dataset contained both the NZ and CCV populations and will hereafter be referred to as NZCA. After SNP discovery, we then filtered the genotypes in NZO and NZCA for quality using VCFtools (Danecek et al. 2011). The methods for filtering between the two datasets diverged at this point, with more stringent filters applied to the NZO dataset, and then separate filtering parameters applied to the NZCA dataset. For the NZO only dataset, we first filtered out genotypes missing in 50% of individuals, a minor allele count of 3, minimum genotype quality value of 30, and a minimum depth of read of 5. We then filtered and removed SNPs that were missing up to 10% of genotypes, and removed individuals with 50% or more missing data. For the final step of quality filtering on the NZO dataset, we filtered out genotypes with minor allele frequencies < 0.01 to remove possible monomorphic loci. To mitigate the possibility of losing too many SNPs from either population, we manually divided the NZCA populations file produced from the catalog into NZ and CCV samples for 62 initial quality filtering, the NZ individual dataset we will hereafter refer to as NZ1 and the CCV individual dataset we will refer to as CA1. Both datasets were filtered separately for max missing genotypes of 0.2, minimum depth of reads of 3, and minimum genotype quality of 30. After that, we generated a list of SNPs still present in both the NZ1 and CA1 VCFfiles, and then used that list to contain only those SNPs in the original populations file generated by gstacks, generating the initial NZCA dataset. To ensure that no low quality reads were left in this newly filtered NZCA dataset, we filtered this dataset with VCFtools to remove any alleles with a minor allele frequency lower than 0.01 and removed individuals with more than 50% missing data (this resulted in the removal of 48 CCV individuals and 2 NZ individuals). After initial quality filtering, both the NZO and NZCA datasets underwent the same final filtering steps. These reads were then filtered for paralogous genes using HDplot due to the whole genome duplication present in Oncorhynchus species(McKinney et al. 2017). After that, we examined genotypes to retain only one SNP per rad tag or region, eliminating those SNPs that were close together (on the same RAD locus) to minimize the possibility of linkage disequilibrium. To explore relatedness, we calculated pairwise 𝛗, using the relatedness2 function in VCFtools (Manichaikul et al. 2010; Danecek et al. 2011). Individuals with a relatedness value ≥ 0.2 (indicative of excessive levels of relatedness such as half-siblings or closer) were removed from the analysis. Statistical analysis To investigate population level relationships between groups, we performed a principal component analysis (PCA) and a discriminant analysis of principal components (DAPC) to infer overall genetic variation among individuals using the “adegenet” package in R (Jombart 2008; Jombart and Ahmed 2011). The number of clusters was determined by running 25 iterations of the find.clusters module, with 30 possible clusters set as the maximum for the NZCA dataset, one group for each river in NZ and one for each river and run timing iteration in CCV. We ran the same analysis on the NZ samples only, this time setting the maximum clusters at 10 to capture any variation among drainages. We then inferred the most appropriate value of K from the Bayesian information criterion (BIC) values. The most likely value of K was then used to define genetically distinct populations within which genetic diversity was explored in the NZO dataset. For NZCA, we used the most likely value of K as a starting point, but also made some assumptions about groups based on already known genetic differentiation within the CCV. 63 Because the low level of SNPs did not resolve population level differences already known in the CCV, we separated those groups by run-timing and one group for NZ populations. To investigate patterns of genetic variation within the NZ populations in the NZO dataset, we used populations defined by the PCA and DAPC. To calculate observed and expected heterozygosities, we used both of the R packages dartR and poppR (Nei 1978; Kamvar et al. 2014; Gruber et al. 2018). DartR was also used to evaluate each population for the presence of private alleles. We utilized the R package ‘PopGenReport’ to analyze levels of allelic richness(Adamack and Gruber 2014). To assess levels of genetic differentiation between the NZ populations and CCV populations in the NZCA dataset as well as within NZ populations within the NZO dataset, we calculated Fst in dartR with 10,000 bootstrap iterations (Gruber et al. 2018). Table 3.2: Individuals and SNPs retained after initial and final filtering steps based on what dataset the individuals were processed in. Note that NZ individuals were both evaluated on their own and with the larger dataset including CCV individuals. Dataset Subset and individuals contained n before filtering SNPs before filtering n after initial filtering Final n SNPs after initial filtering Final SNPs after all filtering NZO (NZ only) NZCA (NZ and CCV) N/A 89 127,103 89 4007 82 1774 NZ1 (NZ only) CA1 (CA only) 564 312,596 517 28,187 593 131 89 312,596 87 19,037 RESULTS Genetic population structure The unfiltered NZO dataset exported from Stacks contained genotypes for 127,103 SNPs in 89 individuals (Table 3.2). After final quality filtering, we retained 1774 SNPs in 82 individuals. The unfiltered dataset for the NZCA dataset exported from Stacks contained genotypes for 312,596 SNPs in 653 individuals (Table 3.2). Unfortunately, due to the amount of SNPs only present in the CCV populations, many of these SNPs had to be removed. When 64 comparing the overlap between the NZ1 and CA1 outputs for the NZCA data set, only 170 SNPs remained. After removing the additional SNPs for the various quality filtering steps described in the methods, only 131 SNPs remained. Because of the low amount of SNPs in the NZCA dataset, we were not able to accurately determine values of relatedness among individuals. We removed 2 individuals from the NZO dataset due to a relatedness value of ≥ 0.2, and removed those same individuals from the NZCA dataset. Results from the PCA based on the NZO dataset revealed that there was some separation along PC axes and that they were likely two or more genetically distinct populations (Fig. 3.2). When performing DAPC, we identified groups of individuals that separated very clearly into at least 2 (BIC = 427) or 3 groups (BIC = 429) (Fig. 3.3). When examining the posterior probabilities for these groups by NZ river, individuals did not show evidence for admixture among groups (Fig. 3.4). In the K = 2 scenario, one group consisted of all Rangitata individuals, the majority of Waitaki individuals, and roughly half the Rakaia individuals, while the other group was almost entirely Rakaia individuals. With a K = 3 scenario, we began to see one group of almost entirely Waitaki origin individuals, one group of Rangitata individuals with roughly half the Rakaia individuals, and another group almost entirely Rakaia individuals. The K = 4 scenario shows signs of overfitting our model as it does not appear the genetic groupings are no longer biologically meaningful. This shows that the models showing K = 2 or 3 are the most likely biologically accurate. Results from the PCA and DAPC NZCA dataset revealed that the CCV populations and NZ populations were genetically differentiated from each other, based on our limited SNP set. The PCA resulted in three groups, with some overlap, showing NZ and winter run CCV individuals beginning to separate out, with some overlap (Fig. 3.5). This was also apparent in the results from the DAPC, which found that a K value of five (BIC = 905) was the most optimal during K means clustering (Fig. 3.6). Although there was overlap in the DAPC , group one consisted entirely of phenotypic winter run samples from the CCV, while group three consisted entirely of NZ origin samples. Groups two, four, and five were largely a mix of phenotypic spring, fall, and late fall run from the CCV (Fig. 3.6). The limited resolution of groups 2-5 is likely due to the low number of SNPs in the dataset. When examining the posterior probabilities of the DAPC, we saw a clear demarcation between winter run, NZ rivers, and the other groups at K = 3 (BIC = 909). As we increased K, the fall, spring, and late fall run groups became less 65 clear, however the separation of winter run and NZ groups remained, and we began to see the fall run groups looking distinct from spring run groups (Fig. 3.7). Figure 3.2: Results from a PCA of the Chinook salmon from NZ analyzed in this study, with axes corresponding to PC 1 and 2. Individuals are colored based on river origin using a dataset with filtered SNPs for only NZ (n = 82, SNPs = 1774). 66 Figure 3.3: Results from a DAPC comparing NZ Chinook salmon. The figure shows separation based on the most informative linear discriminant for the top panel (K = 2) and the bottom panel (K =3). 67 Figure 3.4: Results from posterior probabilities of DAPC analysis comparing NZ Chinook salmon. The figure shows results for K = 2 (top panel), K = 3 (middle panel), and K = 4 (bottom panel). Vertical bars represent individuals and are color coded based on their proportion of membership to a particular DAPC group as illustrated in Figure 3. Individuals are organized by the NZ river of origin. 68 Figure 3.5: Results from a PCA of the Chinook salmon from NZCA dataset, with axes corresponding to PC 1 and 2. Individuals are colored based on run-timing or river origin (NZ individuals) filtered SNPs comparing NZ and CCV (n = 593, SNPs = 131). A) contains all individuals from the NZCA dataset in PC space together B) CCV samples, and C) NZ samples. 69 Figure 3.6: Scatter plot results of a DAPC comparing the NZ vs CCV samples. Group 1 is composed entirely of NZ individuals while group 3 is composed entirely of Winter Run CCV samples. Groups 2, 4, and 5 are a mix of Spring, Fall, and Late Fall run individuals. 70 Figure 3.7: Results from posterior probabilities of DAPC analysis comparing NZ and CCV Chinook salmon. The figure shows results for A) K = 3, B) K = 4, and C) K = 5. Vertical bars represent individuals and are color coded based on their proportion of membership to a particular DAPC group, organized by river origin, and where available, run-timing (CCV individuals). 71 Metrics of genetic diversity among populations and drainages Measures of genetic diversity among all populations in NZ were largely very similar (n = 89, SNPs = 1774). Populations in New Zealand both grouped by river as well as the DAPC groups reported similar high allelic richness as well as observed and expected heterozygosity (Table 3.3). Estimates of the inbreeding coefficients were also very similar and all were low and negative. When comparing statistics of genetic diversity in the NZCA dataset, we see different trends. Because genetic information was limited and did not illustrate the fine scale population structure that is known to exist in the CCV, we first examined populations as defined by known run timings (except in the case of NZ salmon, which were retained as their own group) (Meek et al. 2019b). This showed that Winter run had the lowest heterozygosity and allelic richness, and the highest inbreeding coefficient values, which is as expected given the low contemporary population size of Winter run (Table 3.4). New Zealand had the second lowest heterozygosity and allelic richness values, although those values were much higher than the Winter run group. When comparing levels of genetic differentiation among groups where K = 3, (Table 3.5), we observed relatively low Fst between the groups that contained mainly Rangitata and Waitaki individuals, with a higher level of genetic variation between Rakaia and both the Waitaki and Rangitata rivers. When comparing DAPC groups of K = 2, the differentiation between NZ groups was very similar, largely driven by the difference between some Rakaia individuals and all the other samples (Table 3.6). No private alleles were found in any of the populations in either analysis on either dataset, NZO or NZCA. When comparing NZ and CCV populations and drainages (n=593, SNPs = 131), there were a range of population differentiation scores. In each comparison, Winter run diverged most from NZ drainages, but also from Spring, Fall, and Late Fall run (Table 3.7). The smallest Fst value was between Fall and Late Fall groups, while the 2nd largest Fst values existed between NZ, Late Fall run, and Spring run (after Winter run comparisons). 72 Table 3.3: Heterozygosity (HO=observed and He=expected), inbreeding coefficients (FIS), and allelic richness (aR) for NZO dataset with n=82 and 1774 SNPs. Groups were based first on the K = 3 DAPC groups, named by their majority composition of the 3 river sites in NZ and second on the groups K = 2 groups DAPC assigned, followed by the total metric when the population was considered as a whole. N = sample size. Group Rangitata (K = 3) Rakaia (K = 3) Waitaki (K = 3) Rakaia (K = 2) Rangitata & Waitaki (K = 2) Total N 43 15 24 15 67 82 aR 1.99 1.99 1.99 1.97 1.98 1.99 HO He 0.369 0.355 0.344 0.342 0.360 0.350 0.363 0.341 0.362 0.352 0.344 0.331 FIS -0.039 -0.005 -0.030 -0.027 -0.030 -0.005 Table 3.4: Heterozygosity (HO,=observed and He,=expected), inbreeding coefficients (FIS,), and allelic richness (aR) for NZCA dataset n = 593 and 131 SNPs. Groups were based on known run- timing phenotype in the CCV compared to one group of all NZ individuals. N = sample size. Group Fall Late Fall Spring Winter New Zealand N 315 37 127 29 85 aR 1.61 1.59 1.62 1.45 1.52 HO 0.188 0.185 0.191 0.139 0.178 He 0.184 0.183 0.190 0.140 0.166 FIS -0.019 0.005 -0.002 0.019 -0.063 Table 3.5: Fst Estimates for the NZ rivers where n=89 and 1774 SNPs. Groups for Fst statistics are based on NZ samples grouped by DAPC K = 3 but named as the major tributaries that the majority of the individuals originated from. NZ Fst Estimates - K =3 groups River Rangitata Rangitata Rakaia - - Waitaki 0.0125 0.0289 Rakaia 0.021 - 73 Table 3.6: Fst Estimates for the NZ rivers where n=89 and 1774 SNPs were used. The Fst statistic is based on NZ samples grouped by DAPC (K = 2) but named as the major tributaries that the majority of the individuals in that group originated from. NZ Fst Estimates - 2 Groups Group DAPC Group 2 (Rangitata and Waitaki) DAPC Group 1 (Rakaia) 0.021 Table 3.7: Fst Estimates for CCV vs NZ groups where n = 593 and 131 SNPs were used. Groups were based on known run-timing phenotype in the CCV as compared to one group of all NZ individuals. NZ vs CCV Fst Estimates - 5 groups Group Fall Group Late Fall Group Spring Group Winter Group NZ Group Fall Group Late Fall Group Spring Group Winter Group - - - - DISCUSSION 0.014 - - - 0.019 0.031 - - 0.153 0.165 0.126 - 0.093 0.102 0.096 0.223 Our results show that NZ Chinook salmon have diverged from each other and from CCV Chinook salmon while also maintaining relatively high levels of heterozygosity and allelic richness. Previous work has explored the divergence between NZ and CCV populations, but this is the first time divergence has been demonstrated amongst all run-timing populations within the CCV and with new higher resolution genomic tools (Kinnison et al. 2002). Additionally, this is the first study to include Rangitata to explore differentiation among NZ populations. Our results show there is some evidence for differentiation among NZ river drainages, and that all drainages have similar levels of allelic richness, and heterozygosity compared to CCV spring and fall runs, and low levels of inbreeding. We found that all NZO groups have similar levels of heterozygosity and allelic richness to one another. There was some genetic differentiation between NZ rivers based on Fst values, although it was relatively low. Our results show that Chinook from the Rangitata and Waitaki rivers are most genetically similar, which is perhaps not unexpected as these rivers are next to 74 each other geographically and the Waitaki is where the populations were first introduced. The Rakaia is the furthest north and appears to be the least genetically similar to both the Rangitata and Waitaki. Therefore, there may be a pattern of isolation by distance occurring along the NZ coastline, which will also be an interesting avenue for future study by incorporating additional NZ sampling sites. Using a limited SNP dataset, we found evidence for genetic distinction between CCV and NZ Chinook salmon, though our analyses were limited by the small SNP panel size and were not able to identify which population from the CCV served as the source for the NZ introduction. We do see evidence for distinction of NZ Chinook from CCV Chinook, and that there may be genetic diversity present in these populations that is not found in the CCV, highlighting their potential usefulness in future genetic rescue efforts. In order to further explore which runs and populations served as the source for the original NZ introduction, further study should be done that includes a much greater number of SNPs that are common among the NZ Chinook and CCV Chinook populations and use samples from other individuals in other rivers. This will be an exciting avenue for future study. Because our analysis of the NZCA dataset only contained 131 SNPs, it is possible we did not have the statistical power to distinguish populations, especially considering what is already known about genetically distinct populations in the CCV. However, when comparing the NZ and the CCV, we also found that there is a high level of differentiation between these populations, but most strikingly, that differentiation of Winter run Chinook could still be detected at a low number of loci. The differentiation was highest between Winter run and the combined NZ populations, perhaps indicating preliminarily that the NZ fish did not originate from Winter run, or at least are the least genetically similar to current day populations of Winter run. Interestingly, the lowest Fst value between any CCV group and the NZ group was between the demographically Fall run group and NZ, which is not unexpected as Winter run have been shown to be the most genetically divergent from other CCV run types (Meek et al. 2014; Thompson et al. 2020; O’Leary et al. 2021). Additionally, Winter run was the most divergent population, showing the largest Fst values when compared to any populations, even when NZ and other CCV populations were compared. This indicates that although they have diverged, they are not so diverged that they are substantially different from CCV populations and are a good candidate for exploring rematriation and genetic rescue. 75 Although we did not have the resolution to fully address the cause of these patterns in the NZ Chinook, they are likely driven by genetic drift, adaptation, or a combination of both. At the time of the introduction into NZ, populations of Chinook in the CCV were likely more genetically diverse than today, since they have since faced massive population declines due to anthropogenic factors such as overfishing, dams, and extensive urbanization (Fisher 1992; Yoshiyama et al. 1998; Fisher 2016). This also means CCV populations of Chinook salmon may have diverged simply due to genetic bottlenecks, since population declines have been so severe (Bartley and Gall 1990; Meek et al. 2016). This is particularly true in the case of Winter run, as they may have diverged because they have undergone the most extreme bottleneck, leading to effective population size estimates as low as 174 (well below the recommended value of 500 needed to reduce genetic drift, and swiftly approaching the value of 50 recommended to avoid inbreeding depression) (Franklin 1980; M 1980; Hedrick et al. 2000; O’Leary et al. 2021). It is additionally possible that NZ populations underwent a founder effect due to a small founding population, causing them to look genetically dissimilar to modern CCV populations, which there is some evidence to support this from prior research (Nei et al. 1975; Barton and Charlesworth 1984; Quinn et al. 2001). All of these factors could be leading to the pattern of divergence we see between the populations in this dataset. It is possible that divergence in NZ Chinook means they have evolved different coadapted gene complexes due to their isolation, potential effects of drift, and the novelty of the NZ rivers. If this has occurred, it may limit their utility to recover CA populations. Strong signatures drift and/or selection could mean that these populations are less genetically viable, or have locally adapted to a different environment that will not benefit them when translocated (Templeton et al. 1986; Burton et al. 1999). Not only is NZ in a different hemisphere, meaning the introduced populations would have had to respond to new environmental cues, the rivers are all much shorter than the Sacramento River (Biggs et al. 1990; Jowett and Richardson 1996). Adaptation to a novel environment could mean that introduced NZ Chinook could have decreased survival and fitness or that cross breeding with native Chinook would result in outbreeding depression (Tallmon et al. 2004; Edmands 2007). However, the risks of outbreeding depression are markedly low for these populations given they have been separated for less than 500 years and have no fixed chromosomal differences (Frankham et al. 2011). In fact, for many species (including salmonids), the benefits of avoiding inbreeding and therefore lower fitness far 76 outweigh the risks and effects of outbreeding depression, even among subspecies (Hedrick and Fredrickson 2010; Johnson et al. 2010; Lehnert et al. 2014; Wells et al. 2019; Pregler et al. 2023). The benefit of NZ rivers, however, is that they have been much less impacted by other anthropogenic changes such as damming and heavy urbanization compared to CCV Chinook. This is particularly relevant, as many Chinook populations in their native range have faced massive declines due to anthropogenic changes to their environment, resulting in less genetic variation within populations and therefore less adaptive capacity to respond to change (Weeder et al. 2005; Janowitz-Koch et al. 2019; Thompson et al. 2019). Collectively, this highlights the value in exploring NZ Chinook as possible sources for rematriation in the CCV. Adaptation can happen on scales more rapid than previously assumed (within one to two generations), and understanding how imperiled and introduced species can respond to change is of the utmost importance if we are going to manage them effectively (Christie et al. 2012; Willoughby et al. 2018). Although Chinook Salmon in NZ presumably started from relatively small founding populations highly susceptible to drift, they have colonized and maintained populations in several NZ rivers for over a century and exhibit a diverse set of phenotypes, including unique run–timing (Quinn et al. 2001). This is particularly relevant to locations in North America where some run-types are threatened and/or have become extirpated, and have undergone great losses of genetic diversity since being used for the NZ Chinook introduction (Healey 1994; O’Leary et al. 2021). Further understanding how NZ populations compare genetically to CCV populations can help inform how to proceed with future management efforts to restore CCV populations from their present day small population sizes. It will be important to select NZ populations with the best chance of success for rematiration efforts into historic spawning grounds above the Shasta Dam. Populations with a wide array of genetic diversity will likely be the most resilient to challenges faced as a result of rematriation (Whiteley et al. 2015). Populations with high genetic diversity are often more successful, largely due to increased size and higher reproductive success, but more diverse populations can also provide enhanced ecosystem services (Reynolds et al. 2012; Robinson et al. 2017; Fitzpatrick et al. 2020). We have shown here that the NZ drainages have important pockets of diversity that may allow them to be successful in a system where Chinook are imperiled and genetic diversity is dwindling. This genetic diversity is one important piece of the puzzle that can be used when selecting a source population for rematriation. 77 Our research shows that there is genetic diversity in NZ drainages that has so far been unaccounted for in other analyses. This diversity may be critical for restoring the genetic diversity of imperiled CCV Chinook, but also means that it is possible rapid adaptation and diversification may have taken place in New Zealand. This is promising for the success of imperiled populations in the CCV because this influx of genetic diversity could recover genetic health. Because Chinook genetic diversity has allowed for long-term species resilience and persistence in the face of ecosystem changes, maintaining that diversity is of paramount importance. By understanding the genetic diversity of populations, management agencies can gear objectives towards maintenance of genetic diversity in these systems to maximize biodiversity and overall species resilience. FUTURE DIRECTIONS This work paves the way for exciting next steps to compare populations in NZ and the CCV. Although the same enzyme was used at some point during the design of both populations, the SNPs used for the NZ samples were not explicitly designed at SNP positions with CCV populations in mind. The SNPs were designed for polymorphic sites found across the entire species native range of Chinook, and while potentially informative, did not match a high percentage of the sites found in the CCV populations. A greater amount of genomic information may allow us to tease apart the relationships and further understand the genetic diversity found in both populations and how it relates to each other. Therefore, our next steps are to conduct whole genome resequencing to further explore both signals of rapid adaptation across NZ populations, and also to further disentangle the possible source origin for the NZ introduction. 78 BIBLIOGRAPHY Adamack A, Gruber B. 2014. PopGenReport: simplifying basic population genetic analyses in R - Adamack - 2014 - Methods in Ecology and Evolution - Wiley Online Library. Methods Ecol Evol. 5:384–387. https://doi-org.proxy1.cl.msu.edu/10.1111/2041-210X.12158 Babraham Bioinformatics - FastQC A Quality Control tool for High Throughput Sequence Data. Available from: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/. [accessed 2023 Nov 6]. Baird NA, Etter PD, Atwood TS, Currey MC, Shiver AL, Lewis ZA, Selker EU, Cresko WA, Johnson EA. 2008. Rapid SNP discovery and genetic mapping using sequenced RAD markers. PLoS ONE. 3:. https://doi.org/10.1371/journal.pone.0003376 Bartley DM, Gall G a. E. 1990. Genetic Structure and Gene Flow in Chinook Salmon Populations of California. Trans Am Fish Soc. 119:55–71. https://doi.org/10.1577/1548- 8659(1990)119<0055:GSAGFI>2.3.CO;2 Barton NH, Charlesworth B. 1984. Genetic Revolutions, Founder Effects, and Speciation. Annu Rev Ecol Syst. 15:133–164. https://doi.org/10.1146/annurev.es.15.110184.001025 Beddington JR, Agnew DJ, Clark CW. 2007. Current problems in the management of marine fisheries. Science. 316:1713–1716. https://doi.org/10.1126/science.1137362 Bellard C, Bertelsmeier C, Leadley P, Thuiller W, Courchamp F. 2012. Impacts of climate change on the future of biodiversity. Ecol Lett. 15:365–377. https://doi.org/10.1111/j.1461-0248.2011.01736.x Biggs BJF, Duncan MJ, Jowett IG, Quinn JM, Hickey CW, Davies‐Colley RJ, Close ME. 1990. Ecological characterisation, classification, and modelling of New Zealand rivers: An introduction and synthesis. N Z J Mar Freshw Res. 24:277–304. https://doi.org/10.1080/00288330.1990.9516426 Bottom DL, Jones KK, Simenstad CA, Smith CL. 2009. Reconnecting social and ecological resilience in salmon ecosystems. Ecol. Soc. 14. https://doi.org/10.5751/ES-02734-140105 Bourret SL, Caudill CC, Keefer ML. 2016. Diversity of juvenile Chinook salmon life history pathways. Rev Fish Biol Fish. 26:375–403. https://doi.org/10.1007/s11160-016-9432-3 Brooks TM, Mittermeier RA, Da Fonseca GAB, Gerlach J, Hoffmann M, Lamoreux JF, Mittermeier CG, Pilgrim JD, Rodrigues ASL. 2006. Global biodiversity conservation priorities. Science. https://doi.org/10.1126/science.1127609 Burton RS, Rawson PD, Edmands S. 1999. Genetic Architecture of Physiological Phenotypes: Empirical Evidence for Coadapted Gene Complexes1. Am Zool. 39:451–462. https://doi.org/10.1093/icb/39.2.451 Catchen J, Hohenlohe PA, Bassham S, Amores A, Cresko WA. 2013. Stacks: An analysis tool set for population genomics. Mol Ecol. 22:3124–3140. https://doi.org/10.1111/mec.12354 79 Ceballos G, Ehrlich PR, Barnosky AD, García A, Pringle RM, Palmer TM. 2015. Accelerated modern human-induced species losses: Entering the sixth mass extinction. Sci Adv. https://doi.org/10.1126/sciadv.1400253 Chapin FS, Zavaleta ES, Eviner VT, Naylor RL, Vitousek PM, Reynolds HL, Hooper DU, Lavorel S, Sala OE, Hobbie SE, Mack MC, Díaz S. 2000. Consequences of changing biodiversity. Nature. https://doi.org/10.1038/35012241 Christensen KA, Leong JS, Sakhrani D, Biagi CA, Minkley DR, Withler RE, Rondeau EB, Koop BF, Devlin RH. 2018. Chinook salmon (Oncorhynchus tshawytscha) genome and transcriptome. PLoS ONE. 13:1–15. https://doi.org/10.1371/journal.pone.0195461 Christie MR, Marine ML, French RA, Blouin MS. 2012. Genetic adaptation to captivity can occur in a single generation. Proc Natl Acad Sci. 109:238–242. https://doi.org/10.1073/pnas.1111073109 Clarke SM, Henry HM, Dodds KG, Jowett TWD, Manley TR, Anderson RM, McEwan JC. 2014. A High Throughput Single Nucleotide Polymorphism Multiplex Assay for Parentage Assignment in New Zealand Sheep. PLOS ONE. 9:e93392. https://doi.org/10.1371/journal.pone.0093392 Dallman S, Ngo M, Laris P, Thien D. 2013. Political ecology of emotion and sacred space: The Winnemem Wintu struggles with California water policy. Emot Space Soc. 6:33–43. https://doi.org/10.1016/j.emospa.2011.10.006 Danecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, Handsaker RE, Lunter G, Marth GT, Sherry ST, McVean G, Durbin R, 1000 Genomes Project Analysis Group. 2011. The variant call format and VCFtools. Bioinformatics. 27:2156–2158. https://doi.org/10.1093/bioinformatics/btr330 Edmands S. 2007. Between a rock and a hard place: evaluating the relative risks of inbreeding and outbreeding for conservation and management. Mol Ecol. 16:463–475. https://doi.org/10.1111/j.1365-294X.2006.03148.x Fisher FW. 1992. Chinook salmon, Oncorhynchus tshawytscha, growth and occurrence in the Sacramento-San Joaquin river system. Calif Dep Fish Game Inland Fish Div Red Bluff Calif. Fisher FW. 2016. Past and Present Status of Central Valley Chinook Salmon. 8:870–873. Fitzpatrick SW, Bradburd GS, Kremer CT, Salerno PE, Angeloni LM, Funk WC, Fitzpatrick SW, Bradburd GS, Kremer CT, Salerno PE, Angeloni LM. 2020. Genomic and Fitness Consequences of Genetic Rescue in Wild Populations Report Genomic and Fitness Consequences of Genetic Rescue in Wild Populations. Curr Biol. 30:1–6. https://doi.org/10.1016/j.cub.2019.11.062 Frankham R. 2005. Genetics and extinction. Biol Conserv. 126:131–140. https://doi.org/10.1016/j.biocon.2005.05.002 Frankham R. 2015. Genetic rescue of small inbred populations: meta-analysis reveals large and consistent benefits of gene flow. Mol Ecol. 24:2610–2618. https://doi.org/10.1111/mec.13139 80 Frankham R. 2016. Genetic rescue benefits persist to at least the F3 generation, based on a meta- analysis. Biol Conserv. 195:33–36. https://doi.org/10.1016/j.biocon.2015.12.038 Frankham R, Ballou JD, Eldridge MDB, Lacy RC, Ralls K, Dudash MR, Fenster CB. 2011. Predicting the Probability of Outbreeding Depression. Conserv Biol. 25:465–475. https://doi.org/10.1111/j.1523-1739.2011.01662.x Franklin IR. 1980. Evolutionary change in small populations. Sunderland, Massachusetts, Sinauer Associates, U.S.A. [accessed 2023 Nov 28]. Available from: https://publications.csiro.au/rpr/pub?list=BRO&pid=procite:46c4045a-9333-43d5-973b- 6f55ad9a9cc8 Gray RRR. 2022. Rematriation: Ts’msyen Law, Rights of Relationality, and Protocols of Return. Native Am Indig Stud. 9:1–27. Gruber B, Unmack PJ, Berry OF, Georges A. 2018. dartr: An r package to facilitate analysis of SNP data generated from reduced representation genome sequencing. Mol Ecol Resour. 18:691–699. https://doi.org/10.1111/1755-0998.12745 Healey M c. 1994. Variation in the Life History Characteristics of Chinook Salmon and Its Relevance to Conservation of the Sacramento Winter Run of Chinook Salmon. Conserv Biol. 8:876–877. https://doi.org/10.1046/j.1523-1739.1994.08030863-7.x Hecht BC, Matala AP, Hess JE, Narum SR. 2015. Environmental adaptation in Chinook salmon (Oncorhynchus tshawytscha) throughout their North American range. Mol Ecol. 24:5573–5595. https://doi.org/10.1111/mec.13409 Hedrick PW, Fredrickson R. 2010. Genetic rescue guidelines with examples from Mexican wolves and Florida panthers. Conserv Genet. 11:615–626. https://doi.org/10.1007/s10592-009-9999-5 Hedrick PW, Hedgecock D, Hamelberg S, Croci SJ. 2000. The impact of supplementation in winter-run chinook salmon on effective population size. J Hered. 91:112–116. https://doi.org/10.1093/jhered/91.2.112 Hilborn R, Quinn TP, Schindler DE, Rogers DE. 2003. Biocomplexity and fisheries sustainability. Proc Natl Acad Sci. 100:6564–6568. https://doi.org/10.1073/pnas.1037274100 Houck DL. 2019. Salmon Repatriation: One Tribe’s Battle to Maintain Its Culture and Spiritual Connection to Place. Nat Resour Environ. 34:23. How the Winnemem Winto won their ancestral land back and help save Chinook Salmon - Vox. Available from: https://www.vox.com/climate/23906426/winnemem-wintu-land-back- run4salmon-chinook-california-indigenous-peoples-rights-sovereignty. [accessed 2023 Nov 24]. Intergovernmental Panel on Climate Change. 2014. Climate Change 2014 Synthesis Report - IPCC. IUCN. 2017. IUCN Red List of Threatened Species. In: Version 20171. Available from: www.iucnredlist.org. 81 Janowitz-Koch I, Rabe C, Kinzer R, Nelson D, Hess MA, Narum SR. 2019. Long-term evaluation of fitness and demographic effects of a Chinook Salmon supplementation program. Evol Appl. 12:456–469. https://doi.org/10.1111/eva.12725 Johnson WE, Onorato DP, Roelke ME, Land ED, Cunningham M, Belden RC, McBride R, Jansen D, Lotz M, Shindle D, Howard J, Wildt DE, Penfold LM, Hostetler JA, Oli MK, O’Brien SJ. 2010. Genetic Restoration of the Florida Panther. Science. 329:1641–1645. https://doi.org/10.1126/science.1192891 Jombart T. 2008. adegenet: a R package for the multivariate analysis of genetic markers. Bioinforma Oxf Engl. 24:1403–1405. https://doi.org/10.1093/bioinformatics/btn129 Jombart T, Ahmed I. 2011. adegenet 1.3-1: new tools for the analysis of genome-wide SNP data. Bioinforma Oxf Engl. 27:3070–3071. https://doi.org/10.1093/bioinformatics/btr521 Jowett IG, Richardson J. 1996. Distribution and abundance of freshwater fish in New Zealand rivers. N Z J Mar Freshw Res. 30:239–255. https://doi.org/10.1080/00288330.1996.9516712 Kamvar ZN, Tabima JF, Grünwald NJ. 2014. Poppr: an R package for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ. 2:e281. https://doi.org/10.7717/peerj.281 Kinnison MT, Bentzen P, Unwin MJ, Quinn TP. 2002. Reconstructing recent divergence : evaluating nonequilibrium population structure in New Zealand Chinook salmon Reconstructing recent divergence : evaluating nonequilibrium population structure in New Zealand chinook salmon. Mol Ecol. 11:739–754. https://doi.org/10.1046/j.1365- 294X.2002.01477.x Lehnert SJ, Love OP, Pitcher TE, Higgs DM, Heath DD. 2014. Multigenerational outbreeding effects in Chinook salmon (Oncorhynchus tshawytscha). Genetica. 142:281–293. https://doi.org/10.1007/s10709-014-9774-5 Leonard K, David-Chavez D, Smiles D, Jennings L, Alegado R ʻAnolani, Manitowabi J, Arsenault R, Begay RL, Davis DD. 2023. Water Back: A Review Centering Rematriation and Indigenous Water Research Sovereignty. 16:. Li H, Durbin R. 2009. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics. 25:1754–1760. https://doi.org/10.1093/bioinformatics/btp324 M S. 1980. Thresholds for survival : maintaining fitness and evolutionary potential. Comservation Biol Evol-Ecol Perspect. 151–169. Manichaikul A, Mychaleckyj JC, Rich SS, Daly K, Sale M, Chen W-M. 2010. Robust relationship inference in genome-wide association studies. Bioinformatics. 26:2867– 2873. https://doi.org/10.1093/bioinformatics/btq559 McDowall RM. 1994. The origins of New Zealand’s Chinook salmon, Oncorhynchus tshawytscha. Mar Fish Rev. 56:1–7. McKinney GJ, Seeb LW, Larson WA, Gomez‐Uchida D, Limborg MT, Brieuc MSO, Everett MV, Naish KA, Waples RK, Seeb JE. 2016. An integrated linkage map reveals candidate 82 genes underlying adaptive variation in Chinook salmon ( Oncorhynchus tshawytscha ). Mol Ecol Resour. 16:769–783. https://doi.org/10.1111/1755-0998.12479 McKinney GJ, Waples RK, Seeb LW, Seeb JE. 2017. Paralogs are revealed by proportion of heterozygotes and deviations in read ratios in genotyping-by-sequencing data from natural populations. Mol Ecol Resour. 17:656–669. https://doi.org/10.1111/1755- 0998.12613 Meek MH, Baerwald MR, Stephens MR, Goodbla A, Miller MR, Tomalty KMH, May B. 2016. Sequencing improves our ability to study threatened migratory species: Genetic population assignment in California’s Central Valley Chinook salmon. Ecol Evol. 6:7706–7716. https://doi.org/10.1002/ece3.2493 Meek MH, Stephens MR, Goodbla A, May B, Baerwald MR. 2019a. Identifying hidden biocomplexity and genomic diversity in Chinook salmon, an imperiled species with a history of anthropogenic influence. Can J Fish Aquat Sci. https://doi.org/10.1139/cjfas- 2019-0171 Meek MH, Stephens MR, Goodbla A, May B, Baerwald MR. 2019b. Identifying hidden biocomplexity and genomic diversity in Chinook salmon, an imperiled species with a history of anthropogenic influence. Can J Fish Aquat Sci. https://doi.org/10.1139/cjfas- 2019-0171 Meek MH, Stephens MR, Wong AK, Tomalty KM, May B, R. BM. 2014. Genetic characterization of California’s Central Valley chinook salmon. Ecology. 95:1431. Miller MR, Dunham JP, Amores A, Cresko WA, Johnson EA. 2007. Rapid and cost-effective polymorphism identification and genotyping using restriction site associated DNA (RAD) markers. Genome Res. 17:240–248. https://doi.org/10.1101/gr.5681207 Moore JW, Yeakel JD, Peard D, Lough J, Beere M. 2014. Life-history diversity and its importance to population stability and persistence of a migratory fish: Steelhead in two large North American watersheds. J Anim Ecol. https://doi.org/10.1111/1365-2656.12212 Narum SR, Gallardo P, Correa C, Matala A, Hasselman D, Sutherland BJG, Bernatchez L. 2017. Genomic patterns of diversity and divergence of two introduced salmonid species in Patagonia, South America. Evol Appl. 10:402–416. https://doi.org/10.1111/eva.12464 Narum SR, Genova AD, Micheletti SJ, Maass A. 2018. Genomic variation underlying complex life-history traits revealed by genome sequencing in Chinook salmon. Proc R Soc B Biol Sci. 285:. https://doi.org/10.1098/rspb.2018.0935 National Research Council. 2004. Endangered and Threatened Fishes in the Klamath River Basin: Causes of Decline and Strategies for Recovery. Washington, D.C.: National Academies Press. https://doi.org/10.17226/10838 Nei M. 1978. Estimation of Average Heterozygosity and Genetic Distance from a Small Number of Individuals. Genetics. 89:583–590. Nei M, Maruyama T, Chakraborty R. 1975. The Bottleneck Effect and Genetic Variability in Populations. Evolution. 29:1–10. https://doi.org/10.2307/2407137 83 November 2, 2023. LGC Group. In: lgcgroup. Available from: https://www.lgcgroup.com/. [accessed 2023 Nov 10]. O’Leary SJ, Thompson TQ, Meek MH. 2021. Every cog and wheel: Unraveling biocomplexity at the genomic and phenotypic level in a population complex of Chinook salmon. 2021.03.26.437213. https://doi.org/10.1101/2021.03.26.437213 O’Malley KG, Camara MD, Banks MA. 2007. Candidate loci reveal genetic differentiation between temporally divergent migratory runs of Chinook salmon (Oncorhynchus tshawytscha). Mol Ecol. 16:4930–4941. https://doi.org/10.1111/j.1365- 294X.2007.03565.x Oncorhynchus tshawytscha genome assembly Otsh_v2.0. In: NCBI. Available from: https://www.ncbi.nlm.nih.gov/data-hub/assembly/GCF_018296145.1/. [accessed 2023 Nov 6]. Pregler KC, Obedzinski M, Gilbert-Horvath EA, White B, Carlson SM, Garza JC. 2023. Assisted gene flow from outcrossing shows the potential for genetic rescue in an endangered salmon population. Conserv Lett. 16:e12934. https://doi.org/10.1111/conl.12934 Pyšek P, Richardson DM. 2010. Invasive Species, Environmental Change and Management, and Health. https://doi.org/10.1146/annurev-environ-033009-095548 Quinn TP. 2018. The Behavior and Ecology of Pacific Salmon and Trout. https://doi.org/10.1111/j.1467-2979.2006.00203.x Quinn TP, Kinnison MT, Unwin MJ. 2001. Evolution of chinook salmon (Oncorhynchus tshawytscha) populations in New Zealand: Pattern, rate, and process. Genetica. 492–513. https://doi.org/10.1023/A Raheema N, Talberth J, Colt S, Fleishman E, Swedeen P, Boyle KJ, Rudd M, Lopez RD, O’Higgins T, Willer C, Boumans RM. 2009. The Economic Value of Coastal Ecosystems in California. Ecol Econ. Reynolds LK, McGlathery KJ, Waycott M. 2012. Genetic Diversity Enhances Restoration Success by Augmenting Ecosystem Services. PLOS ONE. 7:e38397. https://doi.org/10.1371/journal.pone.0038397 Rivera-Colón AG, Catchen J. 2021. Population genomics analysis with RAD, reprised: Stacks 2. 2021.11.02.466953. https://doi.org/10.1101/2021.11.02.466953 Robinson ZL, Coombs JA, Hudy M, Nislow KH, Letcher BH, Whiteley AR. 2017. Experimental test of genetic rescue in isolated populations of brook trout. Mol Ecol. 26:4418–4433. https://doi.org/10.1111/mec.14225 Rochette NC, Catchen JM. 2017. Deriving genotypes from RAD-seq short-read data using Stacks. Nat Protoc. 12:2640–2659. https://doi.org/10.1038/nprot.2017.123 Sadovy de Mitcheson Y, Craig MT, Bertoncini AA, Carpenter KE, Cheung WWL, Choat JH, Cornish AS, Fennessy ST, Ferreira BP, Heemstra PC, Liu M, Myers RF, Pollard DA, Rhodes KL, Rocha LA, Russell BC, Samoilys MA, Sanciangco J. 2013. Fishing groupers 84 towards extinction: A global assessment of threats and extinction risks in a billion dollar fishery. Fish Fish. 14:119–136. https://doi.org/10.1111/j.1467-2979.2011.00455.x Tallmon DA, Luikart G, Waples RS. 2004. The alluring simplicity and complex reality of genetic rescue. Trends Ecol Evol. 19:489–496. https://doi.org/10.1016/j.tree.2004.07.003 Templeton AR, Hemmer H, Mace G, Seal US, Shields WM, Woodruff DS. 1986. Local adaptation, coadaptation, and population boundaries. Zoo Biol. 5:115–125. https://doi.org/10.1002/zoo.1430050206 Thompson NF, Anderson EC, Clemento AJ, Campbell MA. 2020. A complex phenotype in salmon controlled by a simple change in migratory timing. 613:609–613. Thompson TQ, Renee Bellinger M, O’Rourke SM, Prince DJ, Stevenson AE, Rodrigues AT, Sloat MR, Speller CF, Yang DY, Butler VL, Banks MA, Miller MR. 2019. Anthropogenic habitat alteration leads to rapid loss of adaptive variation and restoration potential in wild salmon populations. Proc Natl Acad Sci U S A. https://doi.org/10.1073/pnas.1811559115 Tilman D. 2009. Habitat destruction and extinction debt. Nature. https://doi.org/10.1016/j.mbs.2009.06.003 Vrijenhoek RC. 1994. Genetic diversity and fitness in small populations. In: Loeschcke V, Jain SK, Tomiuk J, editors. Conservation Genetics. Basel: Birkhäuser. p. 37–53. https://doi.org/10.1007/978-3-0348-8510-2_5 Weeder JA, Marshall AR, Epifanio JM. 2005. An Assessment of Population Genetic Variation in Chinook Salmon from Seven Michigan Rivers 30 Years after Introduction. North Am J Fish Manag. 25:861–875. https://doi.org/10.1577/M03-227.1 Wells ZRR, Bernos TA, Yates MC, Fraser DJ. 2019. Genetic rescue insights from population- and family-level hybridization effects in brook trout. Conserv Genet. 20:851–863. https://doi.org/10.1007/s10592-019-01179-z Whiteley AR, Fitzpatrick SW, Funk WC, Tallmon DA. 2015. Genetic rescue to the rescue. Trends Ecol Evol. 30:42–49. https://doi.org/10.1016/j.tree.2014.10.009 Williams JG. 2006. Central Valley salmon: a perspective on Chinook and steelhead in the Central Valley of California. San Franc Estuary Watershed Sci. 4:1–393. https://doi.org/10.5811/westjem.2011.5.6700 Willoughby JR, Harder AM, Tennessen JA, Scribner KT, Christie MR. 2018. Rapid genetic adaptation to a novel environment despite a genome-wide reduction in genetic diversity. Mol Ecol. 0–2. https://doi.org/10.1111/mec.14726 Worm B, Hilborn R, Baum JK, Branch TA, Collie JS, Costello C, Fogarty MJ, Fulton EA, Hutchings JA, Jennings S, Jensen OP, Lotze HK, Mace PM, McClanahan TR, Minto C, Palumbi SR, Parma AM, Ricard D, Rosenberg AA, Watson R, Zeller D. 2009. Rebuilding Global Fisheries. Science. 325:578–585. https://doi.org/10.1126/science.1173146 85 Yoshiyama RM. 1999. A History of Salmon and People in the Central Valley Region of California. Rev Fish Sci. 7:197–239. https://doi.org/10.1080/10641269908951361 Yoshiyama RM, Fisher FW, Moyle PB. 1998. Historical Abundance and Decline of Chinook Salmon in the Central Valley Region of California. North Am J Fish Manag. 18:487–521. https://doi.org/10.1577/1548-8675(1998)018<0487 86