(Astragalus Jaegerianus Munz, Fabaceae) with Limited Range Using Microsatellites

California State University, San Bernardino CSUSB ScholarWorks

Electronic Theses, Projects, and Dissertations Office of aduateGr Studies

12-2020

POPULATION STRUCTURE OF A FEDERALLY ENDANGERED PLANT (ASTRAGALUS JAEGERIANUS MUNZ, FABACEAE) WITH LIMITED RANGE USING MICROSATELLITES

SueAnn Neal

Follow this and additional works at: https://scholarworks.lib.csusb.edu/etd

Part of the Biodiversity Commons, Desert Ecology Commons, Genetics Commons, Genomics Commons, Molecular Biology Commons, Molecular Genetics Commons, Other Ecology and Evolutionary Biology Commons, and the Population Biology Commons

Recommended Citation Neal, SueAnn, "POPULATION STRUCTURE OF A FEDERALLY ENDANGERED PLANT (ASTRAGALUS JAEGERIANUS MUNZ, FABACEAE) WITH LIMITED RANGE USING MICROSATELLITES" (2020). Electronic Theses, Projects, and Dissertations. 1150. https://scholarworks.lib.csusb.edu/etd/1150

This Thesis is brought to you for free and open access by the Office of aduateGr Studies at CSUSB ScholarWorks. It has been accepted for inclusion in Electronic Theses, Projects, and Dissertations by an authorized administrator of CSUSB ScholarWorks. For more information, please contact [email protected]. POPULATION STRUCTURE OF A FEDERALLY ENDANGERED PLANT

(ASTRAGALUS JAEGERIANUS MUNZ, FABACEAE) WITH LIMITED RANGE

USING MICROSATELLITES

A Thesis

Presented to the

Faculty of

California State University,

San Bernardino

In Partial Fulfillment

of the Requirements for the Degree

Master of Science

Biology

SueAnn M. Neal

December 2020 POPULATION STRUCTURE OF A FEDERALLY ENDANGERED PLANT

(ASTRAGALUS JAEGERIANUS MUNZ, FABACEAE) WITH LIMITED RANGE

USING MICROSATELLITES

A Thesis

Presented to the

Faculty of

California State University,

San Bernardino

SueAnn M. Neal

December 2020

Approved by:

Dr. Anthony Metcalf, Committee Chair, Biology

Dr. James Ferrari, Committee Member

Dr. David Polcyn, Committee Member

ABSTRACT

Studies on population genetics examine the relationship and effects of population structure, migration, gene flow and demographic history, and are therefore important in the conservation of endangered species. Astragalus jaegerianus, a critically federally endangered species found in a geographically restricted range is investigated to determine population structure and genetic variation. Previous research on A. jaegerianus focused on DNA sequence data for cpDNA and nrDNA showed no variation. Further research on A. jaegerianus utilizing AFLP’s on the whole genome indicated substantial gene diversity and population structure consistent with geographically widespread species. AFLP research is a cost-effective process to identify levels of genetic diversity but does not allow determination of heterozygosity or homozygosity. However, microsatellite analysis identifies tandem base pair repeats, which are co- dominant markers, and heterozygotes can be determined from homozygotes.

Through Illumina paired-end shotgun sequencing, 26 microsatellites were isolated and characterized for A. jaegerianus. Loci were screened across 24 individuals in 5 subpopulations. The 26 identified microsatellites underwent preliminary screening to determine the 7 optimal primers for this study.

Of the 176 initial samples, 140 individuals successfully amplified across a minimum of 4 microsatellite loci. Microsatellite analysis supports Walker and

Metcalf’s (2008) findings of population structure and genetic variation comparable to geographically widespread plant species. The RST values ranged

iii from 0.010 to 0.328, indicating that Brinkman Wash is the most distinct of the populations. In comparing the STRUCTURE analysis to the outcome of the

PCoA, the data indicate there are four distinct populations and that Coolgardie

Mesa and Lane Mountain appear to be more similar and can be viewed as a single population. Both AFLP and microsatellite analyses support the findings that Astragalus jaegerianus has maintained a high level of genetic diversity despite the limited range and reduced population size.

iv ACKNOWLEDGEMENTS

I would like to take this opportunity to thank the village that helped in the completion of this thesis project. First, I would like to thank my thesis advisor, Dr.

Anthony Metcalf, for agreeing to fund and support this project. I would also like to thank my committee members, Dr. James Ferrari and Dr. David Polcyn for agreeing to assist in the completion of my project. Thank you to George Walker for his previous research on the unique little plant that inspired me to take his research to the next level, and to Stacey Nerkowski for first bringing me into the lab and introducing me to an area of research that would lead me to my master’s thesis.

To my amazing family who put up with me through this whole process, although I feel I was very calm, cool and collected. Stop laughing, we all have our delusions. Thank you to my husband Greg Neal, the day I decided it was time to go back to school and find a new path he just rolled with it and told me he was behind me and always took the opportunity to tell me I was capable of great things. When I decided I wanted to go on for my Master’s degree he again rolled with it and was my biggest supporter. Even through the computer crashes, blown up papers and threats of throwing all electronic devices out the window and changing my major to underwater basket weaving he still had my back. He is my rock, my Tech Support, my biggest cheerleader and the best human being I know. I also want to thank my two heathens, Collin and Ethan. They too just rolled with the whole school process and even joined me in the lab for different

v parts of my experiments, most of the time even willingly. Thanks to my parents,

Don and Jackie Stine, for cheering me on and keeping my Starbucks card fully loaded. Caffeine makes the world go round when in grad school! Thanks to my mom for “dragging” on amazing vacations to celebrate my accomplishments, can’t wait for our next adventure.

I owe so much to my lab mates, especially over the last 3 years of this project. Liane Greaver for keeping me sane and laughing, especially when I hit the wall, and for reminding me to just keep swimming. When things went sideways (DNA Analyzers of DEATH, lost samples in the mail…twice) Liane talked me off the cliff and reminded me I don’t look good in prison orange. Caitlin

Hazelquist, aka Lab Lackey, you jumped in and helped me push through the tedious process of DNA extraction. I will forever be grateful I wasn’t the only one in the lab to accidentally knock over a sample or two. I also blame you for new found love of Botany gin, thanks for that! Lauren Morrison, aka Lab Elf, the little ray of sunshine in our lab, thank you for bringing your pure enjoyment of biology to the lab and taking the jokes and pranks we threw at you with a smile. I’ve been lucky to have such a truly amazing group to work with and I am forever grateful for their help and the many, many, many trips for coffee. Thank you to my Brunch

Squad, Dawn Ladd and Renee Bilbrew, for getting me out of my own head and smiling and nodding when I went on one of my school tangents and being the best cheerleaders a girl could ever ask for. Thanks to Bob Trees for always being on ‘Team Greg’, he needed someone on his side every now and then when I was

vi going crazy but seriously it’s time to switch to ‘Team Suzy’. I also owe a huge thanks to the two angels looking out for me; Traci Allen (Princess Bratty Pants

The Quilt Streaker) and Jennie Trees (my bestie). I miss you both every day.

Lastly, I would like to thank the following organizations for their financial support: California State University, San Bernardino’s Biology

Department; California State University, San Bernardino ASI, Inc.; California

State University, San Bernardino Office of Student and the Department Defense,

U. S. Army.

vii DEDICATION

This is dedicated to my husband, Greg Neal. You get me like no one ever could and you keep me grounded while still allowing me to soar. BEST

HUSBAND EVER.

TABLE OF CONTENTS

ABSTRACT ...... iii

ACKNOWLEDGEMENTS ...... v

CHAPTER ONE: INTRODUCTION ...... 1

Overview ...... 1

Geography and Habitat Fragmentation ...... 4

Population Genetics ...... 11

Genetic Structure: Fixation Index (FST) and Slatkin RST ...... 11

Hardy-Weinberg Equilibrium Model ...... 12

Gene Flow and Migration ...... 12

Isolation by Distance ...... 14

Methods for Estimating Genetic Diversity ...... 15

Amplified Fragment Length Polymorphism (AFLP) ...... 16

Microsatellites ...... 17

CHAPTER TWO: MATERIALS AND METHODS ...... 20

Specimen Collection ...... 20

Molecular Methods ...... 20

Microsatellite Acquisition ...... 23

Polymerase Chain Reaction Process ...... 24

Population Genetics ...... 26

Genetic Structure: Fixation Index (FST and RST) ...... 26

CHAPTER THREE: RESULTS ...... 29

Population Genetic Structure ...... 31

CHAPTER FOUR: DISCUSSION ...... 33

Genetic Variation within Astragalus jaegerianus ...... 33

Population Structure and Geographic Fragmentation ...... 35

Population Genetics ...... 37

Conclusions with Conservation Implications ...... 39

APPENDIX A: FIGURES ...... 42

APPENDIX B: TABLES ...... 50

APPENDIX C: MICROSATELLITE DATA ...... 68

APPENDIX D: INPUT FILES ...... 76

REFERENCES ...... 90

CHAPTER ONE

INTRODUCTION

Overview

The Lane Mountain milkvetch (Astragalus jaegerianus Munz) is a federally endangered endemic plant species that is geographically restricted to a small region in the Mojave Desert of Southern California (Appendix A, Figure 1).

Astragalus jaegerianus was first identified by Dr. Edmund Jaeger in 1939 and formally described and named by Philip Munz in 1941 (Barneby, 1964; Munz,

1941). Additional populations were not documented until the mid-1980’s (Kearns,

2003). Three areas of known plant populations were identified in the early 1990’s with an additional plant population documented in the early 2000’s. Census efforts in 2001 counted a population of 5,723 plants in an area that covered

21,350 acres.

Astragalus jaegerianus is a slender plant that is typically associated with a nurse shrub, which is used as a trellis, and grows between 30 to 70 centimeters

(cm) in height during a wet year. They have straggling stems with silvery linear leaflets and when in bloom, the flowers are cream to purple in color. The fruits that form are pencil-shaped pendulum pods with limited dispersal rates due to their size and limited dispersal pathways. Dispersal outside of the parental population location is attributed to granivores or high levels of water flow through the population area (Hohman, 2014; Prigge et al., 2012). In dry years, A. jaegerianus growth tends to be less than 7 cm in height, with some plants

1 exhibiting no growth above ground and believed to remain dormant for up to three years (Hohman, 2014). Researchers estimate the maximum lifespan of A. jaegerianus as 15 years, with only a small number of plants reaching the estimated maximum lifespan (Prigge et al., 2012). Gibson et al (1998) noted that the relationship between the Lane Mountain milkvetch and the host shrub may be mutualistic as the milkvetch is a nitrogen fixer creating a suitable habitat and available resources for nitrogen-fixing bacteria.

Astragalus jaegerianus can be found in the transition zone between creosote bush scrub and Joshua tree woodland areas in the Barstow area of

Southern California (Hohman, 2014). While found in the creosote bush scrub, the

Lane Mountain Milkvetch is primarily found in areas with very few or complete absence of creosote bush. Plants are typically found associated with white bursage (Ambrosia dumosa) and other medium sized shrubs (Baldwin et al.,

2012). Astragalus jaegerianus is found in shallow granitic or sandy soils in areas where bedrock is exposed or near the surface, between 3,000 and 4,000 feet in elevation (Walker & Metcalf, 2008) While the plants are predominantly found associated with a nurse bush that can be used as a trellis, there is no known preference for one particular species of nurse bush. Previous surveys noted an association of Lane Mountain Milkvetch and high density, low canopy shrub.

The plants are found in small groups that are typically separated by several hundred meters (Walker & Metcalf, 2008). These separations and the low dispersal distances of the propagules should lead to a reduced level of gene flow

2 within and among the populations as it is rare for propagules to disperse to adjacent populations. Walker and Metcalf (2008) researched A. jaegerianus in the Fort Irwin/Barstow area as five subpopulations: Goldstone, Brinkman Wash,

Prospectors Wash, Lane Mountain and Coolgardie Mesa Populations (Appendix

A, Figures 2-6). In later studies, the Lane Mountain and Coolgardie populations were combined as one subpopulation Coolgardie-Lane Mountain.

Limited genomic sequencing of chloroplast DNA (cpDNA) and nuclear

DNA (nDNA) analysis completed by Walker and Metcalf (2008), revealed no genetic diversity among individuals from each subpopulation. Amplified Fragment

Length Polymorphism (AFLP) markers using the whole genome showed levels of genetic diversity equivalent to species that are not geographically limited. The lack of genetic diversity in the sequenced markers could be attributed to the reduced population size or some historical event that reduced variation within the species. The higher level of genetic diversity found through AFLP markers was attributed to a recent population contraction from a larger population. Range-wide genetic variation was initially believed to follow that of other geographically limited populations, but research showed levels similar to that of geographically widespread species (Walker & Metcalf, 2008).

While AFLP markers were successful in establishing high levels of genetic variation within, and moderate population structure within and among A. jaegerianus populations, there are several challenges raised by this approach

(Walker, 2005). The first challenge is pragmatic, as the techniques used are time

3 consuming and require highly purified DNA. The second challenge is that AFLP markers are dominant and therefore direct measures of heterozygosity are not possible. In this study we used co-dominant microsatellite markers to re-examine genetic variation and population structure in this federally endangered species.

We further develop these ideas below.

Geography and Habitat Fragmentation

The type of geographic range a species occupies can have an effect on its ability to move beyond that range and the ability of segregated populations to exchange alleles. Factors that might be considered in understanding the effects of geographic distribution include, but are not limited to, a variety of abiotic and biotic factors such as the population dynamic of the species at the range limits, the patterns of growth of the population, individuals within the population, and the genetic variation within the population and at the range limits (Gaston, 2003).

Additionally, how a species reacts to environmental changes and stressors will play a role in reducing or expanding range limits for a particular species.

Environmental stressors can include human encroachment, invasive species and climate change. Overall, there are different abiotic and biotic limiting factors in a species ability to expand beyond or even maintain its current range (Avise, 2004;

Gaston, 2003).

Dispersal barriers due to habitat fragmentation are an important factor in a species ability to expand its range or allow for gene flow between populations.

Plants may be limited in their ability to naturally disperse beyond their given

4 range without assistance from outside sources such as granivores or human intervention (Eriksson, 1998; Young et al., 1996). Thymus serpyllum, commonly referred to as wild thyme, can be found growing in southeastern Sweden in semi- natural pastures and graves from the Iron Age. Plots were studied within the semi-natural pastures and graves, and areas where no graves were found but the soil conditions were considered similar. The plots were further divided into areas that were disturbed, undisturbed, where T. serpyllum were found and areas where they were absent. Eriksson (1998) showed that dispersal limitations were likely the limiting factor in the range distribution of T. serpyllum as the control plots showed no new growth while the disturbed plots, where there were previously no populations showed growth. Historical distribution was credited to the hispid spiky corolla that would attach to the fur of cows that roamed the pastures and were moved periodically between pastures transferring the seeds to a new area. Additionally, in the undisturbed plots the germination and survival rates of the T. serpyllum were sensitive to competition. The pastures where cows roamed tended to be trampled down allowing the T. serpyllum to germinate and grow with minimal competition (Eriksson, 1998).

Pierson and Mack (1990) looked at the occurrence of Bromus tectorum, an invasive grass commonly known as cheatgrass, in the grasslands (steppe) and mature forest areas of eastern Washington and northern Idaho (USA).

Bromus tectorum is found predominantly in the grasslands, with the population decreasing as elevation increases into the mature forested area. A demographic

5 study looked at the experimental introduction of B. tectorum into four different mature plant stands; Pinus, Pseudotsuga, Abies and Thuja. In comparison to the cheatgrass growing in the grasslands, the plants introduced to the mature forested areas were smaller and produced fewer seeds. The difference in growth and reproductive success was determined to be the result of environmental limitations as opposed to dispersal limitations. The forest provided a more shaded, cooler ground cover which is not optimal for the growth of the cheatgrass (Pierson & Mack, 1990).

How a species is geographically distributed is a contributing factor in its genetic structure. A species can be widely distributed across its range or partitioned into different populations; both can have an effect on genetic variation found within and among the population. Plant species are limited in the exchange of genetic material by dispersal ranges, with species found in areas of restricted geographic ranges tending to have limited genetic variation compared to species with widespread ranges (Richards et al., 1999). Factors that may affect the genetic variation and structure of rare/endangered species within a limited geographic range include genetic drift and directional selection. Habitat loss, over-exploitation of a species, pollution and introduction of a new species can all lead to reduced population size and increased fragmentation into small populations (Frankham, 1995). Rare species within limited geographic ranges can become susceptible to limited gene flow, partitioned into genetically different sub-populations and an increased risk of extinction due to both environmental

6 and demographic conditions (Lynch et al., 1995; Metcalf et al., 2001). In a review of endangered species which endured reduction in population size and geographic range, Avise (2004) indicated that while 32 of the 38 endangered species studied display reduced genetic variation, some endangered species were able to maintain diverse genetic variation (Avise, 2004; Frankham, 1995).

Habitat fragmentation is the changing of continuous habitat into segregated and isolated populations leading to individual subpopulations with limited or non-existent gene flow opportunities. Effects of geographic isolation on the genetics of plants include longevity, sexual versus asexual reproduction, pollen and seed dispersal and external interactions. Reduced allelic diversity due to habitat fragmentation creates a genetic bottleneck. Continued isolation with limited to no gene flow will lead to allelic reduction, with increased homozygosity.

Alleles that were originally found in lower frequency are shown to be the first lost in a bottleneck event (Young et al., 1996).

Astragalus cremnophylax var. cremnophylax, a federally listed critically endangered plant found in the Grand Canyon National park in Arizona, is characterized as an herbaceous perennial that can be found most commonly in limestone rock formations. Documentation shows that A. cremnophylax was abundant in the Grand Canyon over 100 years ago, but in recent years the population has declined and become geographically fragmented. There are three documented segregated patches of A. cremnophylax on the rim of the Grand

Canyon; two populations on the South Rim (SR1 and SR2) and one on the North

Rim (NR1). Population size varies between the three populations; SR1 census documented a population size of 500 plants, SR2 showed 2 plants and NR1 with the largest population of around 1000. The South Rim populations experienced the highest level of human impact leading to their lower population numbers

(Travis et al., 1996). Through AFLP research, Travis et al. (1996) determined the genetic diversity within each population varied directly with the population size.

As the populations become geographically segregated the gene flow between populations also becomes very rare. The severe reduction or elimination of gene flow between populations lead to a decrease in allelic variation within the two affected populations (SR1 and SR2), while NR1 was able to maintain a population size large enough that genetic variation was unaffected. The SR1 population showed a lower level of heterozygosity than was expected, which was attributed to population crashes leading to a severe lowering of the population in one event.

Walker and Metcalf (2008) applied the AFLP markers to determine the level of genetic diversity and population structure of A. jaegerianus. The levels of mean gene diversity (Nei, 1987) within the range of the species (0.2660), and population structure (global FST 0.133, p, 0.01) are consistent with geographically widespread species. Research of geographically widespread species such as the

European globeflower (Trollius europaeus L.) found the mean gene diversity among the three regions ranged from 0.158 to 0.299 (Despres et al., 2002) which is similar to what Walker and Metcalf (2008) found with A. jaegerianus. Research

8 by Travis et al. (1996) on the critically endangered Astragalus cremnophylax var. cremnophylax on the South Rim showed an extremely low level of genetic diversity with heterozygosity averages of 0.017 and 0.0373.

Previous research has shown that within the plant taxa there is typically a positive correlation between geographic range and gene diversity or between population size and gene diversity (Despres et al., 2002; Ellstrand & Elam, 1993;

Hamrick & Godt, 1990; Travis et al., 1996). The Walker and Metcalf (2008) study showed a significant positive correlation when comparing population density and gene diversity but did not find the same correlation when comparing gene density and population size nor area. There are several factors that may affect a plants ability to maintain a higher level of genetic variation within and among populations. There appears to be correlation between outcrossing species and genetic variation, along with exhibiting a high level of polymorphism. Life history of the species, such as seed and pollen dispersal patterns, fecundity, and outcrossing ability versus selfing, play an important role in genetic variation

(Hamrick et al. 1979). Endemic species tend to have smaller populations within limited geographic ranges which would typically lead to lower genetic variation but a look at Astragalus showed limited populations in limited geographic ranges exhibiting higher than expected levels of genetic diversity (Hamrick et al. 1979;

Karron, 1991; Travis et al 1996b).

The AFLP markers were successful in establishing evidence of correlation between population size and genetic variation, but they did not directly measure

9 heterozygosity or gene flow. Such measurements are necessary to fully elucidate the population structure, gene flow, and the pattern and history of genetic variation.

Astragalus bibullatus is a federally endangered legume endemic to the limestone cedar glades in Rutherford County, Tennessee (Baskauf and Burke,

2009). Residential expansion has led to a reduced landscape and geographic isolation and ultimately its critical status listing. Previous allozyme research showed low levels of polymorphism but AFLPs indicated higher levels of genetic diversity. While the AFLP research showed higher diversity, the levels are still lower than other plants found in the cedar glen area. Baskauf and Burke (2009) developed novel microsatellite loci to test on 193 individuals from 5 populations and found that the 5 populations were genetically similar. Principle Coordinate

Analysis (PCoA) results showed a significant overlap of individuals from the different populations. The results suggest there is limited differentiation among the populations.

Petunia secreta is a critically endangered rare endemic species found in

Brazil, first described in 1995 and formally described in 2005. It has a limited range with isolated discrete populations (Turchetto et al., 2016). Microsatellite analysis was performed on all known populations, comprised of 73 individuals, using 15 previously identified microsatellite loci. Petunia secreta presented a high level of genetic diversity that was equivalent to more widespread Petunia species found in the area. The maintenance of the high genetic variability was attributed

10 to the retention of ancient variability and a stable founder population. Limited gene flow between the fragmented populations is a concern for conservation efforts.

Population Genetics

Studies on population genetics examine the relationship and effects of population structure, migration, gene flow and demographic history. Allelic variation within populations can be influenced by various processes including evolutionary history, genetic drift, limitation of gene flow between and within populations and dispersal methods and rates (Panda et al., 2015). In the current study, statistical analysis will be applied to the data collected from the microsatellite analysis utilizing the identified primers for A. jaegerianus. The analysis will look at genetic diversity within and among the five identified populations. It will test for isolation by distance and the rate and pattern of migration (gene flow) between geographically segregated populations. First, I briefly review genetic diversity as it applies to this study.

Genetic Structure: Fixation Index (FST) and Slatkin’s RST

F-statistics were introduced by Sewall Wright as a way to calculate and describe population structure in natural populations (Weir, 2012). Sewall Wright made contributions toward the understanding of genetic variation in natural populations and the significance population size plays in maintaining variation

(Wright, 1931, 1965). In an effort to determine the effects of inbreeding in a group of natural populations Sewall Wright defined the fixation index, FST. F-statistics is

11 an informative statistical analysis that indicates reduced heterozygosity in comparison to an infinite population where random mating is occurring. The FST index ranges from 0-1, a value closer to the 0 range indicates there is no genetic divergence between subpopulations, that is breeding individuals can freely move among all populations. A value closer to the 1 range indicates that there is little to no gene flow between populations and an increase in homozygosity.

An analogue to FST, Slatkin’s RST, was developed as a statistical model to analyze population structure utilizing microsatellites. While the FST model assumes an infinite alleles model where all mutations are equally likely, the RST model accounts for Step-wise Mutation Model (SMM) which is characteristic in microsatellite loci (Balloux and Lugon-Moulin, 2002; Slatkin, 1995).

Hardy-Weinberg Equilibrium Model

The Hardy-Weinberg Equilibrium (HWE) model assumes no external or evolutionary forces and an infinite population size with one dominant allele and one recessive allele, as well as random mating. Allelic frequency can be used to calculate the genotypic frequency expected in a population (Hartl and Clark,

2006; Wittke-Thompson et al., 2005). Heterozygosity in a population can be an indicator of increased levels of gene flow among populations and therefore HWE serves as a null hypothesis for evolutionary change.

Gene Flow and Migration

To minimize the effects of reduced allelic diversity due to habitat fragmentation within plant populations, gene flow is mediated by pollination and

12 seed dispersal. Richard et al. (1999) looked at the immigration rate of Silene alba, a short-lived perennial plant. It was shown that immigration, noted as pollen fertilization outside the focal patch, decreased as distance increased. For patches at 20 meters from the focal patch the immigration rate was 47% overall as compared to 80 meters from the focal patch with an overall immigration rate of less than 6% overall. Seed dispersal rates and distance between populations play an important role in gene flow between populations (Richards et al., 1999).

For heavier seed pods, the distance the pods can disperse will depend on environmental factors such as wind and animal interactions. Subdividing a species range can lead to partitioning into distinct sub-populations which in turn can lead to genetic variation among populations (Avise, 2004; Segarra-Moragues et al. 2005). Additional research on the genetic structure of Silene alba

(McCauley, 1997) focused on mating systems and spatial distance between populations. Gene flow relied on two different avenues: seeds and pollen.

Plant density and proximity were found to play an important role in the attracting of pollinators in a study of Lesquerella fendleri, a desert mustard reliant on insect-pollination for reproduction (Roll et al., 1997). Populations with higher density showed a higher level of reproductive success when compared to those with smaller densities which may be subject to inbreeding depression. Smaller, isolated populations have a lower chance of reproductive success when compared to higher density populations of pollinator-dependent species as they are less likely to be attractive to pollinators. It was proposed that populations less

13 likely to attract pollinators may be a factor in reduced population density and size.

Pollinators tend to be generalists in this type of habitat as opposed to specialists.

Multiple different pollinator species have been noted in field observations of A. jaegerianus, with two species showing a higher occurrence in visitations; the leaf cutter bee and the metal leaf cutter bee (Karron, 1987; Rundel et al.,

2007). Seed dispersal is limited due to the pod size. Dispersal relies on granivores and water flow to disperse seeds beyond the population boundaries.

In higher plants, analysis of maternal and paternal genetic markers using nDNA is a better indicator of gene flow through the dispersal of pollen and seed, than is uniparental cpDNA and mtDNA. This study will utilize bi-parentally inherited microsatellite loci.

Isolation by Distance

All species exhibit some level of genetic differentiation across geographic ranges. The greater the distance, typically the lower the dispersal rates, therefore subpopulations that are geographically isolated have a lower probability of migration reducing the gene flow between distant populations. Populations in close proximity to one another will be more similar to each other than populations that are further apart (Avise, John C., 2004; Frankham, 1995). For neutral markers such as microsatellites, genetic structure among populations is correlated with physical distance. To test for isolation by distance, a Mantel test will be constructed consisting of a geographic distance matrix and a genetic distance matrix among sampling sites.

Methods for Estimating Genetic Diversity

Population genetics was reliant on phenotypic variation descriptions as an indicator of genetic variation before molecular techniques were introduced.

Estimating heterozygosity from phenotypic traits such as flower color, leaf size or height of the plant is possible when the traits are Mendelian but is practically impossible to determine in species in natural environments. Phenotypic variation as an indicator of genetic variation encompassed only a small portion of the molecular level of variation within a species and could not be universally applied to all organisms (Hamrick et al., 1979; Hartl & Clark, 2006; Walker, 2005).

The use of electrophoresis in determining protein variants for measuring genetic variation in species, lead to a change from phenotypic variation to molecular basis in estimating levels of variation. This analysis process showed that species, across the board, had higher than expected levels of genetic variation (Hamrick et al., 1979; Hartl & Clark, 2006; Walker, 2005). The use of electrophoresis in determining allozyme variations was a starting point but had its limitations. Allozymes are the expression of a gene but do not reveal hidden DNA sequence variation which led to an underestimation of actual levels of genetic polymorphism (Leberg, 1996; Wang & Szmidt, 2001). Continued advancements in molecular techniques, such as DNA sequencing and the Polymerase Chain

Reaction (PCR), lead to more direct quantifications of genetic variation on a molecular level (Mullis et al., 1986; Sanger et al., 1977). These advancements lead to the ability to determine genetic variation at the individual, population, and

15 species levels. Two of the more common methods used to estimate population structure in plants are Amplified Fragment Length Polymorphism and microsatellites, which are discussed below.

Amplified Fragment Length Polymorphism (AFLP)

Amplified Fragment Length Polymorphism is a commonly used tool for estimating genetic diversity. This process is ideal in determining polymorphism in populations without prior genomic information, as the process utilizes a combination of restriction enzymes, PCR amplification and electrophoresis (Kalia et al., 2011; Vos et al., 1995). The primers used require small amounts of DNA and can generate large amounts of marker fragments, typically between 50 to

100 fragments. Two different restriction enzymes are typically utilized: frequent cutter, which is four base pairs in length and generate the small fragments that are necessary for gel electrophoresis and the rare cutter, which is six base pairs in length and is utilized to reduce the quantity of fragments that will be amplified.

The restriction enzymes cut the DNA in specific sequences that are then amplified exponentially through PCR. The PCR process allows amplification of hundreds of DNA fragments at the same time using the whole genome (Gerber et al. 2000; Mueller and Wolfenbarger, 1999).

One weakness in using AFLP’s is the inability to determine if identical sized fragments are from the same genome section, i.e homologous segments, between samples. Also, this process is dominantly scored, such that the sample either has the sequence or not. Lastly, homozygosity and heterozygosity cannot

16 be determined. Amplified Fragment Length Polymorphism is an effective way to initially determine levels of genetic diversity within a species of interest but for more detailed analysis whole genome sequences or microsatellite analysis may offer more information (Savelkoul et al., 1999).

Microsatellites

Microsatellite analysis is a different avenue in determining population structure. The process of microsatellite analysis identifies short tandem base pair repeats, two to four base pairs that are repeated for a total length between 75-

300 base pairs from the genome. These repeated segments are “selectively neutral” in that they are not known to code for a protein or RNA structure and have a high mutation rate when compared to coding sections of the genome, between 10-5 to 10-3 mutations per generation. The mechanism for change in allele formation is attributed to unequal exchange during meiosis or strand slippage during replication, both of which can lead to an expansion or contraction of the DNA strand. In vitro experiments (Valdes et al., 1993) indicated that slippage occurrence tended to be higher for di-nucleotide repeat units when compared to longer repeat units, and repeats with higher G-C content showed lower occurrence. The length of the entire repeat region did not show an influence on the occurrence of slippage. Stepwise mutation-model (SMM) and

Two-Phase Model (TPM) work to explain different ideals on slippage. SMM explains slippage through the loss and gain of one unit during a slippage event leading to an expansion or contraction of the strand by one unit. This may restrict

17 some slippage events, but more than a one-unit change has been observed.

TPM allows for more than a one-unit change to a strand. The Proportional

Slippage (PS) Model looks at the effects the strand repeat region length has on the occurrence of slippage. It was found that there is a focal point at which expansion and contraction occur, microsatellites below a certain focal point tend toward expansion and those above tend toward contraction (Sainudiin et al.,

2004).

These highly variable repeated sequences allow researchers to estimate population structure, migration rates, and gene flow. Microsatellite analysis allows for a locus-specific focus as opposed to the whole genome with AFLP analysis. The information obtained through this analysis can aid in additionally determining fine scale phylogeny, population structure and demographic history

Two areas of concern are null alleles and stutter/shadow bands. Null alleles are non-amplifying alleles due to a mutation at the identified primer site. A change, such as a point mutation, would prohibit a primer from annealing to the primer target site (Ardren et al., 1999). Null alleles can lead to errors in determining heterozygosity and familial relationship in a population.

Stutter/shadow strands are typically a result in mispairing during the PCR process leading to a ladder of bands 1-2 base pairs apart. These stutter/shadow bands can lead to mis-scoring in alleles (O’Reilly and Wright, 1995). These potential errors can be reduced or mitigated through quality control, and

18 inspection for non-Mendelian ratios. In this study we will employ primer sets that were screened and optimized to reduce these errors (Dewoody et al., 2006).

The identification and characterization of microsatellite loci for A. jaegerianus is the primary focus of this thesis. For this thesis, microsatellite loci will be utilized in the characterization of genetic variation present within and among the five identified populations of A. jaegerianus. I hypothesize that there will be a high level of genetic variation with the narrow endemic A. jaegerianus, with small-partitioned populations. Walker and Metcalf (2008) found a statistically significant level of genetic variation and population structure through the use of

AFLPs. The genetic variation previously documented in A. jaegerianus was found to be similar to that exhibited by geographically widespread plant species. I hypothesize that the high level of genetic variation among the five identified populations can be attributed to the geographic segregation and lack of gene flow between the populations.

CHAPTER TWO

MATERIALS AND METHODS

Specimen Collection

Astragalus jaegerianus samples were collected between 2002-2004

(Metcalf and Hessing, pers comm.) These samples were collected from five geographically separated populations: Coolgardie Mesa, Lane Mountain,

Prospectors Wash, Brinkman Wash and Goldstone. Maps and logs of previously documented populations were acquired from the Department of the Army, Fort

Irwin National Training Center. Maps and logs included Global Positioning

System (GPS) coordinates from a 2001 survey of A. jaegerianus. Documented samples were sought out, along with possible new growth samples, within a 15 to

25-meter range. GPS location and field numbers were assigned for any new samples. Tissue samples were collected by Dr. Anthony Metcalf and Mark

Hessing, placed in plastic bags, and stored on ice. Once in the laboratory, the samples were stored at -20˚C.

Molecular Methods

A total of 176 tissue samples and previously extracted DNA samples in TE

Buffer were available for microsatellite analysis. From the initial 176 samples, two tissue samples were selected for genomic deoxyribonucleic acid isolation using the Qiagen’s Plant Dneasy Kit procedures: one sample from Brinkman Wash population and one from Coolgardie Mesa population. The plant tissue was

20 frozen with liquid nitrogen then ground up with a mortar and pestle for extraction following the Qiagen DNeasy Plant Mini Kit (Chatsworth, California) extraction procedure, then stored in AE buffer.

Samples weighed between 20 mg and 1.0 g and were comprised of stems and leaves. Samples were placed in a mortar then added to the Cryo-

Homogenizing System (Fisher Scientific) that had been prepared with liquid nitrogen, and subsequently pulverized using a Dremel tool. The liquid nitrogen is used to break down the cell wall and cell membranes. Pulverized samples were added to 1.5 ml microcentrifuge tubes followed by the addition of 400 µl Buffer

AP1 and 4 µl RNase A. Microcentrifuge tubes were vortexed then incubated for

10 minutes at 65˚C. During the 10-minute incubation period the microcentrifuge tubes were inverted 3 times, at 3-minute intervals, to further mix the solution and dislodge any tissue that may have become clumped in the bottom of the tube.

Once the incubation period was complete, 130 µl of Buffer P3 was added. The tubes were vortexed and placed on ice for 5 minutes followed by 5 minutes in the centrifuge (Eppendorf Centrifuge 5424) at 14,000 rpm. The lysate was pipetted into a QIAshredder spin column and placed in a 2 ml collection tube, then centrifuged for 2 minutes at 14,000 rpm. The flow through was transferred to a new 1.5 ml microcentrifuge tube and volume was estimated. Then 1.5 times the estimated volume of Buffer AW1 is added and pipette mixed. Measured amounts of Buffer AW1 ranged from 600 µl to 700 µl. An initial 650 µl of the flow through mixed with Buffer AW1 was transferred to the DNeasy Mini column which was

21 placed in a 2 ml collection tube and centrifuged for 1 minute at 8,000 rpm. The flow through was discarded and the remaining mixture was transferred to the column and again centrifuged for 1 minute at 8,000 rpm with flow through discarded when complete. The spin column was transferred to a new 1.5 ml centrifuge tube followed by the addition of 75-100 µl of Buffer AE, preheated to

65˚C. The solution was allowed to incubate at room temperature for 5 minutes then centrifuge for 1 minute at 8,000 rpm. Repeated the addition of preheated

Buffer AE, incubation, and centrifuge steps. The columns were then discarded and the flow through material placed in 4˚C for storage. Variations on the amount of Buffer AE utilized were based on the amount of “brown”, or biologically inactive, plant material used for the process. Increasing the amount of buffer was noted to help increase yield in some DNA extraction procedures.

Using the NanoDrop™ One (Thermo Scientific™) spectrophotometer,

DNA samples of 1.0 µl to 1.5 µl were analyzed for DNA concentration, 260/280 and 260/230. DNA concentrations yield ranged between 3ng/µl to 49ng/µl. The required concentration of 50ng/µl in a 100 ul solution required multiple extraction procedures. In the instance of multiple extractions, they were combined into one

1.5 ml centrifuge tube and placed in the Savant DNA centrifuge SpeedVac for 30 to 45 minutes on medium heat. The 260/280 readings indicate DNA and protein presence in the sample, as nucleic acids have absorbance in the 260 nm spectrum while proteins are around 280 nm. Ratios between 1.8 to 1.9 are considered to be pure DNA, greater than 1.9 is indicative of RNA presence. The

260/230 evaluations are used for contamination detection which has an absorbance in the 230 nm spectrum, ratios of 2.0-2.2 are considered common.

Initial DNA concentrations ranged from 1.0 ng/µl to 37.0 ng/µl.

Additional samples were acquired from previous DNA extraction procedures performed by Katherine Keenan and stored at -20˚C in 1x TE buffer

(pH 7.0). Samples ranged in DNA concentration and were diluted using 1x TE buffer (pH7.0) to reach the required 20 ng/µl concentration. All additional samples were analyzed for DNA concentration, 260/280 and 260/230 levels utilizing the NanoDrop™ One (Thermo Scientific™).

Microsatellite Acquisition

DNA extraction, following the protocol in the Eppendorf Phase Lock Gel

Manual (Mouse Tail Genomic DNA Isolation Protocol), of two samples was used for the isolation of microsatellite loci. An Illumina paired-end shotgun library was prepared and the reads were analyzed with PAL_FINDER_v0.02.03 (Castoe et al., 2012), which identified the reads that contained di-, tri-, tetra-, penta-, and hexa-nucleotide microsatellites (Nunziata et al., 2012). Forty-eight loci were eventually selected. From an additional 22 samples representing the 5 populations, loci were narrowed down to 26 that were found to be useful and polymorphic (Appendix B, Table 1).

The 26 identified microsatellites underwent preliminary screening to determine which seven primers were optimal for this study. Unlabeled primers for all 26 identified loci were tested on three A. jaegerianus samples following a set

PCR process. For each 23 µl PCR reaction, a mixture was prepared that contained 1 µl genomic DNA (21.0 ng/µl to 26.6 ng/µl), 1 µl of each primer sequence, 0.5 µl of Taq polymerase, 1 µl of 10 µm dDNTP solution, 2.5 µl of 10x

Taq polymerase buffer, 1.5 µl of MgCl2, and 14.5 µl dH2O. The PCR program was set using the BioRad C1000 Thermal Cycler. Initial denaturing temperature was set to 95˚C for a 3-minute cycle, followed by 35 cycles starting with 95 ˚C for

30 seconds for denaturation, then the annealing process set at 57 ˚C for 20 seconds followed by 60 ˚C for 30 seconds, At the end of the 35 cycles, there was a final extension of 60 ˚C for 2 minutes then placed on hold at 4 ˚C until removed and placed in the storage at 4 ˚C. All samples were visualized using a 2%

METAPHOR© gel which were then analyzed to determine the optimal seven loci for this study.

Polymerase Chain Reaction Process

Polymerase Chain Reaction procedure was utilized to amplify the identified microsatellite loci. PCR procedures were performed by Savannah River

Ecology Lab following their established protocol. Of the identified primer pairs, one primer sequenced was modified with the addition of a CAG tag to the 5’ end of the sequence which enables the addition of a third primer to the PCR process that will add the fluorescent label. The second primer in the pair was modified with a GTTT sequence without the CAG tag.

Samples were provided to Savannah River Ecology lab in 96 well plates containing between 30 to 50 µl suspension with DNA concentration of 20 ng/µl.

PCR amplifications were performed in a 12.5 L volume with the following conditions,10 mM Tris pH 8.4, 50 mM KCl, 25.0 g/ml BSA, 0.4 M unlabeled primer, 0.04M tag labeled primer, 0.36M universal dye-labeled primer, 3.0 mM

MgCl2, 0.8 mM dNTPs, 0.5U AmpliTaq Gold® Polymerase (Applied Biosystems,

Foster City, CA), and 20 ng DNA template. All PCR amplifications were performed using touch down thermal cycling program (Don et al. 1991) on an

Applied Biosystems GeneAmp 9700. After amplification, PCR products were run on an ABI-3130xl (Applied Biosystems, Foster City, CA) sequencer and sized with Naurox size standard prepared as described in DeWoody et al. (2006), with the exception that the unlabeled primers started with GTTT. Results were analyzed using GeneMapper version 3.7 (Applied Biosystems, Foster City, CA).

All DNA samples were run across the seven identified microsatellite loci a single time and were evaluated for heterozygosity, homozygosity, and null alleles.

Samples that exhibited homozygosity were processed a second time, and null alleles were processed an additional two runs. Of the 176 samples provided, 14 failed to amplify across all seven loci and were removed from the study. Data provided by SREL was then analyzed using various statistical programs to determine population structure, genetic variation among and within populations and heterozygosity of A. jaegerianus.

Population Genetics

Genetic Structure: Fixation Index (FST and RST)

This study was designed to characterize the genetic variation and population structure among and within the five putative populations of A. jaegerianus found in the Barstow area of Southern California. Microsatellites of the 7 loci described above (Appendix B, Table 2) from 176 individuals across 5 populations were amplified and scored using GeneMapper Version 3.7. Of the

176 individuals, 140 successfully amplified across at least 4 of the 7 identified microsatellite loci across all subpopulations. The 140 individuals were utilized in calculating the genetic diversity indices for each of the subpopulations and total geographic range. Calculations included the number of alleles per locus (NA), observed and expected heterozygosity (Ho, He), allelic richness (AR), and the number of private alleles (Peakall and Smouse, 2006). This analysis was utilized in determining the population structure, historical gene flow between adjacent populations, and more complete genetic characterization of A. jaegerianus.

Genetic differentiation was examined using both F-statistics, FST (Wright,

1931) and RST (Slatkin, 1995). The FST values were used to compare the level of genetic variation within a population with values ranging from 0 to 1. Within FST, the closer the value is to 1, the less genetic variation is shared among populations. However, FST is sensitive to mutation rates, and with the assumption of a high mutation rate for microsatellites, I also included RST which is less sensitive to mutation rates. RST values are more accurate under the condition

26 that new mutations in a population are more similar to the new allele’s previous number of repeats than a novel random mutation (Balloux and Lugon-Moulin,

2002). This step-wise mutation process is thought to underlie microsatellite differentiation, hence I included both types of analysis (Hardy et al., 2003).

Significance testing of FST and RST values and variance components were based on 999 permutations with suppression of within individual analysis and evaluated at a 0.05 probability level.

GENALEX version 6.503 (Peakall and Smouse, 2006) was utilized to examine pairwise and overall FST and RST values. Additionally, a Mantel test was performed in GENALEX version 6.503 to compare pairwise FST obtained and geographic distance between the five subpopulations. Centroid calculations were used for the pairwise geographic distances (km). Inferred populations (K) and patterns of genetic variability were calculated using STRUCTURE (version 2.3.4).

The K value was used to infer the number of subpopulations (clusters) that make up a population (Falush et al., 2003, 2007; Nerkowski, 2015)..

Analysis of Molecular Variance (AMOVA), is a statistical method that is utilized to determine population differentiation within and among populations with the use of FST or RST statistics (Meirmans, 2012). GENALEX version 6.503 was utilized for the AMOVA analysis. The AMOVA was analyzed with 5 populations

(with the adjacent Coolgardie Mesa and Lane Mountain as two distinct populations), and with 4 populations (Coolgardie Mesa and Lane Mountain considered as one population).

Principal Coordinate Analysis (PCoA) was performed utilizing GENALEX version 6.503 as a way to visualize dissimilarity between the populations. PCoA was utilized on the four population and five population data sets. A Mantel test was performed utilizing GENALEX version 6.503 to test for isolation by distance among the populations that were sampled. The Mantel Test examined the correlation between pairwise FST values from microsatellite data, and the geographic distance (km) between the populations. Geographic distances were determined using centroid calculations in Arc-GIS. Mantel values range from +1 to -1, positive numbers indicate a positive correlation between genetic and geographic distance displaying isolation by distance, the negative number indicates no isolation by distance was found.

Genetic variability and the number of inferred subpopulations (K) were identified utilizing STRUCTURE (version 2.3.4) (Falush et al., 2003, 2007). K was used to indicate the number of populations based on genetic variation and similarities in the haplotypes of the microsatellite data. The parameters, using the admixture model, were set with 15 independent runs of K = 1-10, K representing the number of putative clusters. The initial burn-in period was set at 10,000 iterations with a run length of 200,000 Markov Chain Monte Carlo iterations

(Evanno et al., 2005; Porras-Hurtado et al., 2013). CLUMPAK was utilized to determine the Best K values represented by the data using the second order rate of change, K and LnP(D) using the Evanno, Regunaut, & Goudet (2005) method.

CHAPTER THREE

RESULTS

This thesis examined the levels of genetic differentiation within and among the Astragalus jaegerianus population and subpopulations found in the Barstow area of Southern California. Populations were subdivided into four and five populations based on geographical separation (Walker and Metcalf, 2008) and population delineation reported in the Charis (2002) report and the U.S. Fish and

Wildlife Service Recovery Plan (2001). Walker and Metcalf AFLP research subdivided the populations into five subpopulations: Coolgardie Mesa (CM 6-10),

Lane Mountain (CM 1-5), Prospector Wash (EP), Brinkman Wash (BW), and

Goldstone (NG) populations (Appendix A, Figures 1-6). The Charis Report

(Kearns, 2002) combined the Coolgardie Mesa and Lane Mountain populations into one subpopulation. Based on the two different delineations of population number, four versus five, genetic analyses were performed on the populations as four (Appendix B, Table 8) and five (Appendix B, Table 9) distinct subpopulations. Genetic Indices calculated include the number of alleles (N), number of alleles found at each locus (NA), the observed (HO) and expected (HE) heterozygosity and the Fixation Index (F) for each of the microsatellite loci. For

ASJA 02, 14 alleles were identified with a base pair range between 204-259,

ASJA 08 exhibited a range of 299-232 bp with 13 alleles identified; ASJA 17 exhibited a range of 273-345 bp with 14 alleles identified; ASJA 19 exhibited a

29 range of 380-345 bp with 18 alleles identified; ASJA 25 exhibited a range of 177-

229 bp with 13 alleles exhibited; ASJA 38 exhibited a range of 182-218 bp with 5 alleles identified; and ASJA 45 exhibited a range of 258-298 with 11 alleles identified.

Mean heterozygosity for the four populations ranged from 0.443

(Coolgardie Mesa population) and 0.550 (Goldstone population), and for the five- population analysis the range was from 0.419 (Lane Mountain population) and

0.550 (Goldstone population). All of the populations, four and five population calculations, showed significant deviation from the expected Hardy-Weinberg proportions to varying degrees across all seven loci. (Appendix B, Tables 10-11).

When looking at the population as 4 subpopulations the sample size varies from 20 to 55 individuals. Allelic richness across all 7 loci ranges from

4.218 and 5.095 and there are 6 private alleles noted for Coolgardie Mesa,

Prospector Wash, and Brinkman Wash and 4 private alleles for Goldstone

(Appendix B, Table 12). When analyzing the population as 5 subpopulations the sample size varies from 20-36 individuals. Allelic richness across the 7 loci ranges from 4.193 to 5.095 and there are 6 private alleles identified for

Prospector Wash and Brinkman Wash, 4 private alleles for Goldstone, 3 private alleles for Coolgardie Mesa and two private alleles for Lane Mountain (Appendix

B, Table 13).

Population Genetic Structure

The geographic structure among the five subpopulations sampled was evaluated using STRUCTURE runs K = 1-10 in STRUCTURE HARVESTER.

Based on the 140 samples from all subpopulations the most probable K was identified as K=4. Principal Coordinates Analyses concurred with the

STRUCTURE results indicated 4 populations, with Coolgardie Mesa and Lane

Mountain as one population and Brinkman Wash as the most distinct of the populations (Appendix A, Figures 7-8). Three variations of the Mantel Test were performed, genetic distance vs geographic distance, FST vs geographic distance, and RST vs geographic distance of subpopulations. The Mantel tests performed showed no statistically significant correlation on all three variations.

Estimated gene flow between the Lane Mountain Milkvetch subpopulations was examined utilizing migration rate (Nm). Table 14 in Appendix

B illustrates the RST values and their significance levels of the overall AMOVA analyzing the population as four and five distinct subpopulations of Lane

Mountain Milkvetch along with the estimated migration numbers. Migration rates ranged from 1.135 to 1.581 showing a constant level of migration between subpopulations, suggesting at least one migration event per generation among the subpopulations (i.e. Nm >1). However, since Nm estimations are directly dependent on FST values we view migration rates among populations as relative values.

The greatest amount of genetic differentiation was observed between

Goldstone and Brinkman Wash (FST=0.100, p=0.001, RST=0.328, p=0.001) both highly significant (Appendix B, Tables 15-18). The least amount of genetic differentiation occurred between CM and LM (RST=0.010, p=0.187, FST=0.001, p=0.006) (Appendix B, Tables 17-18). When analyzing FST values versus RST values it is important to note that high mutation rates can lead to a reduced FST value as the calculation does not consider the Step-wise Mutation Model (Balloux and Lugon-Moulin, 2002). Overall FST among populations accounted for 5% of the variance observed and 95% within populations for both the four and five subpopulations (Appendix B, Tables 17-18). Due to the low levels of variation in the subdivision of Coolgardie Mesa and Lane Mountain, when combined into one subpopulation the highest values remained between Brinkman Wash and

Goldstone. Overall RST for the among populations accounted for 16% of the variance observed and 84% within populations the four subpopulations

(Appendix B, Table 15 a & b) and the among populations accounted for 15% variance and within populations accounted for 85% variance for 5 subpopulations. A Mantel test comparing the cumulative pairwise FST and RST values between each of the five and four subpopulations and geographic locations for all sampling locations showed no statistically significant correlation.

CHAPTER FOUR

DISCUSSION

Genetic Variation within Astragalus jaegerianus

Previous studies on Astragalus jaegerianus revealed no DNA sequence variation in the chloroplast and nuclear genome (Walker & Metcalf, 2008). This lack of genomic sequence variation was attributed to the relatively small population size, which can vary depending on the wet versus dry season cycles, along with the geographic restriction of each subpopulation (Walker, 2005;

Walker & Metcalf, 2008). Walker’s (2005) AFLP data showed not only substantial variation, but it also supported the hypothesis that genetic structure was correlated with geographic isolation by distance (Walker & Metcalf, 2008). The current research project focused on loci-specific microsatellites to further investigate the genetic variation and effects of limited population size on

Astragalus jaegerianus. From samples provided to the Savannah River Ecology

Lab, 24 microsatellite loci were identified of which seven were chosen based on successful amplification and utilized across 176 samples. Amplification of loci was not successful for each sample; samples that were successful across a minimum of 4 loci were chosen for statistical analysis. Of the 140 samples selected for analysis, 45 individuals amplified across all 7 loci: 9 in Coolgardie

Mesa, 7 in Lane Mountain, 16 in Prospector Wash, 6 in Brinkman Wash and 7 in

Goldstone. The number of alleles per locus ranged from 5 (ASJA38) to 18

(ASJA19).

The AFLP research showed substantial genetic variation for Astragalus jaegerianus, genetic diversity ranged from 0.1905 to 0.2505 across the four identified populations. Coolgardie Mesa and Lane Mountain were best represented as a single population unit. The populations’ genetic diversity showed higher than expected values for a species across a limited geographic range (Hamrick et al., 1991; Karron, 1991; Walker & Metcalf, 2008).

Microsatellite analysis indicated a moderate level of genetic diversity, the pairwise population FST values range from 0.001 to 0.100 with the higher levels of genetic diversity found between Brinkman Wash and Coolgardie Mesa (FST =

0.080, p=0.001) and Brinkman Wash and Goldstone (FST = 0.100, p=0.001).

Brinkman Wash and Goldstone are adjacent subpopulations and exhibit the highest level of genetic diversity. Brinkman Wash and Goldstone are also the closest in proximity at approximately 6,281 meters apart, based on population centroid location.

The global RST values (Appendix B, Table 14) show statistically significant levels of molecular variance for the subpopulations whether they are identified as five subpopulations (RST = 0.137, p = 0.001, Nm = 1.581), or as four subpopulations (Coolgardie Mesa and Lane Mountain as one contiguous population), (RST =0.151, p = 0.001, Nm = 1.581), with levels of migration between populations greater than 1. The AMOVA Pairwise RST values for four

34 subpopulations ranged from 0.038 (p=002) to 0.328 (p=0.001), the higher ranges centered on the Brinkman Wash (Appendix B, Tables 15-16). Based on RST values, 86% of the molecular variance is among populations and 14% is within populations, which supports the hypothesis that the Lane Mountain milkvetch exhibits high levels of population structure. The microsatellite analysis supports the findings of Walker and Metcalf (2008) that the Lane Mountain milkvetch may have recently or are currently undergoing population contraction reducing the overall population size but still maintaining genetic diversity (Schmidt and

Jensen, 2000; Walker & Metcalf, 2008). Other factors such as the historic range of the species, bottlenecks, founder effects and microevolutionary changes could be contributing factors in the pattern and extent of genetic variation found for

Astragalus jaegerianus, however, historical factors prior to 1938 when Astragalus jaegerianus was first identified are not available. (Avise, John C., 2004; Barneby,

1964; Munz, 1941; Walker & Metcalf, 2008).

Population Structure and Geographic Fragmentation

Astragalus bibullatus is a federally endangered legume found growing in limestone in the Rutherford County, Tennessee (Trostel, 2017). Initial allozyme studies showed a low level of polymorphism but AFLP analysis indicated higher levels of genetic diversity and the presence of two genetically distinct groups

(Baskauf and Burke, 2009). The diversity levels were still considered lower than other plants found in the cedar glen area where A. bibullatus is found.

Microsatellite analysis using FST indicated that the populations were genetically

35 similar. The PCoA results showed a nearly complete overlap of individuals from all populations, suggesting limited differentiation among populations. While AFLP indicated genetic diversity when considering the whole genome, when focusing on microsatellite loci, the analysis indicated no genetic differentiation. A comparison of AFLP and microsatellite markers in characterizing hops plants

(Humulus lupulus L.) was used to compare the genetic diversity through a comparison of dendrograms (Jakse et al., 2001). The microsatellite markers successfully grouped genetically related individuals while the AFLP indicated clustering of closely related individuals and showed geographically distinct populations. This distinction between AFLP and microsatellite analysis is similar to what was found with the analysis of Astragalus jaegerianus. AFLP analysis showed a high level of genetic diversity and a significant relationship of isolation by distance (Walker & Metcalf, 2008). Microsatellite analysis agrees with Walker and Metcalf’s (2008) finding that there is a high level of genetic diversity with strong population structure indicating four distinct populations but did not indicate isolation by distance as a sole contributing factor. When considering whether

Coolgardie Mesa and Lane Mountain should be seen as one contiguous population or two distinct populations, the PCoA results (Appendix A, Figures 7-

8) support one contiguous population. The PCoA results consistently supported

Brinkman Wash is the most distinct population.

Population Genetics

Moderate levels of gene flow between Astragalus jaegerianus is supported by the global and individual pairwise FST coefficients based on four populations: global FST = 0.049 and pairwise FST ranging from 0.037 to 0.100, p=0.001

(Appendix B, Tables 17-18). Gene flow is mediated by pollination and seed dispersal, and based on the large size of Astragalus jaegerianus seed pod the dispersal distance is believed to be influenced by high winds and water flow, leaving pollen dispersal the more important factor in gene flow between subpopulations (Avise, 2004; Segarra-Moragues et al., 2005). Richard et al.

(1999), in their research on Silene alba, indicated that the pollen fertilization outside the focal patch decreased as distance increased. Based on Walker and

Metcalf’s (2008) findings, population density had a more significant impact on genetic diversity then population size with Astragalus jaegerianus populations, suggesting that local pollination maintained higher than expected levels of variation for a small population.

Plant density and proximity can play an important role in attracting pollinators. Astragalus jaegerianus can be found in small groups separated by several hundred yards which limits the density and possible attraction of pollinators (Walker and Metcalf, 2008). Roll et al (1997) found that plant density and proximity were important in attracting pollinators. Richards, Church,

McCauley (1999) found that the migration rates of Silene alba decreased as the distance between patches increased. Migration rates dropped from 47% overall

37 for patches at 20 meters away from the focal patch, to 6% overall for patches 80 meters away. The pollinators found in the area of Astragalus jaegerianus are found to be generalists, as opposed to specialists which can limit the gene flow between subpopulations. Estimated migration coefficients indicate homogenizing levels of migration between populations: Nm=1.581 based on 5 populations and

Nm=1.410 based on 4 populations (Appendix B, Table 14).

The AFLP and microsatellite analysis supports the hypothesis that

Astragalus jaegerianus utilizes both outcrossing and selfing as a breeding strategy. The low level of observed heterozygosity (Appendix B, Tables 8-9) indicates that inbreeding may be occurring. Walker’s (2005) global FST = 0.133, based on AFLP analysis, supports the hypothesis that Astragalus jaegerianus are also facultative outcrossers and mediated by pollinators, while microsatellite analysis indicates that selfing and higher levels of pollen flow may be contributing factors to lower levels of population structure than shown by AFLP. The AFLP and microsatellite analysis support the hypothesis that outcrossing and selfing occur within the subpopulation of Astragalus jaegerianus, and population density may play a role (Walker and Metcalf, 2008). Plants with a higher density of flowers could receive more pollinators which can increase the levels of pollination occurring in that population; the lower the flower density the less likely pollination will occur (Richards et al., 1999; Roll et al., 1997). Walker and Metcalf (2008) found a lower within-population gene diversity in the low-density populations, a

38 finding which supports the hypothesis that self-fertilization is also occurring and predominantly in low-density populations.

Conclusion with Conservation Implications

The Astragalus jaegerianus is a narrow endemic species found in an area of limited geographic range. Based on the limited population size, limited geographic range and geographic isolation, the species should exhibit very low genetic variation; and yet through both AFLP and microsatellite analysis we show it has significant levels of genetic diversity and significant population structure.

Microsatellite analysis supports Walker and Metcalf’s (2008) findings of population structure and genetic variation comparable to geographically widespread plant species. In comparing the STRUCTURE analysis and the outcome of the PCoA, the data indicates there are four distinct populations, and that Coolgardie Mesa and Lane Mountain appear to be more similar and can be viewed as a single population, with Brinkman Wash as the most distinct of the populations.

On October 6, 1998, Astragalus jaegerianus was classified as a Critically

Federally Endangered Species and a critical habitat rulemaking decision was made in 2005 (“Endangered and Threatened Wildlife and Plants; 12-Month

Finding on a Petition To Reclassify Astragalus Jaegerianus as a Threatened

Species,” 2014). In 2008, a 5-year review of the Lane Mountain milkvetch indicated that the species should be reclassified from Critically Endangered to

Threatened. This recommendation was based on increase in the census count and proposed land management plans. U.S. Fish and Wildlife raised concerns over the Lane Mountain milkvetch habitat based on a change in land usage/ownership. A portion of the habitat is located on National Training Center land at Fort Irwin and is subject to maneuvers performed in the surrounding area.

In addition, off-road high occupancy vehicles and mining operations located on the Bureau of Land Management property, in addition to the effects of climate change, are some of the concerns in the persistence of the Lane Mountain milkvetch. These disturbances can lead to further geographic fragmentation, changes in pollinator behaviors, and increased competition from non-native species leading to further genetic isolation. Based on the above information, in

April 2014 U. S. Fish and Wildlife Service determined that reclassification of the

Lane Mountain milkvetch is not warranted, and it should maintain its critically federally endangered status.

Through AFLP and microsatellite analysis we show that there is population structure which indicates there are four distinct populations,

Coolgardie Mesa/Lane Mountain, Prospector Wash, Brinkman Wash and

Goldstone. Walker (2005) also found that the DNA sequence data shows the effective population size is far less than the 5,000 individuals accounted for in the

2001 census. With the reduced population size and limited geographic range, the focus should be on conserving as many individuals as possible to maintain the high level of genetic diversity that is currently found in the population.

Microsatellite analysis focused on seven specific loci agreed with the

AFLP analysis that Astragalus jaegerianus has significant population structure and has maintained four distinct populations, with Brinkman Wash showing the highest level of differentiation. Coolgardie Mesa and Lane Mountain were the most similar and appeared to have the lowest overall density. With evidence that selfing is occurring in the populations with lower density (J. L. Hamrick et al.,

1979; Karron, 1991; Travis et al., 1996b) a focus should be on the pollinators and their ability to increase facultative outcrossing throughout each of the populations. Areas of concern with the pollinator population is the increase in off road vehicles and rock hounding in the areas where Astragalus jaegerianus are found. An increase in physical disturbance and the dust effects inhibit the pollinators’ ability to access the flowers and therefore inhibit the pollination efforts

(Karron, 1987; Wijayratne et al., 2009).

An important aspect in the conservation of the remaining populations of

Astragalus jaegerianus is at the very least maintaining population size. There needs to be a better understanding of the ecology and how the pollinators affect gene flow. A reduction in the population size can reduce the genetic diversity which leaves the remaining population more susceptible to environmental influences such as climate change, disease, and changes to habitat. Changes in population size also make the plants more susceptible to genetic drift and increased inbreeding depression (Despres et al., 2002; Gaudeul et al., 2004).

APPENDIX A

FIGURES

Figure 1. Ranch map of Astragalus jaegerianus with the five populations. The map of California indicates the approximate location of the study sites.

Figure 2: Map location of the Coolgardie Mesa populations and approximate sample locations. Sample points are noted on the map as CM sites, individual plant identification numbers are listed on Table 4

Figure 3: Map location of the Lane Mountain populations and approximate sample locations. Samples within the Lane Mountain were combined as one sampling location with the Coolgardie Mesa and labeled CM (Coolgardie Mesa). Individual plant identification numbers are listed on Table 3.

Figure 4: Map location of the Brinkman Wash populations and approximate sample locations. Sample points are noted on the map as BW sites, individual plant identification numbers are listed on Table 5.

Figure 5: Map location of the Prospectors Wash populations and approximate sample locations. Sample points are noted on the map as EP sites, individual plant identification numbers are listed on Table 7.

Figure 6: Map location of the Goldstone populations and approximate sample locations. Sample points are noted on the map as NG sites, individual plant identification numbers are listed on Table 6.

Principal Coordinates (PCoA)

NG EP

Coord.2 BW CM

Coord. 1

Figure 7. Principal Coordinate Analysis (PCoA) for the populations identifies K=4. Coolgardie Mesa (CM) and Lane Mountain (LM) represent Region 1, Goldstone (NG) represents Region 2, Prospectors Wash (EP) represents Region 3 and Brinkman Wash (BW) represents Region 4.

Principal Coordinates (PCoA)

BW Coord.2

NG EP

Coord. 1

Figure 8. Principal Coordinate Analysis (PCoA) for the populations identifies K=4. Coolgardie Mesa represent Region 1, Goldstone (NG) represents Region 2, Prospectors Wash (EP) represents Region 3 and Brinkman Wash (BW) represents Region 4.

APPENDIX B

TABLES

Table 2. Identified optimal primers for population structure analysis for Astragalus jaegerianus. Unlabeled primers for the initial 26 primer loci were tested for amplification to determine the optimal loci for this study.

Repeat Locus Forward Primer Reverse Primer motif 2* ATCT TTTGTAGGTTGATAACCAATGTTCG TGATTTAGATAACCAATGTTTCGCC 8 AAAT CATGAACATAGCATTAAAGACCACC TTGACTATTGTTGATAATGATGGAAGC 17* AAAT TGCAAACAAGAGAAGAATCAACC CCAAACATAGCCTAACAATTTCC 19 ATCT TCCAATCAATCAATCTCTTCCG TGAAGCCACAATTGGTACTCTCC 25* ATCT TTGTACCATATGTTAGACAGACATTGC TTCCACCAATGTGAATGTAATGC 38* AAAC CATACGGTTGAGGGTTGTCG CAGGATGTGAATGGGAATGC 45 ATCT TTTCCAAGGGAAAGGAAAGG AAAGTGAGGAAACAAATGAGTGG

Table 3. Samples for Astragalus jaegerianus. The table provides a visual representation of the samples acquired from the Lane Mountain population including the sample points and plant identification numbers.

CM-1 CM-2 CM-3 CM-4 CM-5

703 770 706 640 753

705 771 707 749 755

775 709 756 776 713 757 777 719 758 778 716 759 779 718 760 780 722 762 723 764 724 765 725 767

Plant Identification Numbers Identification Plant 729 768 736 737 738

Table 4. Samples for Astragalus jaegerianus. The table provides a visual representation of the samples acquired from the Coolgardie Mesa population including the sample points and plant identification numbers.

Sample Points CM-6 CM-7 CM-8 CM-9 CM-10

871 831 601 812 834

872 833 801 813 835 873 836 802 814 851 874 804 815 853 875 825 841 876 877 879

lant Identification Number Identification lant 880 P 882 884

Table 5. Samples for Astragalus jaegerianus. The table provides a visual representation of the samples acquire from the Brinkman Wash population including the sample points and plant identification numbers.

Sample Points BW-1 BW-2 BW-3 BW-4 BW-5 BW-6 184 182 104 101 211 109 186 191 106 212 116 192 108 118 193 194 195 199 201 202

BW-7 BW-8 BW-9 BW-10 identification Number identification - 162 132 251 230

164 151 252 233 Plant 153 253 234

Table 6. Samples for Astragalus jaegerianus. The table provides a visual representation of the samples acquired from the Goldstone population including the sample points and plant identification numbers.

Sample Points NG-2 NG-3 NG-5 NG-7 NG-8 NG-9 NG-10

430 303 374 361 308 321 340

432 362 309 325 342 433 364 311 326 343

Number 368 327 344

370 328 345 Plant Identification Identification Plant 329

Table 7. Samples for Astragalus jaegerianus. The table provides a visual representation of the samples acquired from the Prospector Wash population including the sample points and plant identification numbers

Sample Points EP-1 EP-2 EP-3 EP-4 EP-6 EP-5 EP-9 EP-10

496 515 517 522 571 561 402 621 497 518 523 572 562 405 622 499 521 573 563 406 623 501 562 576 564 407 624 502 855 577 408 503 857 578 410 505 859 506 860

Plant Identification Number Identification Plant 507 513 519

Table 8. Genetic Diversity for Coolgardie Mesa, Lane Mountain, Prospectors Wash, Brinkman Wash, and Goldstone populations of Astragalus jaegerianus.

Pop Asja 02 Asja 08 Asja 17 Asja 19 Asja 25 Asja 38 Asja 45 N 21 15 22 23 23 10 20

Na 8 5 6 12 8 4 8

Coolgardie Ne 4.455 4.091 2.898 5.568 4.580 2.597 5.161 Mesa Ho 0.286 0.533 0.227 0.739 0.696 0.100 0.750

HE 0.776 0.756 0.655 0.820 0.782 0.615 0.806 F 0.632 0.294 0.653 0.099 0.110 0.837 0.070

N 30 21 31 28 30 12 27

Na 8 6 8 11 8 3 10 Lane N 4.306 3.445 1.519 8.296 4.478 2.667 6.877 Mountain e Ho 0.233 0.381 0.129 0.821 0.700 0.000 0.667

HE 0.768 0.710 0.342 0.879 0.777 0.625 0.855 F 0.696 0.463 0.623 0.066 0.099 1.000 0.220

N 33 27 34 34 36 27 32

Na 11 9 9 12 11 4 9 Prospectors N 7.723 3.930 2.995 8.377 4.012 2.325 6.302 Wash e Ho 0.394 0.556 0.412 0.882 0.861 0.148 0.594

HE 0.871 0.746 0.666 0.881 0.751 0.570 0.841 F 0.547 0.255 0.382 -0.002 -0.147 0.740 0.294

N 16 15 29 16 29 26 25

Na 10 7 9 10 9 3 9 Brinkman N 7.420 3.814 2.092 6.169 4.163 2.048 3.823 Wash e Ho 0.250 0.400 0.310 0.875 0.793 0.269 0.640

HE 0.865 0.738 0.522 0.838 0.760 0.512 0.738 F 0.711 0.458 0.405 -0.044 -0.044 0.474 0.133

N 19 18 20 20 20 9 19

Na 8 7 8 11 10 4 8

Goldstone Ne 5.232 4.765 4.790 6.667 4.678 2.571 5.309

Ho 0.158 0.667 0.450 0.750 0.750 0.111 0.842

HE 0.809 0.790 0.791 0.850 0.786 0.611 0.812 F 0.805 0.156 0.431 0.118 0.046 0.818 -0.038

Genetic Indices calculated include the number of alleles (N), number of alleles found at each locus (NA), the observed (HO) and expected (HE) heterozygosity and the Fixation Index (F) for each of the microsatellite loci.

Table 9. Genetic Diversity for the Coolgardie Mesa/Lane Mountain, Prospector Wash, Brinkman Wash and Goldstone populations of Astragalus jaegerianus.

Asja Asja Asja Asja Asja Asja Asja Pop 02 08 17 19 25 38 45 Coolgardie Mesa N 51 36 53 51 53 22 47 Na 11 7 8 14 9 4 10 Ne 4.889 3.823 1.975 7.954 4.647 2.696 6.302 HO 0.255 0.444 0.170 0.784 0.698 0.045 0.702 HE 0.795 0.738 0.494 0.874 0.785 0.629 0.841 F 0.680 0.398 0.656 0.103 0.110 0.928 0.165 Prospectors Wash N 33 27 34 34 36 27 32 Na 11.000 9.000 9.000 12.000 11.000 4.000 9.000 Ne 7.723 3.930 2.995 8.377 4.012 2.325 6.302 HO 0.394 0.556 0.412 0.882 0.861 0.148 0.594 HE 0.871 0.746 0.666 0.881 0.751 0.570 0.841 F 0.547 0.255 0.382 -0.002 -0.147 0.740 0.294 Brinkman Wash N 16 15 29 16 29 26 25 Na 10.000 7.000 9.000 10.000 9.000 3.000 9.000 Ne 7.420 3.814 2.092 6.169 4.163 2.048 3.823 HO 0.250 0.400 0.310 0.875 0.793 0.269 0.640 HE 0.865 0.738 0.522 0.838 0.760 0.512 0.738 F 0.711 0.458 0.405 -0.044 -0.044 0.474 0.133 Goldstone N 19 18 20 20 20 9 19 Na 8.000 7.000 8.000 11.000 10.000 4.000 8.000 Ne 5.232 4.765 4.790 6.667 4.678 2.571 5.309 HO 0.158 0.667 0.450 0.750 0.750 0.111 0.842 HE 0.809 0.790 0.791 0.850 0.786 0.611 0.812 F 0.805 0.156 0.431 0.118 0.046 0.818 -0.038

Table 10. Astragalus jaegerianus Populations Deviating from Expected Hardy- Weinberg based on four populations.

Pop Locus p-Value Coolgardie Mesa Asja02 0.000 *** Asja08 0.025 * Asja17 0.000 *** Asja19 0.004 ** Asja38 0.000 *** Asja45 0.000 *** Prospector Wash Asja02 0.000 *** Asja08 0.000 *** Asja17 0.000 *** Asja38 0.000 *** Brinkman Wash Asja02 0.000 *** Asja08 0.000 *** Asja17 0.000 *** Asja25 0.002 ** Asja38 0.010 * Asja45 0.002 ** Goldstone Asja02 0.000 *** Asja17 0.000 *** Asja19 0.008 ** Asja38 0.004 ** * P<0.05, ** P<0.01, *** P<0.001

Statistics based on Chi-Square Tests for Hardy- Weinberg Equilibrium. Only loci that significantly deviate from the expected Hardy-Weinberg Equilibrium are displayed.

Table 11. Astragalus jaegerianus Populations Deviating from Expected Hardy- Weinberg based on five populations.

Population Locus p-VALUE Coolgardie Mesa ASJA 02 0.000 *** ASJA 17 0.000 *** ASJA 19 0.035 * ASJA 38 0.003 ** Lane Mountain ASJA 02 0.000 *** ASJA 17 0.000 *** ASJA 19 0.031 * ASJA 38 0.000 *** ASJA 45 0.012 * Prospectors Wash ASJA 02 0.000 *** ASJA 08 0.000 *** ASJA 17 0.000 *** ASJA 38 0.000 *** Brinkman Wash ASJA 02 0.000 *** ASJA 08 0.000 *** ASJA 17 0.000 *** ASJA 25 0.002 ** ASJA 38 0.010 * ASJA 45 0.002 ** Goldstone ASJA 02 0.000 *** ASJA 17 0.000 *** ASJA 19 0.008 ** ASJA 38 0.004 **

* P<0.05, ** P<0.01, *** P<0.001

Statistics based on Chi-Square Tests for Hardy- Weinberg Equilibrium. Only loci that significantly deviate from the expected Hardy-Weinberg Equilibrium are displayed.

Table 12. Astragalus jaegerianus Genetic Diversity Summaries for each of the four populations. The number of samples (n), Allelic Richness across all seven loci (AR), observed heterozygosity (HO), expected heterozygosity (HE), and private alleles (PA) are identified.

Population n AR HO HE PA Coolgardie Mesa 55 4.612 0.443 0.737 6 Prospector Wash 36 5.095 0.55 0.761 6 Brinkman Wash 29 4.218 0.505 0.710 6 Goldstone 20 4.859 0.533 0.778 4

Table 13. Astragalus jaegerianus Genetic Diversity Summaries for each of the five populations. The number of samples (n), Allelic Richness across all seven loci (AR), observed heterozygosity (HO), expected heterozygosity (HE), and private alleles (PA) are identified.

Population n AR HO HE PA Coolgardie Mesa 23 4.193 0.476 0.744 3 Lane Mountain 32 4.513 0.419 0.708 2 Prospector Wash 36 5.095 0.550 0.761 6 Brinkman Wash 29 4.218 0.505 0.710 6 Goldstone 20 4.859 0.533 0.778 4

Table 14. RST and NM values for Analysis of Molecular Variance Comparisons

RST p- Comparison NM Values Values

Coolgardie Mesa, Lane Mountain, Prospectors 0.137 0.001 1.581 Wash, Brinkman Wash, Goldstone

Coolgardie Mesa (includes Lane Mountain), Prospector 0.151 0.001 1.410 Wash, Brinkman Wash, Goldstone

RST values and their significance levels for the overall analyzing the populations as 4 subpopulations and 5 subpopulations.

Table 15. (a) AMOVA Pairwise RST Values Among the Four Subpopulations (Coolgardie Mesa and Lane Mountain considered as one population) of A. jaegerianus and (b) AMOVA Pie Graph Representing the Percentage of Molecular Variation Among and Within the Four Subpopulations.

(a)

Coolgardie Prospector Brinkman Goldstone Mesa Wash Wash Coolgardie Mesa 0.000 0.002 0.001 0.002 Prospector Wash 0.038 0.000 0.001 0.002 Brinkman Wash 0.211 0.211 0.000 0.001 Goldstone 0.063 0.063 0.328 0.000

Pairwise RST values are displayed below the diagonal and bolded p-values represent significant values based on 999 permutations above the diagonal.

(b)

Percentages of Molecular Variance

Among Pops 16%

Within Pops 84%

Table 16. (a) AMOVA Pairwise RST Values Among the Five Subpopulations of A. jaegerianus and (b) AMOVA Pie Graph Representing the Percentage of Molecular Variation Among and Within the Five Subpopulations.

(a) Coolgardie Lane Prospectors Brinkman Goldstone Mesa Mountain Wash Wash Coolgardie Mesa 0.000 0.187 0.001 0.001 0.009 Lane Mountain 0.010 0.000 0.001 0.001 0.003 Prospectors Wash 0.038 0.046 0.000 0.001 0.004 Brinkman Wash 0.259 0.186 0.211 0.000 0.001 Goldstone 0.062 0.063 0.063 0.328 0.000

Pairwise RST values are displayed below the diagonal and bolded p- values represent significant values based on 999 permutations above the diagonal.

(b)

Percentages of Molecular Variance

Among Pops 15%

Within Pops 85%

Table 17. (a) AMOVA Pairwise FST values among the 5 Subpopulation (b) AMOVA Pie Graph Representing the Percentage of Molecular Variation Within and Among Populations and Among Individuals within the 5 subpopulations.

(a) Coolgardie Lane Prospector Brinkman Goldstone Mesa Mountain Wash Wash 0.000 0.006 0.001 0.001 0.001 Coolgardie Mesa 0.015 0.000 0.001 0.001 0.001 Lane Mountain 0.042 0.047 0.000 0.001 0.001 Prospector Wash 0.095 0.087 0.046 0.000 0.001 Brinkman Wash 0.028 0.049 0.025 0.100 0.000 Goldstone

Pairwise FST values are displayed below the diagonal and bolded p-values represent significant values based on 999 permutations above the diagonal.

(b)

Percentages of Molecular Variance Among Pops 5%

Within Pops 95%

Table 18. (a) AMOVA Pairwise FST values among the 4 Subpopulation (b) AMOVA Pie Graph Representing the Percentage of Molecular Variation Within and Among Populations and Among Individuals within the 4 subpopulations.

(a) Coolgardie Prospector Brinkman Goldstone Mesa Wash Wash

0.000 0.001 0.001 0.001 Coolgardie Mesa 0.042 0.000 0.001 0.001 Prospector Wash 0.086 0.046 0.000 0.001 Brinkman Wash 0.038 0.025 0.100 0.000 Goldstone

Pairwise FST values are displayed below the diagonal and bolded p-values represent significant values based on 999 permutations above the diagonal.

(b)

Percentages of Molecular Variance Among Pops 5%

Within Pops 95%

APPENDIX C

MICROSATELLITE DATA

ASJA02 FREQUENCIES Locus Allele/n CM LM EP BW NG LocusAsja02 N 21 30 33 16 19 204 0.000 0.000 0.000 0.031 0.000 207 0.000 0.017 0.061 0.000 0.053 211 0.024 0.050 0.197 0.125 0.000 215 0.048 0.000 0.045 0.000 0.000 219 0.071 0.000 0.091 0.031 0.053 223 0.286 0.067 0.182 0.094 0.316 227 0.310 0.400 0.106 0.188 0.211 231 0.190 0.167 0.106 0.031 0.158 235 0.000 0.167 0.136 0.094 0.053 239 0.024 0.083 0.000 0.188 0.105 243 0.000 0.050 0.045 0.156 0.053 247 0.000 0.000 0.015 0.000 0.000 251 0.000 0.000 0.015 0.063 0.000 259 0.048 0.000 0.000 0.000 0.000

Allele Frequecy for ASJA02 0.450 0.400 Coolgardie Mesa 0.350 Lane Mountain 0.300 Prospector Wash 0.250 Brinkman Wash 0.200 Goldstone Frequency 0.150 0.100 0.050 0.000 204 207 211 215 219 223 227 231 235 239 243 247 251 259 ASJA 02

ASJA08 FREQUENCIES Locus Allele/n CM LM EP BW NG LocusAsja08 N 15 21 27 15 18 299 0.000 0.000 0.019 0.033 0.000 301 0.000 0.000 0.000 0.033 0.111 303 0.233 0.310 0.093 0.200 0.083 305 0.000 0.000 0.000 0.067 0.000 307 0.367 0.405 0.352 0.400 0.361 309 0.000 0.000 0.056 0.033 0.000 311 0.100 0.143 0.333 0.233 0.167 313 0.000 0.000 0.019 0.000 0.028 314 0.000 0.024 0.000 0.000 0.000 315 0.133 0.095 0.074 0.000 0.139 317 0.000 0.000 0.019 0.000 0.000 319 0.167 0.000 0.037 0.000 0.111 323 0.000 0.024 0.000 0.000 0.000

Allele Frequency Coolgardie Mesa 0.450 Lane Mountain 0.400 Prospector 0.350 Wash 0.300 Brinkman Wash 0.250 0.200

Frequency 0.150 0.100 0.050 0.000 299 301 303 305 307 309 311 313 314 315 317 319 323 ASJA08

ASJA17 FREQUENCIES Locus Allele/n Pop1 Pop2 Pop3 Pop4 Pop5 LocusAsja17 N 22 31 34 29 20 273 0.000 0.000 0.000 0.017 0.000 281 0.000 0.016 0.000 0.000 0.050 285 0.545 0.806 0.544 0.672 0.350 289 0.000 0.000 0.000 0.017 0.000 293 0.114 0.065 0.059 0.017 0.000 297 0.000 0.000 0.029 0.000 0.000 301 0.023 0.016 0.059 0.017 0.075 305 0.114 0.016 0.147 0.138 0.025 309 0.091 0.016 0.074 0.052 0.175 313 0.114 0.048 0.044 0.052 0.175 317 0.000 0.000 0.029 0.000 0.125 321 0.000 0.016 0.015 0.000 0.000 337 0.000 0.000 0.000 0.017 0.000 345 0.000 0.000 0.000 0.000 0.025

Allele Frequency Coolgardie 0.900 Mesa Lane Mountain 0.800 0.700 Prospector Wash 0.600 Brinkman Wash 0.500 0.400

Frequency 0.300 0.200 0.100 0.000 273 281 285 289 293 297 301 305 309 313 317 321 337 345 ASJA17

ASJA19 FREQUENCIES Locus Allele/n Pop1 Pop2 Pop3 Pop4 Pop5 LocusAsja19 N 23 28 34 16 20 380 0.000 0.000 0.015 0.000 0.000 388 0.000 0.000 0.000 0.000 0.025 396 0.000 0.000 0.000 0.063 0.000 400 0.087 0.054 0.059 0.031 0.000 404 0.065 0.107 0.044 0.000 0.025 408 0.043 0.089 0.118 0.281 0.150 412 0.022 0.071 0.191 0.156 0.075 416 0.022 0.054 0.176 0.188 0.100 420 0.370 0.179 0.118 0.063 0.275 424 0.065 0.143 0.088 0.063 0.125 428 0.065 0.179 0.029 0.094 0.125 432 0.087 0.000 0.059 0.000 0.000 436 0.000 0.054 0.074 0.031 0.050 440 0.000 0.054 0.029 0.031 0.025 444 0.087 0.018 0.000 0.000 0.000 448 0.022 0.000 0.000 0.000 0.000 452 0.065 0.000 0.000 0.000 0.000 456 0.000 0.000 0.000 0.000 0.025

Allele Frequency Coolgardie Mesa 0.400 Lane Mountain 0.350 Prospector 0.300 Wash 0.250 Brinkman Wash 0.200 0.150 0.100 0.050 0.000 380 388 396 400 404 408 412 416 420 424 428 432 436 440 444 448 452 456 LocusAsja19

ASJA25 FREQUENCIES Locus Allele/n Pop1 Pop2 Pop3 Pop4 Pop5 LocusAsja25 N 23 30 36 29 20 177 0.000 0.000 0.000 0.000 0.050 185 0.348 0.300 0.167 0.293 0.125 189 0.087 0.083 0.014 0.000 0.000 193 0.065 0.167 0.431 0.362 0.400 197 0.022 0.000 0.014 0.069 0.025 201 0.022 0.033 0.042 0.069 0.000 205 0.239 0.300 0.153 0.017 0.075 209 0.152 0.083 0.083 0.034 0.125 213 0.000 0.000 0.014 0.103 0.100 217 0.065 0.017 0.056 0.017 0.025 221 0.000 0.017 0.000 0.034 0.050 225 0.000 0.000 0.014 0.000 0.025 229 0.000 0.000 0.014 0.000 0.000

Allele Frequency Coolgardie Mesa 0.500 Lane Mountain 0.450 0.400 Prospector 0.350 Wash Brinkman Wash 0.300 0.250

0.200 Frequency 0.150 0.100 0.050 0.000 177 185 189 193 197 201 205 209 213 217 221 225 229 ASJA25

ASJA38 FREQUENCIES Locus Allele/n Pop1 Pop2 Pop3 Pop4 Pop5 LocusAsja38 N 10 12 27 26 9 182 0.050 0.000 0.000 0.000 0.056 194 0.350 0.250 0.296 0.154 0.167 198 0.500 0.500 0.574 0.654 0.556 202 0.100 0.250 0.111 0.192 0.222 218 0.000 0.000 0.019 0.000 0.000

Allele Frequency 0.700

0.600

0.500 Coolgardie Mesa 0.400 Lane Mountain 0.300

Frequency Prospector Wash 0.200 Brinkman Wash 0.100 Goldstone 0.000 182 194 198 202 218 ASJA38

ASJA45 FREQUENCIES Locus Allele/n Pop1 Pop2 Pop3 Pop4 Pop5 LocusAsja45 N 20 27 32 25 19 258 0.000 0.000 0.016 0.000 0.053 262 0.075 0.074 0.094 0.140 0.026 266 0.000 0.037 0.000 0.020 0.053 270 0.350 0.241 0.234 0.460 0.211 274 0.125 0.185 0.094 0.040 0.000 278 0.075 0.074 0.219 0.080 0.237 282 0.150 0.148 0.109 0.100 0.237 286 0.075 0.056 0.141 0.100 0.158 290 0.125 0.056 0.078 0.020 0.026 294 0.025 0.111 0.016 0.000 0.000 298 0.000 0.019 0.000 0.040 0.000

Allele Frequency 0.500 Coolgardie Mesa 0.450 Lane Mountain 0.400 Prospector Wash 0.350 Brinkman Wash 0.300 0.250 Goldstone

0.200 Frequency 0.150 0.100 0.050 0.000 258 262 266 270 274 278 282 286 290 294 298 ASJA45

APPENDIX D

INPUT FILES

GENALEX MICROSATELLITE DATA FILE – 4 POPULATIONS

GENALEX MICROSATELLITE DATA FILE – 5 POPULATIONS

STRUCTURE PARAMETER SET: SIMULATION CONFIGURATION Parameter Set: Long Run

Running Length Length of Burn-in Period: 10000 Number of MCMC Reps after Burn-in: 1000000 Ancestry Model Info Use Admixture Model *Infer Alpha *Initial Value of ALPHA (Dirichlet Parameter for Degree of Admixture): 1.0 *Use Same Alpha for all Populations *Use a Uniform Prior for Alpha **Maximum Value For Alpha: 10.0 **SD of Proposal for Updating Alpha: 0.025

Frequency Model Info Allele Frequencies are Correlated among Pops * Assume Different Values of Fst for Different Subpopulations * Prior Mean of Fst for Pops: 0.01 * Prior SD of Fst for Pops: 0.05 * Use Constant Lambda (Allele Frequencies Parameter) * Value of Lambda: 1.0

Advanced Options

Estimate the Probability of the Data Under the Model Frequency of Metropolis update

REFERENCES

Ardren, W. R., Borer, S., Thrower, F., Joyce, J. E., & Kapuscinski, A. R. (1999).

Inheritance of 12 microsatellite loci in Oncorhynchus mykiss. Journal of

Heredity, 90(5), 529–536.

Avise, John C. (2004). Molecular Markers, Natural History, and Evolution (2nd

Edition). Sinauer Associates, Inc. Publishers.

Baldwin, B. G., Goldman, D., Keil, D. J., Patterson, R., & Rosatti, T. J. (2012).

The Jepson Manual: Vascular Plants of California. University of California

Press.

Balloux, F., & Lugon-Moulin, N. (2002). The estimation of population

differentiation with microsatellite markers. Molecular Ecology, 11(2), 155–

165.

Barneby, R. C. (1964). Atlas of North American Astragalus, New York Botanical

Garden.

Baskauf, C. J., & Burke, J. M. (2009). Population genetics of Astragalus

bibullatus (Fabaceae) using AFLPs. Journal of Heredity, 100(4), 424–431.

Bateman, A. J. (1947). Contamination of seed crops. Journal of Genetics, 48(2),

257–275.

Castoe, T. A., Poole, A. W., Koning, A. P. J. de, Jones, K. L., Tomback, D. F.,

Oyler-McCance, S. J., Fike, J. A., Lance, S. L., Streicher, J. W., Smith, E.

N., & Pollock, D. D. (2012). Rapid Microsatellite Identification from Illumina

Paired-End Genomic Sequencing in Two Birds and a Snake. PLOS ONE,

7(2), e30953.

Despres, L., Loriot, S., & Gaudeul, M. (2002). Geographic pattern of genetic

variation in the European globeflower Trollius europaeus L.

(Ranunculaceae) inferred from amplified fragment length polymorphism

markers. Molecular Ecology, 11(11), 2337–2347.

Dewoody, J., Nason, J. D., & Hipkins, V. D. (2006). Mitigating scoring errors in

microsatellite data from wild populations. Molecular Ecology Notes, 6(4),

951–957. https://doi.org/10.1111/j.1471-8286.2006.01449.x

Don, R. H., Cox, P. T., Wainwright, B. J., Baker, K., & Mattick, J. S. (1991).

“Touchdown” PCR to circumvent spurious priming during gene

amplification. Nucleic Acids Research, 19(14), 4008.

Ellstrand, N. C., & Elam, D. R. (1993). Population Genetic Consequences of

Small Population Size: Implications for Plant Conservation. Annual Review

of Ecology and Systematics, 24(1), 217–242.

Endangered and Threatened Wildlife and Plants; 12-Month Finding on a Petition

to Reclassify Astragalus jaegerianus as a Threatened Species. (2014). In

the Federal Register / FIND (Vol. 79, Issue 85, pp. 25084-). Federal

Information & News Dispatch, LLC.

Eriksson, Är. (1998). Regional distribution of Thymus serpyllum: Management

history and dispersal limitation. Ecography, 21(1), 35–43.

Evanno, G., Regnaut, S., & Goudet, J. (2005). Detecting the number of clusters

of individuals using the software structure: A simulation study. Molecular

Ecology, 14(8), 2611–2620. https://doi.org/10.1111/j.1365-

294X.2005.02553.x

Falush, D., Stephens, M., & Pritchard, J. K. (2003). Inference of population

structure using multilocus genotype data: Linked loci and correlated allele

frequencies. Genetics, 164(4), 1567–1587.

Falush, D., Stephens, M., & Pritchard, J. K. (2007). Inference of population

structure using multilocus genotype data: Dominant markers and null

alleles. Molecular Ecology Notes, 7(4), 574–578.

Frankham, R. (1995). Conservation Genetics. Annual Review of Genetics, 29,

305–327.

Gaston, K. J. (2003). Structure and Dynamics of Geographic Ranges: Structure

and Dynamics of Geographic Ranges. Oxford University Press.

Gaudeul, M., Till-Bottraud, I., Barjon, F., & Manel, S. (2004). Genetic diversity

and differentiation in Eryngium alpinum L. (Apiaceae): Comparison of

AFLP and microsatellite markers. Heredity, 92(6), 508–518.

https://doi.org/10.1038/sj.hdy.6800443

Gerber, S., Mariette, S., Streiff, R., Bodénès, C., & Kremer, A. (2000).

Comparison of microsatellites and amplified fragment length

polymorphism markers for parentage analysis. Molecular Ecology, 9(8),

1037–1048.

Gibson, A. C. (1998). Photosynthetic Organs of Desert Plants. BioScience, 48(11), 911–920.

Hamrick, J. L., Godt, M. J. W., Murawski, D. A., & Loveless, M. D. (1991).

Correlations between species traits and allozyme diversity: Implications for

conservation biology. Genetics and Conservation of Rare Plants, 75–86.

Hamrick, J. L., Linhart, Y. B., & Mitton, J. B. (1979). Relationships Between Life

History Characteristics and Electrophoretically Detectable Genetic

Variation in Plants. Annual Review of Ecology & Systematics, 10, 173–

200.

Hamrick, J. Lꎬ, & Godt, M. W. (1990). Allozyme diversity in plant species. Plant

Population Genetics, Breeding, and Genetic Resources., 43–63.

Hardy, O. J., Charbonnel, N., Fréville, H., & Heuertz, M. (2003). Microsatellite

allele sizes: A simple test to assess their significance on genetic

differentiation. Genetics, 163(4), 1467–1482.

Hartl, D. L., & Clark, A. G. (2006). Principles of Population Genetics. (4th ed.).

Sinauer Associates, Inc. Publishers.

Hohman, J. (2014). Species Report for Lane Mountain Milk-vetch (Astragalus

jaegerianus). U.S. Fish and Wildlife Service Ventrua Fish and Wildlife

Office, 132.

Jakse, J., Kindlhofer, K., & Javornik, B. (2001). Assessment of genetic variation

and differentiation of hop genotypes by microsatellite and AFLP markers.

Genome; Ottawa, 44(5), 773–782.

Kalia, R. K., Rai, M. K., Kalia, S., Singh, R., & Dhawan, A. K. (2011).

Microsatellite markers: An overview of the recent progress in plants.

Euphytica, 177(3), 309–334.

Karron, J. D. (1987). The pollination ecology of co-occuring geographically

restricted and widespread species of Astragalus (Fabaceae). Biological

Conservation, 39(3), 179–193.

Karron, J. D. (1991). Breeding Systems in Rare Plant Species. Genetics and

Conservation of Rare Plants, 87.

Kearns, D. (2003). Lane Mountain milk-vetch pollination report. Prepared by

Charis Professional Services Corporation, Temecula, California for the U.

S. Army National Training Center, Fort Irwin, California.

Leberg, P. L. (1996). Applications of allozyme electrophoresis in conservation

biology. Molecular Genetic Approaches in Conservation, 87–103.

Lynch, M., Conery, J., & Burger, R. (1995). Mutation accumulation and the

extinction of small populations. The American Naturalist, 146(4), 489–518.

McCauley, D. E. (Vanderbilt U. (1997). The relative contributions of seed and

pollen movement to the local genetic structure of Silene alba. The Journal

of Heredity (USA), 88, 257–263.

Meirmans, P. G. (2012). AMOVA-Based Clustering of Population Genetic Data.

Journal of Heredity, 103(5), 744–750.

https://doi.org/10.1093/jhered/ess047

Metcalf, A. E., Nunney, L., & Hyman, B. C. (2001). Geographic Patterns of

Genetic Differentiation Within the Restricted Range of the Endangered

Stephens’kangaroo Rat Dipodomys stephensi. Evolution, 55(6), 1233–

1244.

Mueller, U. G., & Wolfenbarger, L. L. (1999). AFLP genotyping and fingerprinting.

Trends in Ecology & Evolution, 14(10), 389–394.

Mullis, K., Faloona, F., Scharf, S., Saiki, R., Horn, G., & Erlich, H. (1986).

Specific enzymatic amplification of DNA in vitro: The polymerase chain

reaction. Cold Spring Harbor Symposia on Quantitative Biology, 51 Pt 1,

263–273.

Munz, PA. (1941). Interesting western plants—V. Leaflets of Western Botany, 3,

49–50.

Nei M. 1987. Molecular Evolutionary Genetics. New York: Columbia University

Press. 512 pages.

Nerkowski, S. A. (2015). Microsatelite analysis of population structure in Santa

Ana speckled dace (Rhinichthys osculus). California State University, San

Bernardino.

Nunziata, S. O., Lance, S. L., Jones, K. L., Nerkowski, S. A., & Metcalf, A. E.

(2012). Development and characterization of twenty-three microsatellite

markers for the freshwater minnow Santa Ana speckled dace (Rhinichthys

osculus spp., Cypreinidea) using paired-end Illumina shotgun sequencing.

Conservation Genetic Resources, 5(5), 145–148.

O’Reilly, P., & Wright, J. M. (1995). The evolving technology of DNA

fingerprinting and its application to fisheries and aquaculture. Journal of

Fish Biology, 47, 29–55.

Panda, S., Naik, D., & Kamble, A. (2015). Population structure and genetic

diversity of the perennial medicinal shrub Plumbago. AoB PLANTS, 7.

Peakall, R., & Smouse, P. E. (2006). genalex 6: Genetic analysis in Excel.

Population genetic software for teaching and research. Molecular Ecology

Notes, 6(1), 288–295.

Pierson, E. A., & Mack, R. N. (1990). The population biology of Bromus tectorum

in forests: Distinguishing the opportunity for dispersal from environmental

restriction. Oecologia, 84(4), 519–525.

Porras-Hurtado, L., Ruiz, Y., Santos, C., Phillips, C., Carracedo, Á., & Lareu, M.

V. (2013). An overview of STRUCTURE: Applications, parameter settings,

and supporting software. Frontiers in Genetics, 4.

https://doi.org/10.3389/fgene.2013.00098

Prigge, B. A., Sharifi, M. R., & Rundel, P. W. (2012). Community Dynamics and

Soil Seed Bank Ecology of Lane Mountain Milkvetch (Astragalus

jaegerianus Munz). DTIC Document.

Richards, C. M., Church, S., & McCauley, D. E. (1999). The influence of

population size and isolation on gene flow by pollen in Silene alba.

Evolution, 53(1), 63–73.

Roll, J., Mitchell, R. J., Cabin, R. J., & Marshall, D. L. (1997). Reproductive

Success Increases with Local Density of Conspecifics in a Desert Mustard

(Lesquerella fendleri). Conservation Biology, 11(3), 738–746.

Rundel, P. W., Prigge, B. A., & Sharifi, M. R. (2007). The Effect of Amount and

Frequency of Precipitation on Seedling Establishment and Survival of

Lane Mountain Milkvetch (Astragalus jaegerianus Munz).

Sainudiin, R., Durrett, R. T., Aquadro, C. F., & Nielsen, R. (2004). Microsatellite

Mutation Models. Genetics, 168(1), 383–395.

Sanger, F., Nicklen, S., & Coulson, A. R. (1977). DNA sequencing with chain-

terminating inhibitors. Proceedings of the National Academy of Sciences

of the United States of America, 74(12), 5463–5467.

Savelkoul, P. H. M., Aarts, H. J. M., Haas, J. de, Dijkshoorn, L., Duim, B., Otsen,

M., Rademaker, J. L. W., Schouls, L., & Lenstra, J. A. (1999). Amplified-

Fragment Length Polymorphism Analysis: The State of an Art. Journal of

Clinical Microbiology, 37(10), 3083–3091.

Schmidt, K., & Jensen, K. (2000). Genetic structure and AFLP variation of

remnant populations in the rare plant Pedicularis palustris

(Scrophulariaceae) and its relation to population size and reproductive

components. American Journal of Botany, 87(5), 678–689.

Segarra-Moragues, J. G., Palop-Esteban, M., González-Candelas, F., & Catalán,

P. (2005). On the verge of extinction: Genetics of the critically endangered

Iberian plant species, Borderea chouardii (Dioscoreaceae) and

implications for conservation management. Molecular Ecology, 14(4),

969–982.

Slatkin, M. (1995). A measure of population subdivision based on microsatellite

allele frequencies. Genetics, 139(1), 457–462.

Travis, S. E., Maschinski, J., & Keim, P. (1996). An analysis of genetic variation

in Astragalus cremnophylax var. Cremnophylax, a critically endangered

plant, using AFLP markers. Molecular Ecology, 5(6), 735–745.

Trostel, K. (2017). Population Genetics of the Rare Cedar Glade Endemic

Astragalus bibullatus (Fabaceae) Using Nuclear Microsatellite. Middle

Tennessee State University.

Turchetto, C., Segatto, A. L. A., Mäder, G., Rodrigues, D. M., Bonatto, S. L., &

Freitas, L. B. (2016). High levels of genetic diversity and population

structure in an endemic and rare species: Implications for conservation.

AoB PLANTS, 8. https://doi.org/10.1093/aobpla/plw002

U.S. Fish and Wildlife Service. 2001. Draft recovery plan for the Lane Mountain

Milk vetch (Astragalus jaegerianus). U.S. Fish and Wildlife Service,

Portland, Oregon. Xx-pages.

Valdes, A. M., Slatkin, M., & Freimer, N. B. (1993). Allele frequencies at

microsatellite loci: The stepwise mutation model revisited. Genetics,

133(3), 737–749.

Vos, P., Hogers, R., Bleeker, M., Reijans, M., Van de Lee, T., Hornes, M., Friters,

A., Pot, J., Paleman, J., Kuiper, M., & others. (1995). AFLP: A new

technique for DNA fingerprinting. Nucleic Acids Research, 23(21), 4407–

4414.

Walker, G. F. (2005). Analysis of Molecular Variation in the Federally

Endangered Astragalus jaegerianus (Fabaceaea, Papilionoideae): A

Species with a Restricted Geographic Range. California State University,

San Bernardino.

Walker, G. F., & Metcalf, A. E. (2008). Genetic variation in the endangered

Astragalus jaegerianus (Fabaceae, Papilionoideae): A geographically

restricted species. Bulletin, Southern California Academy of Sciences,

107(3), 158–177.

Wang, X.-R., & Szmidt, A. E. (2001). Molecular Markers in Population Genetics

of Forest Trees. Scandinavian Journal of Forest Research, 16(3), 199–

220.

Weir, B. S. (2012). Estimating F-Statistics: A Historical View. Philosophy of

Science, 79(5), 637–643.

Wijayratne, U. C., Scoles-Sciulla, S. J., & Defalco, L. A. (2009). Dust Deposition

Effects on Growth and Physiology of the Endangered Astragalus

jaegerianus (Fabaceae). Madroño, 56(2), 81–88.

https://doi.org/10.3120/0024-9637-56.2.81

Wittke-Thompson, J. K., Pluzhnikov, A., & Cox, N. J. (2005). Rational Inferences

about Departures from Hardy-Weinberg Equilibrium. The American

Journal of Human Genetics, 76(6), 967–986.

Wright, S. (1931). Evolution in Mendelian Populations. Genetics, 16(2), 97–159.

Wright, S. (1965). The Interpretation of Population Structure by F-Statistics with

Special Regard to Systems of Mating. Evolution, 19(3), 395–420.

Young, A., Boyle, T., & Brown, T. (1996). The population genetic consequences

of habitat fragmentation for plants. Trends in Ecology & Evolution, 11(10),

413–418. https://doi.org/10.1016/0169-5347(96)10045-8