Examination of the population structure of darkbarbel catfish (Pelteobagrus vachelli) in the Upper Yangtze River Basin, China, using novel microsatellite markers

By Adrienne Powell

A thesis submitted to the Department of Biology in conformity with the requirements for the degree of Master of Science

Queen’s University Kingston, Ontario, Canada June, 2012 Copyright © Adrienne Powell, 2012 ii

Abstract:

The darkbarbel catfish (Peltoebagrus vachelli) is a small bodied benthic fish that inhabits the Yangtze River in China and is commercially valued as food. The objectives of this project were to develop the first set of microsatellite primers specific to P. vachelli and use them to examine the levels of genetic variability and population structure of three populations collected at three sites (Yibin, Luzhou, and Song Ji) along an unfragmented stretch of the upper Yangtze River. Microsatellite primers were designed from an enriched library of genomic DNA and a total of eight primer pairs were optimized to produce reliable polymorphic PCR amplicons on a LI-COR 4200 IR2 platform.

A high level of variability was detected amongst the isolated loci with the number of alleles per locus ranging from 14 – 29 and mean observed and expected heterozygosity values of 0.84 and 0.90 respectively across the entire sample set. Overall levels of genetic diversity were high within the three populations with mean observed and expected heterozygosities ranging from 0.823 – 0.869 and 0.864 – 0.921 respectively.

Little evidence of genetic structure was detected (global FST = 0.0065, p < 0.05) within the sampled region and pairwise tests of differentiation were not significant (all FST p >

0.05). These results imply historically large populations of P. vachelli in the upper portion of the Yangtze River that, in the absence of artificial barriers, are part of a single panmictic unit. As plans have been put forth for construction of hydroelectric installations within the sampled region, this study provides a baseline estimate of the levels of genetic variation present in P. vachelli within the remaining undammed stretch of the Upper Yangtze River which will serve as a foundation for future analyses on the iii

effects of river fragmentation within this region. Additionally the isolation and characterization of the microsatellites will provide useful molecular markers for a variety of other applications in this and closely related species such as parentage analysis, determination of stocks, and maintenance of genetic variation in stocking practices.

iv

Acknowledgements:

I would like to thank my supervisors, Drs. Peter Boag and Yu-Xiang Wang, without whom this project would not have been possible. My appreciation also goes to

Dr. John Casselman as an integral member of my committee who gave support and insightful comments throughout the process of completing my Master’s degree.

Dr. Brent Murray from the University of Northern British Columbia was instrumental in the designing, planning, and carrying out the sampling of specimens used in this study. Thank you also to the students who helped with the sample collection: Bo-Jian Chen, Lily Yao, Yong-Peng Chen, and Dr. Shi-Jian Fu from Chongqing

Normal University as well as Mi Wang, Hong Qing and Xiang Chang from Southwestern

University.

I would also like to express my sincere gratitude to Zhengxin Sun for providing me with the protocol for the microsatellite library development and guiding me through the process as well as locating supplies necessary to complete the task. I appreciate the support and encouragement that you extended over the years.

Thank you to also to members of the Boag lab including Candace Scott for designing the primers and providing invaluable knowledge with regards to laboratory methods amongst other things. Other members of the lab including Christopher Harris,

Stephanie McCormack, Ozge Danisment, and Jesse Lai provided technical support and laboratory assistance during the data collection process. v

From the Wang lab, I would like to extend my appreciation to Victoria Kopf, Dr.

Yi-Ping Luo, Mingzhi Qu and various other office buddies for their company and random conversations along the way.

I would also like to express my appreciation to Dr. Peter Van Cooverden deGroot for his support, encouragement, and advise on a variety of issues.

Others in the department including Anne Curtis, Amy Chabot, Andy Wong,

Sharilyn Hoobin, Roxane Razavi, and Lori Parker provided knowledge and support along the way.

I would also like to thank my family, particularly my grandmothers, my parents and my many brothers and sister as well as my friends for their support throughout this often challenging process.

vi

Table of Contents:

Abstract ii

Acknowledgements iv

List of Tables ix

List of Figures and Illustrations x

List of Abbreviations xi

Chapter 1: Introduction and Literature Review 1

River Fragmentation 1 Genetic Consequences of River Fragmentation 1 Biodiversity in the Yangtze River 2 High Levels of Biodiversity in the Upper Yangtze River 3 Dam Development in the Yangtze Basin 3 Consequences of Dam Construction along the Yangtze River 4 Case Study: The Chinese Sturgeon 6 Population Structure in the Yangtze River 7 Genetic Impacts of Fragmentation along the Yangtze 8 Molecular Tools Applicable to Fisheries Biology 8 Microsatellites 9 Catfish Evolutionary History 10 Siluriformes 11 Catfish Phylogeny 12 The Family Bagridae 12 Genera of the Family Bagridae 13 The study species: Pelteobagrus vachelli 14 Population Structure of Pelteobagrus vachelli 16 Effects of Dam Construction on Pelteobagrus vachelli 17 Purpose of This Study 18

Chapter 2: Materials and Methods 20

The Study Area: The Yangtze River 20 The Upper Yangtze 20 Sample Collection 22 Microsatellite Library Development 22 Isolation of Genomic DNA 22 Restriction Enzyme Digestion of Genomic DNA 24 Preparation of Double Stranded SNX Linkers 25 Ligation of DNA Fragments to SNX Linkers and Verification 26 vii

Enrichment of the Microsatellite gDNA-Linker with Biotin-oligos 26 PCR to Test the Results of the Enrichment 28 Dot Blot of the Enriched Microsatellite gDNA-Linker 29 Ligation of Enriched Microsatellite gDNA-Linker to TOPO TA Vector 31 Identification of Positive Colonies by Hybridization with Probes 32 Growth of Positive Colonies 34 Plasmid Isolation and Sequencing of Inserts 35 Primer Design 35 Primer Optimization 35 Multiplex PCR Optimization 37 Genomic DNA Isolation of the Sample Set 37 PCR Amplification of Loci 38 Amplification Error Estimation 38 Data Analysis 40

Chapter 3: Results 43

Microsatellite Library Construction and Optimization 43 Genotyping Error Estimates 43 Microsatellite Genotyping 44 Assessment of Genotyping Technical Errors 44 Genetic Variability of the Optimized Alleles 44 Tests of Linkage Disequilibrium across Populations 46 Private Alleles 46 Genetic Variation within Populations 46 Deviations from Hardy-Weinberg equilibrium within Populations 49 Linkage Disequilibrium between Pairs of Loci within Populations 50 Population Structure 51

Chapter 4: Discussion 55

Microsatellite Variability 55 Comparisons with Other Bagrid Species 56 High Levels of Genetic Diversity 57 Population Structure 59 Models of Gene Flow 61 High Variability and Lack of Population Differentiation 62 Limitations of this Study 65 Final Considerations 66

Literature Cited 67

Appendix 1: Allele frequencies per population 82

Appendix 2: Three dimensional plot of the principle coordinates analysis 85 viii

Appendix 3: Literature summary of the size range of microsatellite alleles 86

Appendix 4: Literature summary of levels of heterozygisty in fish species 87

ix

List of Tables:

Table 1: Sampling sites, GPS coordinates, the number of individual Pelteobagrus 23 vachelli sampled, and the range of fork length (FL) in cm from the three locations along the Upper Yangtze River, China.

Table 2: Microsatellite loci isolated in this experiment with the locus specific 39 forward and reverse primer sequences, the optimized annealing temperature

(Tm), whether or not it was amplified as part of a multiplex, and the expected allele size. MgCl2 concentration was 1.5 mM in all cases.

Table 3: Summary statistics per locus and averages across loci including the 45 number of alleles, standardized allelic richness (AR), range in allele size (in base pairs), expected heterozygosity (He), observed heterozygosity (Ho), and the polymorphic information content of each locus (PIC) and the averages across loci for the 61 samples included in the analyses.

Table 4: Number of private alleles (P), total number of alleles (A) and the 47 percentage (%) of the total identified for each locus within the Song Ji, Luzhou, and Yibin populations and over all populations.

Table 5: Genetic diversity at eight microsatellite loci for three populations along 48 the upper Yangtze River, China, including the total number of alleles (A), allelic richness (AR) standardized to 20 individuals per population, expected heterozygosity (HE), observed heterozygosity (HO), and the p-value associated with the exact test for deviation from Hardy-Weinberg equilibrium.

Table 6: AMOVA analysis averaged across loci. Global FST = 0.00648 over all loci 52 (distance matrix calculated from the number of different alleles), 95% C.I. (0.00137 – 0.01110), P (rand > obs value) = 0.04703.

Table 7: Pairwise FST values among the three populations (above the diagonal) 53 with the corresponding p-values (below the diagonal).

x

List of Figures and Illustrations:

Figure 1: Map of the upper Yangtze River system where: XLD represents Xiluodu; 5 XJB Xiajiaba; PCZ the Pianchuangzi; SP, Shipeng; ZYX, Zhuyangxi; XNH, Xiaonanhai; HC, Hechuan; LXH, Longxie; PS, Pengshui; DXK, Daxikou; TGD, Three Gorges Dam; GZB, Gezhouba Dam. The life history information refers to Dabry’s sturgeon (Acipenser dabryanus). (Figure taken from Zhang et al., 2011)

Figure 2: Photo of an individual Pelteobagrus vachelli fish collected during the 15 sampling procedure (Photo courtesy of Dr. B. Murray)

Figure 3: Map of the sampling sites along the upper Yangtze River, China, 21 between the cities of Yibin and Chongqing. The sampling sites are marked 1 = Yibin, 2 = Luzhou, and 3 = Song Ji.

Figure 4: Two dimensional plots of the samples from the Song Ji, Luzhou, 54 and Yibin populations on a) principle coordinate axes 1 and 2; b) axes 2 and 3; c) axes1 and 3.

xi

List of Abbreviations:

A Total number of alleles AR Allelic richness AMOVA Analysis of molecular variance bp Base pairs C.I. Confidence interval CIAP Calf Intestine Alkaline Phosphotase DMSO Dimethyl sulphoxide DIG Digoxigenin ds Double stranded EDTA Ethylenediamine tetra acetic acid dipotassium salt dehydrate FST Fixation index FL Fork length gDNA Genomic DNA HE Expected heterozygosity HO Observed heterozygosity HWE Hardy-Weinberg equilibrium MPC Magnetic Particle Collecting unit M13 F M13 Forward primer P Private alleles PBS Phosphate Buffered Saline SDS Sodium dodecyl sulfate SNX “SNX” linker (Hamilton et al., 1999) SSC Saline sodium citrate buffer Tm Annealing temperature TAE Tris-Acetate-EDTA TBE Tris-Borate-EDTA TLE Tris-Low-EDTA Tris Tris(hydroxmethyl)aminomethane PIC Polymorphic information content

1

Chapter 1: Introduction:

River Fragmentation

River fragmentation through the construction of hydroelectric installations may cause severe impacts on local biological diversity and ecosystems in the surrounding area

(Pringle et al., 2000). As such, dam development is increasingly viewed as the largest anthropogenic threat to aquatic species diversity in many rivers globally (Fan et al.,

2006). In particular dams affect fish by blocking their migratory routes, fragmenting populations, disturbing their spawning grounds, and changing the hydrological conditions to which they are adapted (Pringle et al., 2000; Wu et al., 2003; Fu et al., 2003; Chen et al., 2009). For these reasons dams can produce many changes in fish community structure including: a decrease in the number of species, especially flood dependent and migratory ones; an increase in exotic species; decreases in population sizes; and the elimination of rare species (Pringle et al., 2000; Chen et al., 2009).

Genetic Consequences of River Fragmentation

Until recently, few studies had investigated the genetic changes in fish populations associated with the construction of dams (Eg. Neeras and Spruell, 2001;

Heggenes and Roed, 2006; Deiner et al., 2007; Reid et al., 2008a). However, it is now recognized that river fragmentation through the construction of dams and other artificial barriers to flow including weirs and sluice gates can generate changes in the levels and distribution of genetic diversity within and among fish populations (Eg. Meldgaard et al.,

2003; Laroche and Durand, 2004; Beneteau et al., 2009). These barriers to fish 2 movement can lead to decreased genetic diversity in isolated populations, increased differentiation among populations, altered patterns of gene flow, and upstream to downstream gradients in genetic variation (Yamamoto et al., 2004; Dehais et al., 2010;

Yamazaki et al., 2011). Moreover, the effects of river fragmentation have been shown to affect different species to varying degrees, depending on factors including the permeability of barriers, life history traits and the natural distribution of the species of interest, as well as the structure of the aquatic landscape (Pringle et al., 2000; Haponski et al., 2007; Skalski et al., 2008; Reid et al., 2008a).

Biodiversity in the Yangtze River

As the third largest river in the world at over 6300 km in length, the Yangtze River is home to rich biodiversity (Fu et al., 2003; Chen et al., 2009). The changing geological and climatic conditions associated with the uplift of the Tibetan plateau during the

Pleistocene led to a wide assortment of ecological environments throughout the river basin which has resulted in a vast collection of species with varying life-history characteristics (Chen et al., 2009; Duan et al., 2009). With more than 350 fish species, of which 177 are endemic, this river is the most species rich in the palearctic region (Fu et al., 2003; Park et al., 2003). Concerns about the maintenance of this valuable biological resource have arisen as fish in the Yangtze are depleted through a combination of factors including habitat fragmentation, overfishing, water quality deterioration and invasion of exotic species (Chen et al., 2009). Of these, construction of hydrological installations is considered to pose the greatest risk. Hence there is a need for scientific studies that provide information on the fish resources in the Yangtze River to form and implement 3 effective management strategies for the protection and utilization of fishery stocks (Fu et al., 2003; Chen et al., 2009)

High Levels of Biodiversity in the Upper Yangtze River

Within the Yangtze basin, the richest levels of fish biodiversity are found in the upper reaches (Fu et al., 2003). In the main stem alone, there are 162 species including

44 that are endemic and at serious risk of extinction with continued dam development

(Park et al., 2003). This has led some authors to propose that conservation initiatives, especially for fish in the upper reaches of the Yangtze basin, should be considered an immediate priority (Fu et al., 2003; Fan et al., 2006).

Dam Development in the Yangtze Basin

The Yangtze River Basin is currently the most threatened aquatic environment in the world in terms of development pressure in the form of hydroelectric dams, with more than 40 large dams either planned or under construction (WWF, 2004). The construction of hydroelectric impoundments in the main stem of the Yangtze River began with the construction of the Gezhouba Dam in 1981 (Fan et al., 2006). This was followed with the construction of the Three Gorges Dam, the largest hydroelectric project in the world with a potential for producing 2.65 Billion Kw hours of electricity per year, and which began producing power in 2003 (Xiao et al., 2006). This dam modified the flow of the Yangtze which had historically been a fast flowing river in this section and created a reservoir that stretches for more than 600 km upriver to the municipality of Chongqing

(Fu et al., 2003). The greatest potential for hydropower along the Yangtze River is in the 4 upper reaches and the majority of dams that are now planned or under construction are being developed within this portion of the river (Yonghui et al., 2006). Development began within this region in 2003 with the construction of the Xiluodu Dam which was soon followed by the Xiangjiaba Dam completed in 2008 (Yonghui et al., 2006; Zhang et al., 2011). Several other dams are now either planned or under construction which will lead to dramatic changes to the hydrologic conditions and pose serious risk to the continued survival of many of the rare and culturally important species within the region

(Yonghui et al., 2006; Zhang et al., 2011).

Consequences of Dam Construction along the Yangtze River

Since the commencement of hydroelectric development along the Yangtze River, significant changes to fish community structure have already been noted including a decline in the number of species, particularly migratory ones, a decrease in the size and age of fish, and an increase in the number of exotic species (Chen et al., 2009; Fu et al.,

2003). In 2000, a national level aquatic nature reserve was established in the upper

Yangtze region to protect rare and endemic species that covers a stretch of the main stem from Leibo to Hejiang (Chongqing) and in 2005 it was expanded to include some of the major tributaries (Fan et al., 2006). This reserve was developed to mitigate the impacts of dam construction. However, despite the initial objectives of this reserve, there continues to be development in the form of hydroelectric installations within the boundaries of the reserve (Zhang et al., 2011; Figure 1). As Fu et al. (2003) noted, there is a need for more stringent conservation practices especially for the fish in the upper

Yangtze region. 5

Figure 1: Map of the upper Yangtze River system where: XLD represents Xiluodu; XJB Xiajiaba; PCZ the Pianchuangzi; SP, Shipeng; ZYX, Zhuyangxi; XNH, Xiaonanhai; HC, Hechuan; LXH, Longxie; PS, Pengshui; DXK, Daxikou; TGD, Three Gorges Dam; GZB, Gezhouba Dam. The life history information refers to Dabry’s sturgeon (Acipenser dabryanus). (Figure taken from Zhang et al., 2011)

6

Case Study: The Chinese Sturgeon

Some of the more serious effects on fish from the construction of dams along the

Yangtze River have been recorded in large migratory species as there are no fish passage facilities incorporated in most of the dams in China. Three species in particular, the

Chinese paddlefish (Psephurus gladius), the Yangtze or Dabry’s sturgeon (Acipenser dabryanus) and the Chinese sturgeon (A. sinensis) have suffered serious population declines and are currently listed as critically endangered on the IUCN Red List (IUCN

2010). For example, the Chinese sturgeon, is an anadromous species that formerly migrated a distance of approximately 3000 km from the East China Sea to their breeding ground in the upper Yangtze River. These fish have been restricted to the lower section of the river since the construction of the Gezhouba Dam in 1981 which blocked their migration route (Xie, 2003). A new, smaller, spawning ground has been established below the dam, however there has been a significant reduction in population size (Qiao et al., 2006; Yang et al., 2006a). Despite an active breeding program, the spawning stock has been diminishing and the age structure and sex ratios have changed, resulting in more older individuals and a higher female to male sex ratio with poor gemmate quality amongst the remaining males (Hui et al., 2006). Molecular analyses have found there is still a great deal of genetic diversity among the remaining sturgeon yet the question remains as to the effect that changing water conditions through additional dam construction along the river will pose to the survival of this species (Zhang et al., 2003;

Xie, 2003).

7

Population Structure in the Yangtze River

Few studies have focused on the natural genetic population structure found at the local scale within river watersheds (eg. Laroche et al., 1999; Carvalho-Costa et al.,

2008; Carlsson and Nilsson, 2000; Perdices et al., 2003). One exception to this focused on the four major carp species in the Yangtze River; Silver Carp (Hypophthalmichthys molitrix), grass carp (Ctenopharyngodon piceus), bighead carp (Aristichthys nobilis), and black carp (Mylopharyngodin piceus) (Lu et al., 1997). All are native to East Asia and together are the major source of freshwater fish production in China and rank amongst the top in the world (Lu et al., 1997). Prior to the construction of the Three Gorges Dam, a study was carried out to investigate the natural population structure and levels of genetic diversity of these fish (Lu et al., 1997). High genetic variability was found among the silver, bighead, and black carp from different nurseries using mitochondrial DNA indicating that they belonged to different genetic stocks but the there was little variation amongst the sampled grass carp populations which the authors suggested may be linked to past demographic differences among the species (Lu et al., 1997). These findings indicate that at least some fish inhabiting the Yangtze basin cannot be managed as a single unit and support the notion that further research focussing on the natural distribution patterns of a variety of species are needed if effective conservation and management strategies are to alleviate the pressures of anthropogenic activities such as the construction of hydroelectric dams on fish biodiversity in the Yangtze (Lu et al., 1997;

Chen et al., 2009).

8

Genetic Impacts of Fragmentation along the Yangtze

There are few studies in the English literature focussing on the genetic effects of fragmentation of local fish population structure within the Yangtze basin, and those that do exist, have focussed on economically important species (Wang et al., 2006; Zhang and

Tan, 2010; Yang et al., 2011). In the largemouth bronze gudgeon (Coreius guichenoti), a benthic potadromodous fish that migrates to the upper Yangtze to lay eggs, molecular analyses have found little population structure along the unfragmented stretch of the main stem of the Yangtze River, but populations blocked by hydroelectric projects on tributaries and below the Gezhouba Dam show significant differentiation (Zhang and

Tan, 2010). In another study, the distribution of genetic variance among populations of the longsnout catfish (Leiocassis longirostris), a benthic semi-migratory species, was examined and little evidence of structure associated with river fragmentation was detected using both mitochondrial and nuclear DNA (Yang et al., 2011). This conflicted with previous work which had reported low levels of genetic variation within this species and significant genetic distance among populations along different stretches of the river

(Wang et al., 2006). This highlights the need for more research examining the impacts of dam construction along the Yangtze River and the importance of studying a variety of species with varying life history characteristics.

Molecular Tools Applicable to Fisheries Biology

During the past few decades a variety of molecular tools have been developed to investigate genetic variation within and among individuals and populations that can 9 answer a variety of questions in fish biology. Genetically based fisheries studies began in the 1950’s using blood type variants and soon after allozyme analyses were introduced to determine stock composition, mainly in commercially important salmonid species

(Ward and Grewe, 1994). Since then a plethora of new genetic marker systems have been employed including: Amplified Fragment Length Polymorphisms (AFLPs), Random

Fragment Length Polymorphisms (RFLP’s), Randomly Amplified Polymorphic DNA

(RAPD’s), mitochondrial DNA, Microsatellites, Single Nucleotide Polymorphisms (SNP’s), and Quantitative Trait Loci (QTLs). Polymerase Chain Reaction (PCR) approaches have made it easy to gather large amounts of information from small and historic samples (Lui and Cordes, 2004). These markers are now used in a wide variety of applications in fisheries research and management including the determination of phylogenetic relationships among species and taxa, genetic mapping, forensic identification, and quantifying levels of variation (Wirgin and Waldman, 1994; Lui and Cordes, 2004; Withler et al., 2004).

Microsatellites

Due to their relative ease of use and high resolution in detecting variation among individuals, populations, and closely related taxa, microsatellites have become the marker of choice in many population studies (eg. Estoup et al., 1998; Seeb et al., 2007;

Aspi et al., 2008; Kruckenhauser et al., 2008). They are co-dominant, inherited in a

Mendelian fashion, typically highly polymorphic, ubiquitous throughout the genome, and are generally thought of as selectively neutral markers (Lui and Cordes, 2004; Chistiakov et al., 2006). Microsatellites consist of repeated units of DNA, with repeats usually 2 to 6 10 base pairs (bp) in length, and various alleles are determined by differences in the number of repeats (Lui and Cordes, 2004). As they are mainly found in non-coding regions of nuclear DNA and display high levels of polymorphism, microsatellites are now widely used to answer many questions in fisheries biology including: parentage analysis

(Bernatchez and Duchesne, 2000; Venturelli et al., 2010), quantifying levels of variation in brood stocks (eg. Li et al., 2004; Withler et al., 2007), delineating population structure

(eg. Spruell et al., 1999; Salgueiro et al., 2003; Wilson et al., 2004; Whiteley et al.,

2006b), assignment of individuals to putative populations (Primmer et al., 2000), species and stock identification (Withler et al., 2004), detecting fisheries induced evolution

(Hutchinson et al., 2003), and detecting genetic signatures of recent environmental fragmentation (Zhang et al., 2006; Reid et al., 2008a and b).

Catfish Evolutionary History

Catfish are members of the Ostariophysi which is the second most diverse superorder within the Teleostei (Saitoh et al., 2003). Included in the Ostariophysi are five orders: the Gonorynchiformes (milkfish), Cypriniformes (minnows, carps, and loaches),

Characiformes (tetras), Siluriformes (catfish), and the Gymnotiformes (electric eels).

Within the Ostariophysi, the Gonorynchiformes have traditionally been thought of as the sister group to the remaining four orders, collectively termed the Otophysi (Saitoh et al.,

2003). This group of fish has historically been considered monophyletic however recent molecular analyses of complete mitochondrial DNA sequences have revealed that the gonorynchiformes are more closely related to the clupeiforms, thereby calling into question the monophyly of the Ostariophysians (Saitoh et al., 2003; Peng et al., 2006a). 11

Several studies have investigated the origin and evolutionary history of the

Otophysan fishes. Of the non-marine vertebrates, they are thought to be one of the most informative groups of animals in determining historical continental relationships as they are mainly restricted to freshwater environments and are not transported by other means such as wind or birds (Briggs, 2005). The prevailing view is that the otophysians have a Pangean origin and only certain lineages expanded in time to become established on the various continents (Saitoh et al., 2003). As such, it is thought that the otophysians originated in South America as this is home to the Gymnotiformes, the most apomorphic of the Characiformes and Siluriformes, and two of the most ancestratral siluriform clades

(Briggs 2005, Sullivan et al., 2006).

The otophysian fish are characterized by having a Weberian apparatus which is an articulated series of transformed anterior vertebrae that extends from the middle ear to the swim bladder (Saitoh et al., 2003). This specialized bony connection permits acute auditory perception which provides a selective advantage and has contributed to the evolutionary success of the otophysian orders (Briggs, 2005).

Siluriformes

Catfish form an incredibly diverse group of fish with 3093 species recognized

(Ferraris, 2007) and another estimated 1750 species yet to be discovered and described

(Sabaj et al., 2004). They are distributed among 477 genera and 36 families and have a global distribution, with fossil catfish discovered even in Antarctica (Diogo, 2005;

Ferraris, 2007). The majority of catfish inhabit freshwater environments although there 12 are some that are found in brackish and marine environments (Ferraris, 2007). In recent years, catfish have become the subject of large scale projects such as the All Catfish

Species Inventory (ACSI), a global initiative devoted to discovering, naming and classifying catfish species as well as determining the higher level phylogenetic relationships within this group (Sabaj et al., 2004; Sullivan et al., 2006).

Catfish Phylogeny

It has been found that Siluriformes form a robustly monophyletic group (Saitoh et al., 2003; Peng et al., 2006a; Sullivan et al., 2006; Wang et al., 2011a). Within the

Siluriformes, the prevailing view, based on morphological data (Mo, 1991) and molecular analysis using the complete sequence of the mitochondrial cytochrome b gene was that

Diplomystidae represented the most basal of the catfish groups (Hardman, 2005).

However, recent molecular analysis using rag 1 and rag 2 sequences found the South

American Locricarioidei to be the sistergroup to the remaining catfishes which were separated into Diplomystidae and Siluroidei (Sullivan et al., 2006). Within the Siluroidei, results of these analyses indicate that there was rapid speciation from a common ancestor followed predominantly by intracontinental diversification (Sulivan et al., 2006).

The Family Bagridae

Bagrid catfish are traditionally recognized as a primitive group of catfish that are widely distributed in Africa and Asia in both fresh and brackish water (Mo, 1991; Ku et al., 2007). In a recent checklist of catfishes it was reported that the family Bagridae comprises 23 genera and 183 species including 11 fossil species (Ferraris, 2007). Within 13 the Yangtze River, the family Bagridae is one of the most diverse families accounting for

5.8% of the total number of species including 3.9% of those that are endemic (Fu et al.,

2003).

As currently defined, the family Bagridae is not monophyletic (Mo, 1991; Harman,

2005). Based on morphological data Mo (1991) proposed that the group be divided into two subfamilies, namely Ritinae and Bagrinae (Mo, 1991); this split has recently been supported by molecular analyses that have found the bagirds to be monophyletic with the exclusion of Rita (Hardman, 2005; Sullivan et al., 2006). Subsequently, molecular analyses using mitochondrial cytochrome b sequences have found that the bagrids in

East Asia and specifically those in China form a monophyletic group (Hardman, 2005;

Peng et al., 2006a). A molecular clock calibration using cytochrome b sequence data for

East Asian bagrids suggests that although bagrids have a fossil record dating back to the

Eocene (Ng, 2003), the majority of the extant species in East Asia resulted from rapid speciation within the past 10 million years (Ku et al., 2007).

Genera of the Family Bagridae

Within the Bagridae, there appears to be polyphyleticism of genera however the species assigned to these genera form a robustly monophyletic group with relatively low genetic divergence (Hardman, 2005; Ku et al., 2007). Ku et al. (2007) investigated the intrafamilial relationships of East Asian bagrid catfish using morphological data combined with sequence data from mitochondrial cytochrome b fragments and proposed that the structure of the maxillary barbels and the presence or absence of serrations on the 14 anterior edge of the pectoral spines seem to be appropriate indicators of the phylogenetic relationships within this group. Thus it has been proposed that the genera of the family Bagridae are in need of revision, and that synonymising the genera

Pelteobagrus with Pseudobagrus and including some Leicassis species would more accurately reflect the evolutionary relationships in this group (Hardman, 2005; Ku et al.,

2007).

The study species: Pelteobagrus vachelli

The darkbarbel catfish (Pelteobagrus vachelli) is a small, benthic, omnivorous fish distributed in China, Vietnam, and Korea (Ferraris, 2007; Fishbase) (Figure 2). Within

China, it is found in the majority of watersheds and throughout the entire length of the

Yangtze with the exception of the head-water basin (Fu et al., 2003; Zeng et al., 2010).

Adult P. vachelli are known to feed at dawn and dusk (Gan et al., 2007) and tend to eat insect larvae, molluscs, and occasionally small shrimp and fish as well as vascular plant debris (Wang et al., 2011b). Physiological studies of P. vachelli have shown that they have a high aerobic capacity, and that they are robust and tolerant to a variety of environmental conditions including temperatures that are well below their optimal range

(Peng et al., 2007; Zhu et al., 2010). Unusual characteristics of this fish include their venomous pectoral spine and high lipid content, ranging from 8.72 to 14.3 % of their body weight and 30.5 to 43.7 % dry mass (Wang and Xie, 2003; Fu et al., 2009).

Pelteobagrus vachelli in the lower part of the Yangtze River are known to spawn once annually from April to July and those in the upper reaches from May to July (Wang 15

Figure 2: Photo of an individual Pelteobagrus vachelli fish collected during the sampling procedure (Photo courtesy of Dr. B. Murray) 16 et al., 2011b; Duan and Sun, 1999b). They breed in areas with submerged vegetation in the main stem of the Yangtze River and its tributaries (Wang et al., 2004; Wang et al.,

2011b). The age of sexual maturity is 2 years and the age of P. vachelli in the reproductive population ranges from 2-7 years in males and 2-4 in females, with the dominant age group 2-3 years old (Duan and Sun, 1999b; Yang, 2006). In this species, males grow faster than females and significant differences between the sexes can be seen after two years of age (Duan and Sun, 1999a).

P. vachelli is an economically important species and together with P. fulvidraco, accounts for more than 10 % of the catch in the Banan region of the upper Yangtze (Duan et al., 2002). Due to their high market price and prominence as a food fish, P. vachelli has become a popular aquaculture species and has been stocked in inland China (Ma et al., 2003; Zheng et al., 2010). Successful artificial reproduction and breeding of P. vachelli was achieved in the late 1990’s (Zeng et al., 2010) and since then there have been several studies focussing on the diet and rearing of these fish in order to increase yields (eg. Wang et al., 2001; Xiao et al., 2002; Long et al., 2005; Wang et al., 2005; Wang et al., 2007; Gan et al., 2008). However, this fish is currently being over exploited in the wild; although it has been increasing in the percent catch (Chen et al., 2009), total numbers have been declining throughout the past decade (Zeng et al., 2010).

Population Structure of Pelteobagrus vachelli

One previous study has investigated the genetic structure of P. vachelli and P. fulvidraco in the Yangtze River using PCR-RFLP analysis of the mitochondrial ND5/6 and 17

D-loop fragments (Wang et al., 2004). Three populations were examined along the length of the river with one population sampled in the upstream portion and the remaining two from the below the Gezhouba Dam in the middle and lower sections of the Yangtze River (Wang et al., 2004). There were higher levels of genetic diversity in P. vachelli compared to P. fulvidraco and no haplotypes were shared among the upstream and mid-downstream populations in either species (Wang et al., 2004). This suggests that the Three Gorges region formed a natural barrier to gene flow in the Yangtze River.

However, a lower degree of differentiation among the upstream and mid-downstream populations was detected in P. vachelli compared to P. fulvidraco, and the authors suggest that this difference may be due to differences in spawning behaviour or a limited population size of P. vachelli in the Yangtze River (Wang et al., 2004).

Effects of Dam Construction on Pelteobagrus vachelli

There is uncertainty as to the effects that future construction of hydroelectric installations without fish passage facilities will have on the genetic diversity of populations of P. vachelli in the Upper Yangtze Basin. There is a possibility that the division of the river may lead to isolated populations with reduced genetic diversity or alternatively, under the changing environmental conditions associated with reservoir formation, this species may proliferate and population sizes expand. A species similar to

P. vachelli in terms of ecology, found in North America and Asia and likewise tolerant to a broad range of environmental conditions, is Brown bullhead (Ameirus nebulosus) (Scott and Crossman, 1973). Molecular analyses of this species using nuclear and allozyme data 18 have found lower levels of genetic diversity in populations from historically degraded sites compared to more pristine locations (Silbiger et al., 2001). This would suggest then that the levels of genetic diversity in P. vachelli along the upper stretch of the Yangtze may be at risk due to both the effects of population isolation and the changing aquatic conditions associated with dam construction.

Purpose of this Study

Scientific studies, particularly those that investigate genetic diversity and population structure of culturally and economically important fish species are needed to inform comprehensive management and protection strategies of fishery resources and biodiversity in the upper Yangtze Basin (Chen et al., 2009). Furthermore, it has been shown that common species are capable of providing valuable information that can be combined with knowledge of species of more concern to guide management and conservation directives (Whiteley et al., 2006a). This project provides a case study of the levels of genetic diversity and population structure of a non-migratory benthic species along the remaining undammed stretch of the upper Yangtze River. This study provides a baseline for future temporal analyses as dams are constructed within this area.

The specific objectives of this experiment were to: 1) develop a set of microsatellite loci specific to P. vachelli, and 2) employ these genetic markers to investigate the levels of genetic variance within and among three populations along the upper Yangtze River. The isolation and optimization of a set of microsatellites loci will be valuable in aquaculture applications for this species, for parentage analysis or 19 maintenance of genetic variation of brood stocks. Additionally, the knowledge gained about the genetic structure of this species will not only add to the understanding of the natural distribution of this species but may also provide information useful in determining future management and conservation strategies for this species and others in the upper Yangtze region.

20

Chapter 2 - Materials and Methods:

The Study Area: The Yangtze River

At over 6300 km in length, the Yangtze River is the third longest river in the world and the longest river in China (Fu et al., 2003). The waters of the Yangtze river originate in the Tibetan plateau at an elevation of 5000 m and the river follows a sinuous path through southeast China until it reaches the East China Sea at Shanghai (Fu et al., 2003;

Peng et al., 2006b; Figure 3). The drainage area of the Yangtze basin is 1 800 000 km2, one fifth the area of China, (Chen et al., 2008) and the river has a mean discharge of

28200 m3/s (Chen et al., 2001). There are more than 3000 tributaries and 4000 lakes contained in the drainage basin which intertwine to form a complex network of aquatic habitats (Fu et al., 2003).

The Upper Yangtze

The Yangtze basin is typically divided in to the upper, middle and the lower reaches, mainly based on geomophological differences (Fu et al., 2003). The Upper

Yangtze, also known as the Jinsha River, is the Western most section of the Yangtze

River. It is 2360 km in length and flows from the Eastern Tibetan plateau to the city of

Yichang in the province of Hubei (Yonghui et al., 2006; Heiner et al., 2011). This section of the river flows through a mountainous terrain with a fall in elevation of 3,280 m and drains an area of over 1 million km2 (Yonghui et al., 2006). 21

The Yangtze River

Chongqing

Y 3

Yibin

2 1

200 km

Figure 3: Map of the sampling sites along the upper Yangtze River, China, between the cities of Yibin and Chongqing. The sampling sites are marked 1 = Yibin, 2 = Luzhou, and 3 = Song Ji. 22

Sample Collection

Samples were collected in December 2008 from three sampling locations along an unfragmented stretch of the Upper Yangtze River, China (Figure 3, Table 1). Fish were purchased in early morning from local fishermen when they returned to shore from their daily catch. Species identity was confirmed, fork length measured and samples of the adipose and tail fins were collected and stored in a dimethyl sulphoxide (DMSO) salt solution (20 % DMSO, 0.25 M EDTA, and NaCl to saturation) (Seutin et al., 1991). GPS readings were taken at each sampling location which were in an approximate range of 2 km from the catch site (Table 1).

Microsatellite Library Development

The protocol for the microsatellite library construction and screening was compiled by Zhengxin Sun using Karen Samis’ lab notes, Glen and Schable’s protocol

(2005), Hamilton et al.’s protocol (1999), Elmer et al. (2006), and the Roche DIG

Application Manual for Filter Hybridization. The resulting protocol for the development and isolation of microsatellite loci for P. vachelli was as follows.

Isolation of Genomic DNA

Total genomic DNA (gDNA) taken from one fish sample was extracted following the protocol for the DNeasy Blood and Tissue Kit (Qiagen , Valencia, USA; cat # 69506).

Approximately 20 mg of tissue was washed twice with 500 μl of Phosphate Buffered

Saline (PBS) to remove any DMSO prior to extraction. The quality of the genomic DNA 23

Table 1: Sampling sites, GPS coordinates, the number of individual Pelteobagrus vachelli sampled, and the range of fork length (FL) in cm from the three locations along the Upper Yangtze River, China.

Population Site GPS Coordinates Number of fish FL (cm) 1 Yibin 28.77708 104.62469 20 7.2 – 11.1 2 Luzhou 28.88262 105.45469 20 8.0 – 16.0 3 Song Ji 29.06060 105.93778 21 8.1 – 13.4

24 was assessed visually by electrophoresing 2 μl gDNA combined with 0.5 μl of 6 x bromophenol blue loading dye on a 1 % agarose gel in 1 x TAE (Tris-Acetate-EDTA) pre- stained with ethidium bromide and then placing the gel on a UV transilluminator. gDNA was quantified using a Biochrom Ultrospec 2100 Pro (Biochrom Ltd., Cambridge, UK; cat

# 802112-21).

Restriction Enzyme Digestion of Genomic DNA

Genomic DNA was then digested with various restriction enzymes to find a combination that produced fragments of 200 – 1000 base pairs (bp), with an average size of approximately 500 bp. Both single and double digestions were carried out in search of this combination. Conditions for the single digests were: 300 ng of gDNA, 3 μl of 10 x

Fermentas Buffer Tango, 2 U of restriction enzymes (Alu I, Bsu RI, or Rsa I respectively)

(Fermentas) and H2O to a final volume of 30 μl. The double digestions consisted of 300

ηg of gDNA, 3 μl of 10 x Fermenteas Buffer Tango, 2 U of each restriction enzyme in combinations of Alu I/Bsu RI, Alu I/Rsa I, Alu I/SSP I, Bsu RI/Rsa I, Bsu I/SSP I, or Rsa I/ SSP

I, and H2O to make a total reaction volume of 30 μl. All reactions were incubated overnight at 37 ˚C.

To visualize the fragment lengths of the restriction digests, 3 μl of 6 x bromophenol blue loading dye were added to 15 μl of the reaction volume and 18 μl were run on 2 % agarose gel in 1 x TAE pre-stained with ethidium bromide and then placed on a UV transilluminator. The enzyme that best produced fragments between

200 and 1000 bp was Bsu RI and it was selected for further use in the development of the 25 microsatellite library. The digestion was repeated to produce a larger quantity of product by mixing 5 μg of gDNA, 12 μl of 10 x Fermentas Buffer Tango, 20 U of Bsu RI and

66 μl H2O for a total reaction volume of 120 μl and incubating overnight at 37 ˚C.

The gDNA fragments were dephosphorylated to prevent fragments from ligating together by adding 0.5 μl of undiluted Calf Intestine Alkaline Phosphotase (CIAP) (1 U/μl)

(Invitrogen, Carlsbad, USA; cat # 18009-027) to the 120 μl solution and incubating at 50

˚C for 30 minutes. Following this, a second 0.5 μl of undiluted CIAP was added and the solution incubated for another 30 minutes at 50 ˚C. 5 μl of this solution was removed to use as a non-purified dephosphorylated control in future gel electrophoresis.

The DNA was then purified following the protocol for the PureLinkTM Quick Gel

Extraction Kit (Invitrogen, cat # K2100-12). To ensure that the DNA had been successfully recovered, 5 μl of enzyme digested and dephosphorylated DNA and 5 μl of digested, dephosphorylated, and purified DNA were each mixed with 1 μl of 6 x bromophenol blue loading dye and run on 1 % agarose gel pre-stained with ethidium bromide in 1 x TAE.

The gel was subsequently visualized using a UV transilluminator.

Preparation of Double Stranded SNX Linkers

To make the double stranded SNX linkers, 50 μl of 100 μM SNX-F (sequence 5’-

CTAAGGCCTTGC-3’) and 50 μl of 100 μM SNX-R (sequence 5’-pAGCAAGGCCTTAGAAAA-

3’) (Invitrogen) were incubated at 94 ˚C for 2 minutes, 65 ˚C for 10 minutes, then at room temperature for one hour. The double stranded SNX linker (50 μM dsSNX) was stored at

-20 ˚C 26

Ligation of DNA Fragments to SNX Linkers and Verification

The double stranded SNX linkers were added to the digested gDNA fragments by mixing 66 μl gDNA, 3 μl dsSNX linker (50 μM), 8 μl 10 x ligase buffer, 1 μl Xmn I (20 U/ μl),

2 μl T4 ligase (5 U/μl) and H2O to make a final volume of 80 μl. This solution was then incubated overnight at 16 ˚C. To ensure that this reaction was successful, 3 separate PCR reactions were carried out by mixing 1.5 μl 10 x KCL Buffer, 0.2 μl 10 mM dNTP mix, 1 μl

10 μM SNX-F, 10 μl H2O, 0.1 μl Taq DNA polymerase (5 U/μl) (Qiagen; cat # 201203), and

1 μl of either the ligation product, the dephosphorylated and purified gDNA fragments, or water (as a control) respectively. Thermocycling conditions were as follows: 95 ˚C for

2 minutes, 30 x (95 ˚C for 20 seconds, 60 ˚C for 20 seconds, and 72 ˚C for 90 seconds), and 72 ˚C for 5 minutes. Four μl of each of these 3 reactions were then mixed with 1 μl of 6 x bromophenol blue loading dye and electrophoresed on 1 % agarose gel in 1 x TBE

(Tris-Borate-EDTA). The gel was then stained with ethidium bromide and placed on a UV transilluminator for visualization of the PCR products. The successful ligation of the dsSNX linkers to the gDNA fragments was confirmed as this was the only reaction to produce PCR products and resulted in a smear from approximately 200 - 1000 bp.

Enrichment of the Microsatellite gDNA-Linker with Biotin-oligos

The digested DNA ligated to the SNX linkers was then hybridized with dinucleotide repeat biotin-oligos by mixing 50 µl 2 X Hybridization buffer (12 X SSC, 0.2%

SDS), 2 µl 10 µM 3’-biotin-(AC)13, 2 µl 10 µM 3’-biotin-(AG)13, 46 µl DNA –linker and then placing the solution in a thermocycler under the conditions: 95 °C for 10 minutes, 70 °C 27 for 1 hour, 65 °C for 1 hour, 60 °C for 1 hour, 55 °C for 1 hour, 50 °C for 1 hour and then held overnight at 10 °C.

Dynabeads were prepared by placing 80 µl of Dynabeads® M-280 Streptavidin (10 mg/ml) (Invitrogen, cat# 112-05D) into a 1.5 ml microcentrifuge tube and then capturing the beads with a Magnetic Particle Collecting unit (MPC) and removing the supernatant.

Beads were then washed with 250 µl of TE and vortexed. Beads were then recaptured using the MPC and the supernatant discarded. This step was repeated with a second 250

µl TE and then twice with 250 µl 1 x Hybridization buffer (6 x SSC, 0.1 % SDS). When using the MPC, beads were left for a minimum of 1 minute before removing the supernatant in order to minimize loss of beads. Beads were then resuspended in 100 µl

1 x Hybridization buffer.

To collect the microsatellite enriched DNA, the biotin-oligo and DNA–SNX linker hybridization reaction mixture was then added to the resuspended Dynabeads and incubated on a rotor at slow speed for 1 hour at room temperature. The magnetic beads were captured using the MPC and the supernatant removed and stored at -20 °C. The beads were then washed twice with 400 µl 2 x SSC, 0.1 % SDS at room temperature.

They were left to incubate for five minutes at room temperature during each wash then the supernatant was removed using the MPC to secure the beads, as in all subsequent washes, and the supernatant stored at -20 °C. The beads were then washed twice with

400 µl of 1 x SSC, 0.1 % SDS at 45 °C. Beads in solution were incubated for 5 minutes and the supernatant removed and stored at -20 °C. Then the beads were washed with 400 µl of 1 x SSC, 0.1 % SDS preheated to 55 °C and left to sit at room temperature for 5 28 minutes. The supernatant was removed and stored at -20 °C. The beads were resuspended in 200 µl of TLE (10 mM Tris, pH 8.0, 0.1 mM EDTA) and incubated at 95 °C for 5 minutes. They were then promptly put on the magnet and left to sit for 1 minute.

The supernatant was transferred to a fresh tube, 22 µl of 3 M Sodium Acetate was added and the solution mixed by tapping. 444 µl of 100% Ethanol was then added and the solution inverted to mix and placed at -20 °C for 1 hour. The DNA was then centrifuged at 14000 g for 15 minutes and the supernatant removed. The DNA was then washed with cold 70 % Ethanol and then spun for 5 minutes at 14000 g in the centrifuge. The supernatant was aspirated and the DNA was air dried in a fume hood for 30 minutes.

The DNA pellet was then resuspended in 25 µl of TLE.

PCR to Test the Results of the Enrichment

To confirm the enrichment was successful, a PCR was done with four different reactions by combining: 2.5 µl of 10 x PCR buffer, 1 µl of 50 ηM MgCl2, 0.375 µl of 10 mM dNTP mix, 1.3 µl of 10 µM SNX-F, 2.5 µl of 100 X BSA (10 mg/ml), 15.125 µl H2O, 0.2 µl of

Taq DNA polymerase (5 U/µl), and 2 µl of either the SNX ligation mixture, the enriched gDNA-linker, the dephospholyrated and purified DNA, or H2O as a control.

Thermocycling conditions were as follows: 95 °C for 2 minutes, 30 x (95 °C for 20 seconds, 60 °C for 20 seconds, 72 °C for 90 seconds), 72 °C for 30 minutes and then held at 4°C. 4 µl were then taken from each reaction mixture and mixed with 1 µl 6 x bromophenol blue loading dye and run on 1% agarose gel in 1 x TBE and post stained with ethidium bromide. The gel was then visualized using a UV transilluminator to ensure the enrichment was successful. 29

Dot Blot of the Enriched Microsatellite gDNA-Linker

Oligonucleotide probes {(AC)13 and (AG)13} were labelled with digoxigenin (DIG) using the DIG Oligonucleotide 3’ – End Labelling Kit, 2nd generation (Roche, Indianapolis,

USA; cat # 03353575910) and subsequently quantified through dot blot analysis following the procedure outlined in the DIG Application Manual for Filter Hybridization

(Roche). Additional materials used included a nylon membrane (Roche; cat #

11699083001), the DIG Wash and Block Buffer set (Roche; cat # 11 585 762 001) and a

SpectrolinkerTM XL-1000 UV Crosslinker (Spectronics Corp.).

To ensure that the olginucleotide enrichment was successful, dot blot hybridizations were set up to compare the relative intensities. Four different DNA samples were used including: the gDNA fragments, the gDNA-linker ligation mix, PCR product of the ligation mix, and the PCR product of the enriched microsatellite gDNA- linker. 1:2, 1:5, 1:10 dilutions of the two PCR products were made using 1 µl of the original solutions. 1.5 µl of formamide (Roche; cat # 11814320001) was then added to all six of these samples, including the dilutions, and 1.5 µl of H2O was used as a control.

These mixtures were then heated at 95 °C for 5 minutes in a thermocycler and then put on ice. A pencil was used to draw grids on a nylon membrane and 3 µl of denatured DNA and the control were placed in the grids and air dried. DNA was then fixed to the membrane using a Spectrolinker XL-1000 UV Crosslinker.

The membrane was pre-hybridized with 10 ml of pre-heated DIG Easy Hyb

(Roche; cat # 11 603 558 001) at 50 °C for 30 minutes in a hybridization bottle in a 30

ProBlotTM L6 hybridization oven (Labnet International, Inc.). The probe/hybridization solution was made by diluting 10 ρmol of DIG-labeled oligo probe in 10 ml pre-heated

DIG Easy Hyb. The prehybridization solution was removed and the probe/hybridization solution was immediately added to the membrane and incubated at 50 °C for 1 - 6 hours in the hybridization oven. The probe/hybridization solution was then removed and the membrane washed twice with 15 ml 2 x SSC, 0.1 % SDS for 5 minutes each time in a rotisserie oven at room temperature. The membrane was then washed twice with 15 ml

1 x SSC, 0.1 % SDS, each time placing in the hybridization oven at 50 °C for 15 minutes.

This was followed by a wash with 20 ml of Washing buffer [0.1 M maleic acid, 0.15 M

NaCl, pH 7.5, 0.3 % (v/v) Tween-20] at room temperature, shaking it in an orbital shaker for 2 minutes. The Washing buffer was then removed and the membrane was incubated for 30 minutes in 10 ml freshly diluted 1 x Blocking solution. The Blocking solution was then discarded and the membrane placed in a fresh container. The anti-DIG-AP antibody was diluted in 10 ml of Blocking solution and then the membrane was incubated in this antibody solution for 30 minutes at room temperature. The antibody solution was then removed and the membrane washed twice with 10 ml of Washing buffer at room temperature for 15 minutes each time, discarding the Washing buffer at the end of each wash. 10 ml of Detection buffer [0.1 M Tris-HCL, 0.1 M NaCl, pH 9.5] was added to the membrane and it was incubated for 5 minutes at room temperature. The membrane was then removed and placed on a transparency DNA side up. Several drops of CDP-Star

(Sigma, Saint Louis, USA; cat # C 0712) were applied to the membrane until the membrane was completely covered. A second transparency was placed on top, making 31 sure that there were no air bubbles, and the membrane was allowed to sit for 5 minutes at room temperature. Excess liquid was squeezed out from between the transparencies and they were placed inside a cassette and exposed to X-ray film for 5 minutes in a dark room and then the film developed. The darkest band appeared where the PCR product of the enriched microsatellite gDNA-linker sample was placed indicating that the enrichment was successful.

Ligation of Enriched Microsatellite gDNA-Linker to TOPO TA Vector

The enriched microsatellite gDNA-linker was then ligated to pCR2.1®-TOPO® TA vectors following the protocol for the TOPO® TA Cloning Kit (Invirogen, cat # K4500-01).

4 µl of the PCR product from the enriched microsatellite gDNA-linker was mixed with 1 µl of Salt Solution and 1 µl of TOPO® TA vector to obtain a final volume of 6 µl. The solution was then incubated at room temperature for 30 minutes and then placed on ice for 30 minutes.

To do the transformation, S.O.C. medium was warmed to room temperature and two 15 cm diameter LB agar plates containing 60 mg/L of ampicillin and 24 mg/L X-GAL were incubated at 37 °C until needed. One vial of One Shot® TOP 10 chemically competent cells was thawed on ice and then 2 µl of the ligation mix was added to the competent cells and the cells gently mixed by tapping. Cells were then incubated on ice for 30 minutes followed by a heat shock at 42 °C for 30 seconds without shaking. The tube was then immediately transferred to ice and 250 µl of the room temperature S.O.C. medium was added to the tube. Cells were then shaken horizontally at 37 °C for 1 hour 32 at 200 rpm. Two 15 cm diameter plates were then plated with 80 µl each of the cell solution. Plates were left bottom up at 37 °C overnight.

Identification of Positive Colonies by Hybridization with Probes

Four 15 cm diameter LB/Amp agar plates with grids on the bottom were made and clearly labelled. Sterile pipette tips were used to pick single white colonies from the agar plates and then place smears on duplicate plates in the same grid location. These plates were then incubated at 37 °C overnight.

The colonies then needed to be lifted and prepared to bind the DNA to membranes. One of each of the duplicate plates was cooled for approximately 30 minutes at 4 °C. Two nylon membranes (Roche, cat# 11699083001) were labeled on the back (reverse side) with pencil and then carefully placed on the agar plates to avoid air bubbles and left to sit for 5 minutes. A sterile syringe was used to place asymmetric markings on the membranes and plates in order to mark the orientation of the membrane to the plate. 2 ml of Denaturation Solution (0.5 M NaOH, 1.5 M NaCl) was placed on either end of the transparency. The membranes were then carefully removed from the plates and the bottoms dried on a sheet of Whatman 3MM paper. The membranes were then placed on the Denaturation Solution on the transparency and left to sit for 15 minutes. The membranes were then dried again on a new sheet of

Whatman 3MM paper and then placed on a new transparency each in 2 ml of

Neutralization Solution (1.5 M NaCl, 1.0 M Tris-HCL, pH 7.4) and left to sit for 15 minutes.

When the neutralization process was complete, the nylon membranes were removed 33 and placed on a new sheet of Whatman 3MM paper to dry and then placed on a new transparency each in 2 ml of 2 x SSC solution. The membranes were left to sit for 15 minutes and then air dried briefly before cross-linking the DNA with a Spectrolinker XL-

1000 UV Crosslinker.

To remove the cell debris, the membranes were wet with ddH2O and placed in hybridization bottles. Fifteen ml of digestion buffer (1 x TE with 0.5 % SDS) and 75 µl of

Proteinase K (10 mg/ml) were added to each bottle and they were placed in the hybridization oven at 50 °C for 1 hour. The digestion solution was then removed and 10 ml of DIG Easy-Hyb preheated to 50 °C was added to each bottle. The membranes were then put in the rotisserie for 1 hour at 50 °C. The pre-hybridization solution was then removed and 10 ml of probe/hybridization solution pre-heated to 50 °C was added and the bottles placed in the hybridization oven overnight at 50 °C.

The probe/hybridization solution was removed and the membranes were then washed three times with 15 ml of 2 x SSC with 0.1 % SDS, each time placing the bottle in the hybridization oven at room temperature for 15 minutes. The membranes were then washed twice with 1 x SSC with 0.1 % SDS preheated to 50 °C by placing them in the hybridization oven at 50 °C for 15 minutes each time. This was followed by a wash with

20 ml of 1 x Washing buffer for 5 minutes in the hybridization oven at room temperature in order to remove the SSC. The Washing buffer was then removed and the membranes were blocked by adding 15 ml of freshly made 1 x Blocking solution to each bottle and placing them in the hybridization oven at room temperature for 45 minutes. The

Blocking solution was then removed and 15 ml of Antibody solution (1:10000 Anti-DIG- 34

AP antibody (Roche; cat # 1093274) in 1 x Block solution) were added to each bottle and the membranes incubated for 30 minutes in the hybridization oven at room temperature. The antibody solution was then removed.

Membranes were then washed three times with 15 ml of 1 x Washing buffer for

10 minutes each time by placing them in the hybridization oven at room temperature.

To equilibrate the membranes, they were then washed twice with 15 ml of 1 x Detection buffer (0.1 M Tris-HCL, 0.1 M NaCl, pH 9.5) for 5 minutes in the hybridization oven at room temperature, disposing of the solution after each wash.

Membranes were then removed from the hybridization bottles, dried on

Whatman 3MM paper, and then placed on a transparency DNA side up. The membranes were then each covered with 2 ml of CDP-Star and left to sit for 5 minutes. A second transparency was then placed overtop, excess liquid squeezed out, and the orientation marked by placing a glow in the dark sticker on the top of the transparency. The whole assembly was then exposed to X-ray film for 5 minutes in a dark room. Positive colonies were identified as those producing a dark band on the X-ray paper.

Growth of Positive Colonies

Positive colonies were removed from the replicate plates corresponding to those identified on the X-ray film using sterile toothpicks and were cultured overnight in 7 ml

LB Broth containing ampicillin (60 mg/L) at 37 °C on a rocking bed. 1.5 ml of each bacterial culture was transferred to a 2 ml cryovial tube, mixed with 100 µl of glycerol, and then snap frozen in liquid nitrogen for storage as permanent cultures at -80 °C. 35

Plasmid Isolation and Sequencing of Inserts

DNA was isolated from 96 of the positive cultures following the protocol for the

PurelinkTM Quick Plasmid Miniprep Kit (Invitrogen, cat# K2100-11) and the concentration of DNA was quantified using a Biochrom Ultrospec 2100 Pro. Samples were then diluted with ddH2O to a concentration of 50 ηg/µl and sent to Plateforme de Séquencage et de

Génotypage des Génomes for sequencing (Québec, Canada).

Primer Design

Sequences were visualized using Microsoft Excel and primers were designed using

Primer 3 (Rozen and Skaletsky, 2000). The following changes to the Primer 3 parameters were made: product size range 100-300; primer size max 22; primer GC % = 55; GC clamp

= 1). A first set of ten primer pairs was ordered from Invitrogen (Invitrogen) and a second set of nine primer pairs was ordered from Eurofins MWG Operon (Huntsville,

USA).

Primer Optimization

gDNA samples were extracted from six individual fish according to the protocol for the PurelinkTM Genomic DNA Mini Kit (Invitrogen; cat # K1820-02). Approximately 20 mg of tissue was washed twice with 500 µl of phosphate buffered saline (PBS) to remove any DMSO prior to extraction. DNA was quantified using a Biochrom Ultrospec 2100 Pro.

Quality was ensured by running 15 µl of each sample combined with 3 µl of 6 x bromophenol blue loading dye on 1% agarose gel pre-stained with ethidium bromide and then placing the gel on a UV transilluminator. 36

Primers were diluted to 100 µM with TE Buffer and then to 10 µM with sterile water. PCR reactions were then conducted to test if primers produced products by combining: 1 µl of gDNA, 1 µl of 10 x QIAGEN PCR buffer, 0.1 µl of dNTP’s [10 mM], 0.1

µl [10 mM] forward primer, 0.1 µl [10 mM] reverse primer, 0.05 µl Taq DNA polymerase

(5 U/µl), and 7.65 µl sterile water for a total reaction volume of 10 µl. Conditions for the

PCR included a range of annealing temperature (Tm) from 55-63 °C for the first set of ten primers and from 52.7 to 63.2 °C for the second set of nine primers. Thermocyling parameters were set as follows: 94 °C for 3 minutes followed by 35 cycles of (94 °C for 15 seconds, Tm for 30 seconds, 72 °C for 30 seconds) and a final extension of 72 °C for 7 minutes and then held at 4 °C. 10 µl of OG loading dye were then added to each 10 µl reaction volume and 10 µl were electrophoresed on 2 % agarose gel in 1 x TBE and then post stained with ethidium bromide. Gels were placed on a UV transilluminator and examined for the presence and quality of product. Primers that produced viable PCR amplicons were selected for further testing and the optimal annealing temperature for these loci were selected as the Tm that produced the clearest and brightest bands.

The primers that produced PCR amplicons were then further tested on the LI-COR

4200 IR2 platform (LI-COR 4200 Global IR2 SystemTM, LI-COR Inc., Lincoln, USA) to identify those that produced polymorphic products. Forward primers were synthesized with the

LI-COR M13 universal sequence (Sequence 5’-CACGACGTTGTAAAACGAC-3’) at the 5’ end of the locus-specific microsatellite primer sequences to be used with the fluorescent

M13 Forward (M13F) universal primer (LI-COR Inc., M13F-700IRD; cat# 4200-20). PCR reactions were carried out by combining 1 µl of gDNA, 1 µl of 10 x QIAGEN PCR buffer, 37

0.1 µl of dNTP’s [10 mM], 0.1 µl of [10 mM] forward primer, 0.1 µl of [10 mM] reverse primer, 0.05 µl of M13F [1 ρmol/µl], 0.05 µl of Taq DNA polymerase (5 U/µl), and 7.6 µl of sterile H2O for a total reaction volume of 10 µl. Thermocycling conditions were identical to those used during the first stage of primer optimization with the annealing temperature as determined through the optimization process (Table 2).

Multiplex PCR Optimization

In order to increase the efficiency of PCR reactions and reduce amplification error, selected loci were combined into multiplex PCR reactions capable of amplifying two loci at once. All multiplex reactions were optimized using the LI-COR 4200 IR2 platform. PCR reactions were optimized using the QIAGEN Multiplex PCR Kit (cat #

206143) by combining: 1µl of gDNA, 5µl of 2 x QIAGEN Multiplex PCR Master MixTM, 1µl of 10 x Primer cocktail (containing 1.7 µM of each forward and reverse primer per unique locus), 1µl of M13F-IRD700 primer, and 2 µl of sterile H2O for a total reaction volume of

10µl. Thermocycling conditions for all multiplex reactions were as follows: 95 °C for 15 minutes, 35 x (94 °C for 1 minute, 55 °C for 1.5 minutes, 72 °C for 1.5 minutes) and a final extension of 72 °C for 15 minutes and then held at 4 °C.

Genomic DNA Isolation of the Sample Set

Genomic DNA of the sample set was extracted using the PurelinkTM Genomic DNA

Mini Kit (Invitrogen; cat # K1820-02) according the manufacturer’s protocol. Quality was ensured through electrophoresing 10 µl of genomic DNA combined with 2 µl of 6 x bromophenol blue loading dye in 1% agarose gel in 1 x TAE pre-stained with ethidium 38 bromide and then placing the gel on a UV transilluminator. Samples were quantified using a Biochrom Ultrospec 2100 Pro.

PCR Amplification of Loci

PCR reactions were then carried out using gDNA from the entire sample set according to the optimization conditions, either as single loci or combined in multiplexes

(Table 2) according to the reaction protocols outlined above and run on the LI-COR 4200

IR2 platform. Template DNA ranged in concentration from approximately 5 to 30 ng/µl.

The software program Gene ImagIRTM Version 4.05 (Scanalytics Inc.) was used to analyze the resulting electrophoretograms and PCR amplicons were sized using commercially available 50 - 350 bp sizing standards (LI-COR cat # 829-06157 and 829-05343).

Amplification Error Estimation

To estimate genotyping error, two methods were employed. For the first, two different researchers, using two different primer cocktails, amplified a subset of samples using the same gDNA. The second method consisted of the same researcher amplifying a subsample of gDNA using two separate primer cocktails. In both cases, the same researcher scored and compared the resulting genotypes. To ensure that the relative sizes were consistent among electrophoretograms, either subsets of the samples were run a second time/in replicate on single gel loads or for some loci, the entire sample set was run a second time in an order corresponding to the initial size determination. 39

Table 2: Microsatellite loci isolated in this experiment with the locus specific forward and reverse primer sequences, the optimized annealing temperature (Tm), whether or not it was amplified as part of a multiplex, and the expected allele size. MgCl2 concentration was 1.5 mM in all cases.

Locus Motif Forward Primer Reverse Primer Tm Multiplex e Expected Size

Pvi 004 (AC)21 CGAGCTCGAACTAAACTACACC ATGTGTGCTTCTGTCTGTGC 58°C Yes/No 151

Pvi 026 (AC)26 ATGAAATTGCGTCCAAAAGG GCTTCCATCGCTGAAGAAAC 58°C No 245

Pvi 038 (AC)24 ACACAGCTGTAGCACCATGC TTCTGGAGCAAATGAAACACAC 55°C No 183

Pvi 045 (GT)27 GCATCTGCAGGTGAAATGTG GTGAGAGCCTGGAAACTTGC 58°C No 215

Pvi 052 (TG)29 CATGGCTGGAGGAAACACAC GAGTTGGACACATGGAGCTG 55°C Yes/No 238

Pvi 055 (TG)23 ACTGTGGTATGCGCTGTCTG ACGATGACAGCCTGGGTTAC 58°C Yes

Pvi 066 (AC)26 ACATATCCAAGCGGTTGAGG AGCAGTTTTTCCCCCAAATG 58°C Yes

Pvi 114 (TG)33 CTGGAAATGTGGGGAAGATG TGCATCACTGGCTGGTTTAG 63°C No 206

For the loci marked yes/no, a multiplex reaction was used in 2009 and not in 2011

40

Data Analysis

Genotypes were examined for the possibility of genotyping errors due to large allele dropout, stuttering, and the presence of null alleles using the program Micro-

Checker version 2.2.3, with the number of randomizations set to 10 000 (Van Oosterhout et al., 2004).

Standard genetic diversity indices including the allele frequencies, the number of alleles per locus (A), the observed (HO) and expected (HE) heterozygosities (Nei, 1987), and the polymorphic information content (PIC) were computed using the Microsoft Excel add-in utility Microsatellite Toolkit (Park, 2001). Allelic richness, which is the mean number of alleles per locus standardized to the smallest sample size (n = 20), was calculated with the program Fstat version 2.9.3.2 (Goudet, 2002).

Linkage disequilibrium between pairs of loci across the entire sample set was tested using Genepop version 4.1.1 (Rousset, 2008). Tests for significant associations between pairs of loci within populations were calculated using a likelihood ratio test of linkage disequilibrium for genotypic data, gametic phase unknown (Slatkin and Excoffier,

1996). These tests were computed using Arlequin version 3.5.1.3 (Excoffier et al., 2005) with 16 000 permutations and the expectation-maximum set to 3 (Guo and Thompson,

1992).

Exact tests for deviations from Hardy-Weinberg equilibrium (HWE) for each locus within populations were computed using Arlequin version 3.5.1.3. The probability of significant departure from HWE was calculated using a Markov chain procedure with 1 x 41

106 steps and 100 000 dememorization steps (Guo and Thompson, 1992). Estimates of heterozygote deficiency or excess were performed using exact tests (Guo and Thompson,

1992; Goudet et al., 1996) implemented in Genepop version 4.1.1 using the default settings. To minimize the risk of Type I errors occurring with multiple tests, the significance level for p-values (p < 0.05) was adjusted through sequential Bonferroni corrections for all simultaneous statistical inference techniques (Rice, 1989).

Due to indications of significant deviation from HWE in each of the three populations and a relatively high level of linkage disequilibrium associated with locus Pvi

114, this locus was excluded from further analyses exploring the population structure among the three sampling locations.

Analysis of molecular variance (AMOVA) (Excoffier et al., 1992) to partition genetic diversity within and among populations was carried out using Arlequin version

3.5.1.3 (10 000 permutations). Pairwise FST values (Weir and Cockerham 1984; Excoffier et al., 1992) between population pairs and the corresponding p-values were also computed using Arlequin version 3.5.1.3 with the number of permutations set to 10 000.

FST was used as opposed to RST because simulations have shown that FST is more conservative when population sizes are small and relatively few microsatellite markers are used (Gaggiotti et al., 1999) and is more effective in detecting isolation-by-distance

(Estoup et al., 1998).

To further investigate the presence of genetic structure among the three populations, an exact test of differentiation based on sample genotype frequencies 42

(Raymond and Rousset, 1995; Goudet et al., 1996) was conducted using Arlequin version

3.5.1.3 with a Markov chain length of 1 000 000 and the number of dememorization steps set to 100 000 and tests of genic differentiation were conducted using Genepop version 4.1.1.

To visualize the relationship in genetic distance among individuals and populations, principle co-ordinates analysis of pair-wise individual genetic distances based on differences in number of alleles (covariance matrix with data standardization)

(Nei, 1987) was done using GenAlEx (Peakall and Smouse, 2006) and the first three principle co-ordinates plotted in two dimensions using Microsoft Excel and three dimensions using SigmaPlot version 11.2.0.

43

Chapter 3– Results:

Microsatellite Library Construction and Optimization

The microsatellite library development resulted in 237 (of 312) colonies containing oligonucleotide enriched DNA fragments which were cultured and stored at -

80 ˚C as permanent cultures. Plasmid DNA was isolated from 96 of the positive colonies and sequences were obtained for the inserted fragments. These sequences were used to design 19 primer pairs specific to the flanking regions of the oligonucleotide repeats, all of which corresponded to dinucleotide repeats between 24 and 66 base pairs in length.

Of the 19 designed primer sets, PCR amplicons from 11 loci were visualized on agarose gels. Of these, eight amplified polymorphic loci and were optimized to produce reliable products on the LI-COR 4200 IR2 platform (Table 2), while the remaining three primer sets amplified monomorphic loci.

Genotyping Error Estimates

Genotyping analysis using duplicate locus specific primer cocktail mixes produced identical genotypes in all samples that amplified in both instances for all loci. Among all of the electrophoretograms, two instances of a different genotype at one locus for particular samples were observed; however as multiple runs of the same samples were done in various orders on the gels, the correct genotype was evident and the differences were found to be due to human error.

44

Microsatellite Genotyping

All 8 optimized loci were amplified and genotyped across the 61 samples collected from 3 populations along the upper Yangtze River (Tables 1 and 2).

Assessment of Genotyping Technical Errors

Analyses of the data using Micro-Checker provided no indication of large allele dropout or error due to stuttering. However, the presence of null alleles for two of the loci, each within separate populations (Pvi 052 within the Songji population and Pvi 114 within the Luzhou population) was suggested by the presence of homozygote excess ranging across most of the allele size classes. Additionally, more than 50% of the alleles for locus Pvi 114 in the Yibin population were of the same size (178 bp) and thus binomial analysis (Weir, 1996) was not possible.

Genetic Variability of the Optimized Alleles

There was a high level of genetic variability across the entire sample set with a total of 171 alleles amongst the 8 loci. The number of alleles per locus across the three populations ranged from 14 to 29 with an average of 21.4 alleles per locus (Table 3).

Observed and expected heterozygosities per locus ranged from 0.66 – 0.93 and 0.84 –

0.95 with averages of 0.84 and 0.90 respectively (Table 3). The PIC values were high for each locus with a range of 0.819 – 0.936 when the data from the 61 samples were considered together (Table 3), indicating that the optimized primers are detecting large amounts of variation within and amongst the sampled populations. 45

Table 3: Summary statistics per locus and averages across loci including the number of alleles, standardized allelic richness (AR), range in allele size (in base pairs), expected heterozygosity (He), observed heterozygosity (Ho), and the polymorphic information content of each locus (PIC) and the averages across loci for the 61 samples included in the analyses.

Locus # of Alleles Range He Ho PIC Pvi 004 25 162-240 0.918 0.902 0.905 Pvi 026 21 235-283 0.935 0.885 0.923 Pvi 038 29 181-247 0.947 0.934 0.936 Pvi 045 14 212-254 0.852 0.885 0.829 Pvi 052 23 232-282 0.926 0.803 0.913 Pvi 055 15 101-129 0.850 0.754 0.824 Pvi 066 23 322-370 0.937 0.902 0.925 Pvi 114 21 178-258 0.836 0.656 0.819 Average 21.4 - 0.900 0.840 -

46

Tests of Linkage Disequilibrium across Populations

Tests of disequilibrium between locus pairs for the entire sample set indicated two locus combinations were in linkage disequilibrium at an alpha level of 0.05. These were loci Pvi 038 and Pvi 114 (p = 0.04583) and Pvi 052 and Pvi 114 (p = 0.03512). After adjustment of the significance level through sequential Bonferroni correction, neither of these associations was significant at the table wide alpha level (k = 28).

Private Alleles

Among the three sampled populations, private alleles were detected for each locus (Table 4). Over all populations, the number of private alleles found per locus ranged from 19 % – 43 % of those detected among all individuals (Table 4). Unique alleles were detected in each of the populations, with 30.4 % of the total number of alleles observed being private to individual sampling sites and no allele was characteristic of a particular population (Table 4, Appendix 1).

Genetic Variation within Populations

Within each of the three populations, levels of genetic variation were high. The number of alleles per locus within the populations ranged from 10 – 20, 10 – 21, and 10 –

19 with averages of 15.5, 16.0, and 13.6 in the Song Ji, Luzhou, and Yibin sampling sites respectively (Table 5). As the same number of individuals was sampled in each of the populations, with the exception of Song Ji which had one additional individual, the allelic 47

Table 4: Number of private alleles (P), total number of alleles (A) and the percentage (%) of the total identified for each locus within the Song Ji, Luzhou, and Yibin populations and over all populations.

Locus Song Ji Luzhou Yibin Over all populations P A % P A % P A % P A % Pvi 004 4 17 23.5 1 17 5.88 4 17 23.5 9 25 36 Pvi 026 2 15 13.3 4 17 23.5 0 12 0 6 21 28.6 Pvi 038 2 20 10 4 21 19.0 3 19 15.8 9 29 31.0 Pvi 045 2 10 20 2 10 20 2 10 20 6 14 42.9 Pvi 052 5 18 27.8 1 15 6.67 3 13 23.1 9 23 39.1 Pvi 055 0 10 0 1 14 7.14 1 11 9.09 2 15 13.3 Pvi 066 3 18 16.7 3 19 15.79 1 15 6.67 7 23 30.4 Pvi 114 2 16 12.5 2 15 13.3 0 12 0 4 21 19.0 Total 20 124 16.1 18 128 14.1 14 109 12.8 52 171 30.4

48

Table 5: Genetic diversity at eight microsatellite loci for three populations along the upper Yangtze River, China, including the total number of alleles (A), allelic richness (AR) standardized to 20 individuals per population, expected heterozygosity (HE), observed heterozygosity (HO), and the p-value associated with the exact test for deviation from Hardy-Weinberg equilibrium.

Locus A AR HE HO P - value Population 1 = Yibin (n = 20) Pvi 004 17 17 0.906 0.900 0.344 Pvi 026 12 12 0.913* 0.800 0.035 Pvi 038 19 19 0.927 0.950 0.347 Pvi 045 10 10 0.858 0.850 0.623 Pvi 052 13 13 0.913 0.900 0.650 Pvi 055 11 11 0.794 0.650 0.086 Pvi 066 15 15 0.913 0.900 0.175 Pvi 114 12 12 0.691* 0.550 0.033 Mean 13.6 13.6 0.864 0.813 Population 2 = Luzhou (n = 20) Pvi 004 17 17 0.935* 0.900 0.025 Pvi 026 17 17 0.947 0.950 0.143 Pvi 038 21 21 0.949 0.950 0.883 Pvi 045 10 10 0.874 0.900 0.164 Pvi 052 15 15 0.906 0.850 0.647 Pvi 055 14 14 0.905 0.850 0.167 Pvi 066 19 19 0.959 0.900 0.175 Pvi 114 15 15 0.892** 0.650 0.001 Mean 16.0 16.0 0.921 0.869 Population 3 = Song Ji (n = 21) Pvi 004 17 16.6 0.923 0.905 0.475 Pvi 026 15 14.8 0.916 0.905 0.862 Pvi 038 20 19.5 0.943 0.905 0.359 Pvi 045 10 9.90 0.812 0.905 0.661 Pvi 052 18 17.7 0.934** 0.667 0.001 Pvi 055 10 9.81 0.830* 0.762 0.039 Pvi 066 18 17.6 0.940 0.905 0.244 Pvi 114 16 15.6 0.868* 0.762 0.029 Mean 15.5 15.2 0.896 0.839 * indicates locus population combinations that deviate from HWE at p < 0.05 ** indicates those that deviate at the table-wide alpha (k = 24 or 8) after sequential Bonferroni correction (Rice, 1989)

49 richness for each locus and averaged across loci was identical to the number of alleles within the Luzhou and Yibin populations (Table 5). For the Song Ji population, the allelic richness per locus was 15.2, slightly lower than the average number of alleles (Table 5).

Expected heterozygosity for each locus within the three populations was high, ranging from 0.794 – 0.949 with the exception of one locus, Pvi 114 in the Yibin population, which had an expected heterozygosity of 0.691 (Table 5). Within each of the three populations, locus Pvi 038 had the highest expected heterozygosity ranging from

0.927 – 0.943 and the highest number of alleles per locus (Table 5). Mean expected heterozygosity for each population was similarly high, ranging from 0.864 in the Yibin population to 0.921 in the Luzhou population (Table 5). Mean observed heterozygosity

(HO) within each of the populations was lower than the expected heterozygosity (HE) and ranged from 0.813 – 0.869 (Table 5). Within each of the populations the range in observed heterozygosity was greater than that of the expected heterozygosity (Table 5).

The smallest range in observed heterozygosity was from 0.762 – 0.905 in the Song Ji population and the largest range was from 0.550 – 0.950 in the Yibin population (Table

5). Within the Yibin and Luzhou populations, locus Pvi 114 had the lowest observed heterozygosity and in the Song Ji population Pvi 052 had the lowest (Table 5).

Deviations from Hardy-Weinberg equilibrium within Populations

Exact tests for deviations from Hardy-Weinberg equilibrium (HWE) for each locus within the three populations identified 7 of the 24 population-locus combinations that did not conform to Hardy-Weinberg expectation at the p < 0.05 significance level (Table 50

5). Of these, 3 were in the Song Ji population and 2 were in each of the Luzhou and Yibin populations (Table 5). For the most part, there was no consistency amongst loci within populations with the exception of Pvi 114 which was found to significantly deviate from

HWE in all three populations at the p < 0.05 significance level. After applying the sequential Bonferroni correction for simultaneous statistical inference, only 2 population-locus combinations were significant for departure from HWE at the table wide alpha level (k = 8) (Table 5). There was no indication of heterozygosity excess and exact tests of heterozygosity deficiency indicated that only one locus-population combination showed significant deviation from Hardy-Weinberg equilibrium once the sequential Bonferroni correction was taken into account (Pvi 052 in the Songji population, p < 0.0001).

Linkage Disequilibrium between Pairs of Loci within Populations

Likelihood ratio tests of linkage disequilibrium between pairs of loci within populations detected 16 significant associations out of a total of 84 possible combinations at the 0.05 significance level. No strong associations between particular pairs of loci were detected however, locus Pvi 114 was included in 8 of the 16 tests that were significant for linkage disequilibrium across the three populations. After applying the sequential Bonferroni correction for multiple tests (k = 28) only one pair of loci (Pvi

055 and Pvi 114) in the Yibin population showed significant linkage disequilibrium.

51

Population Structure

AMOVA results indicated weak but statistically significant population structure among the three sites (Global FST = 0.00648, 95% CI 0.00137 – 0.01110, p = 0.04703). The majority of the variation, 99.35 %, was attributed to within population differences in alleles and only 0.65 % of the variation was explained by differences among the populations (Table 6). This weak population structure was also confirmed through tests of genic differentiation based on allele frequencies for each population pair. Of the three population pairs tested, only the Song Ji and Yibin populations were significantly differentiated (p = 0.030654) however this value was no longer significant once the sequential Bonferroni correction was taken into account (k = 3). Similarly, exact tests of population differentiation found no indication of differentiation between any of the population pairs. Additionally, pairwise FST values between each of the population pairs ranged from 0.00353 – 0.00856, none of which were statistically significant (all p > 0.05)

(Table 7).

For the principal co-ordinates analysis the first three axes explained 59.18 % of the variance in the data, with the first axis explaining 25.77 %, the second axes explaining

17.39 % and the third explaining 16.02 %. Scatter plots of the individual samples grouped according to their sampled population plotted against the co-ordinate axes provide a visual representation of the lack of structure among the three sites (Figure 2).

A three dimensional scatter plot of the samples against the first three principal co- ordinates axes further illustrates the lack of genetic structure among the sampled populations in the upper Yangtze River (Appendix 2). 52

Table 6: AMOVA analysis averaged across loci. Global FST = 0.00648 over all loci (distance matrix calculated from the number of different alleles), 95% C.I. (0.00137 – 0.01110), P (rand > obs value) = 0.04703.

Source of Variation Sum of Squares Variance components % Variation Among populations 8.013 0.02066 0.64812 Within populations 376.839 3.16672 99.35188 Total 384.852 3.18737

53

Table 7: Pairwise FST values among the three populations (above the diagonal) with the corresponding p-values (below the diagonal).

Population Song Ji Luzhou Yibin Song Ji - 0.00722 0.00856 Luzhou 0.07475 - 0.00353 Yibin 0.06504 0.25780 -

54

a)

Song Ji

ord. 2 -

Co Luzhou Yibin

Co-ord. 1

b)

Song Ji ord. 3

- Luzhou Co Yibin

Co-ord. 2

c)

Song Ji ord. 3

- Luzhou Co Yibin

Co-ord. 1

Figure 4: Two dimensional plots of the samples from the Song Ji, Luzhou, and Yibin populations on a) principle coordinate axes 1 and 2; b) axes 2 and 3; c) axes 1 and 3.

55

Chapter 4– Discussion:

This study was successful in the isolation and characterization of eight microsatellite loci representing the first set of primers for the specific analysis of P. vachelli. Additionally it is the first to investigate the levels of genetic variability and population differentiation for this species within a localized area. The sampling sites used provide a baseline estimate of the levels of genetic variation and population structure along an unfragmented stretch of the upper Yangtze River, providing a foundation for future analyses of the effects of fragmentation on genetic variation in this species as development of hydroelectric installations occurs.

Microsatellite Variability

The microsatellite loci isolated and optimized in this study revealed a high level of genetic variation within P. vachelli. The level of variation detected amongst the sampled populations is much higher than that generally found in other freshwater species and more similar to that reported in marine fish (DeWoody and Avise, 2000; Ward et al.,

1994). In their review, of the available literature to that date, DeWoody and Avise (2000) reported that variation at microsatellite loci was related to the ecology and life history strategies of fish with levels of genetic variation generally highest in marine species, intermediate in anadromous populations, and lowest in freshwater fish. However, more recent studies have found a range in measures of genetic diversity in freshwater species with both the number of alleles per locus and average levels of expected heterozygosity much higher than the average values of 7.1 and 0.58 reported at the time (DeWoody and 56

Avise, 2000) (Appendices 3 and 4). Compared to other riverine species, all eight loci characterized in this study could be considered highly variable with a range of 14 – 29 alleles per locus (Table 3, Appendix 3). Likewise, the observed and expected heterozygosities calculated within the three populations of P. vachelli in the upper

Yangtze River (mean HO = 0.84 and mean HE = 0.90) are much higher than those found in other taxa with varying migratory patterns (Appendix 4). The levels of genetic diversity detected in this study exceed those reported in most studies of freshwater or anadromous fish and likely reflect the particular loci isolated, the lifehistory traits of P. vachelli, and historically large population sizes in the upper Yangtze River.

Comparisons with Other Bagrid Species

Both the number of alleles per locus and levels of heterozygosity reported here are much higher than in most other closely related bagrid species. For example, in the

Southeast Asian river catfish (Mystus memurus), between 2 and 11 alleles per locus were detected among 20 microsatellite loci and the observed and expected heterozygosities ranged from 0.4418 – 0.5730 and 0.4350 – 0.5410 respectively in six populations sampled in Malaysia (Usmani et al., 2008). Similarly lower levels of genetic diversity have been reported in yellow catfish (Horabagridae brachysoma and H. nigricollaris) where the number of alleles ranged from 1 – 8 amongst five loci and observed heterozygosity was 0.463 and 0.443 in one sample of each of the two species from India (Abdul Muneer et al., 2011). In the Yangtze River, studies of the longsnout catfish (Leiocassis longirostris) have also reported relatively low levels of genetic diversity within and among the populations using mitochondrial DNA but not using microsatellite markers 57

(Wang et al., 2006; Yang et al., 2011). In one study, levels of genetic variability were similar to those reported here, with between 6 and 29 alleles found at 20 microsatellite loci and values of observed and expected heterozygosity ranging from 0.8205 – 0.8697 and 0.7757 – 0.8300 respectively (Yang et al., 2011).

High Levels of Genetic Diversity

The high level of allelic diversity and the above average levels of heterozygosity detected imply large population sizes of P. vachelli in the upper region of the Yangtze

River (Nei, 1987; Knaepkens et al., 2004). This is supported by the fact that although individual darkbarbel catfish are small (approximately 10 – 30 g) they make up a large proportion of the fishing yield by weight within the sampled region of the upper Yangtze

River (Duan et al., 2002). The large population size and high levels of variation within P. vachelli are in the range of those reported in marine species (DeWoody and Avise, 2000;

Tzeng et al., 2009) and have most likely been influenced by the geological history of the

Yangtze River. The Yangtze River is the longest river in China and with its vast length and historical lack of physical barriers, is capable of housing large populations of freshwater species with high levels of gene flow. Unlike the majority of freshwater environments in the northern hemisphere where the majority of ichthyological research has been conducted, the Yangtze River was largely unaffected by glaciation during the Quaternary

(Yang et al., 2011; Lui, 2003). The stable history of this region combined with the size of the habitat has thus allowed for a relatively long evolutionary period for the populations of P. vachelli to expand and accumulate mutations at microsatellite loci. 58

Alternatively the high genetic variability may indicate a previously large population size that is currently undergoing a population bottleneck as a result of increasing fishing pressure during the past two decades (Chen et al., 2009; Zeng et al.,

2010). Recent population bottlenecks are usually evident by an increased level of heterozygosity and a loss of rare alleles such that the number of alleles decreases faster than the range in allele size (Garza and Williamson, 2001). This was seen in the distribution of allele frequencies (Appendix 1); however tests for deviations from HWE found no significant excess in heterozygosity. Thus this study does not indicate recent population declines (Cornuet and Luikart, 1996). However, some studies have found that the genetic signature of a reduction in population size can take several generations to become evident (Whiteley et al., 2006b; Yang et al., 2011) and accordingly these results should be interpreted with caution. On the other hand, previously larger population sizes could have kept allele frequencies at mutation-drift equilibrium and prevented the genetic changes associated with reductions in population size (Tzeng et al., 2009).

Other factors that may be influencing the high levels of genetic variation detected in this study are the ecology and reproductive strategy of P. vachelli. Amongst freshwater species, the highest levels of variability have been identified in habitat generalists (Heithaus and Laushman, 1997; Reid et al., 2008a) and non-migratory fish lacking parental care (Lassala and Renesto, 2007). As stated above, P. vachelli have a wide niche range throughout the Yangtze River and are capable of enduring a variety of environmental conditions. These fish also have high lifetime reproductive efforts as they spawn multiple times throughout their lifespan and high relative fecundity levels have 59 been reported with females capable of laying 1100 – 19675 eggs per year (Duan and Sun,

1999b; Yang, 2006). In the absence of physical barriers, the combination of their broad ecological tolerance and high recruitment rates would allow for large populations with continuous distributions and provide the mechanism for accumulating mutations leading to higher levels of variability at microsatellite loci in this species.

Population Structure

Analyses of population structure among the sampling sites found little indication of population differentiation along the unfragmented stretch of the upper Yangtze River.

Results of the AMOVA analysis did support a weak but statistically significant structure among the populations (global FST = 0.0065, p < 0.05), however nearly all (99.35 %) of the variation was attributed to within population differences in allele frequencies and only a slight portion was due to differences among the populations (Table 6). This proportion of within population variation is higher than what is generally reported in other riverine species with varying migratory patterns (eg. 85.00 % of variation within populations in the anadromous eulachon (Thaleichthys pacificus), McLean and Taylor, 2001; 98.43 % in the non-migratory Northern sheatfish (Silurus soldatovi), Quan et al., 2006; 94.24 % in the potadromodous mud carp (Cirrhina molitorella), Zhao et al., 2011) including the closely related semi-migratory longsnout catfish (L. longirostris) in the Yangtze River where 97.26 % of the variance was detected within populations and just 2.74 % was among regions (Yang et al., 2011). This high level of intra-population variance and low level of divergence among populations most likely reflects the high variability of the loci and large population sizes with gene flow among them in the absence of barriers. 60

Exact tests for population differentiation found no significant structure among the three populations. This was evidenced in the three dimensional plot of the samples along the first three principal coordinate axes where samples from all three locations formed a single cluster (Appendix 2). In addition, pairwise FST values were low, ranging from 0.0035 – 0.0086, and not significant (Table 7) providing further support that P. vachelli inhabiting the sampled stretch of the upper Yangtze River are part of a single large population with free gene flow. These low FST values are similar to those reported amongst populations of marine species (eg. FST = 0.007 in Blue mackerel (Scomber australasicus) from the East and South China Seas, Tzeng et al., 2009; FST = 0.0066 –

0.0081 in the Korean rockfish (Sebastes schlegeli) around the coast of Korea, An et al.,

2012) and in the migratory sutchi catfish (Pangasianodon hypophthalmus) in the lower

Mekong River in Cambodia (FST range 0.0001 - 0.0065, So et al., 2006). Comparable values have also been found amongst populations of migratory benthic redhorse species

(Moxostoma ssp.) and Greenside darters (Etheostoma blennioides) in rivers in North

America which are both unfragmented and recently dammed yet permeable to fish movement (FST values ranging 0.004 - 0.016 in the Muskegon river and 0.005 - 0.032 in the divided Trent and Grand Rivers in redhorse, Reid et al., 2008a and b; FST ranging

0.001 - 0.079 in the unfragmented Sydenham and Ausable rivers and 0.001 - 0.024 in the fragmented Grand and Thames rivers in Greenside darters, Beneteau et al., 2009). Thus caution should be taken in interpreting these results as indicative of a homogeneous population of P. vachelli with on-going gene flow within the sampled region of the upper 61

Yangtze River as some low levels of differentiation may be present amongst the populations.

Models of Gene Flow

While none of the pairwise FST values were significant, it is noteworthy that the level of genetic differentiation between the putative populations related to the geographic distances among the sampling sites. This implies the possibility of a slight effect of isolation-by-distance on gene flow amongst the populations. Previous studies of population structure in fish with a continuous distribution within river basins have reported patterns of gene flow that follow a stepping-stone model in one dimension where dispersal occurs between adjacent populations that are linearly arranged along a river resulting in a positive correlation between calculations of FST and geographic distance (Hernandez-Martich and Smith, 1997; Estoup et al., 1998; Carlsson and Nilsson,

2000). Although a strong relationship between geographic distance and pairwise measures of differentiation was not observed in this study, values were highest among the Yibin and Song Ji populations, the two that are the furthest apart. Under the stepping stone model of gene flow, the lack of a strictly increasing monotonic relationship between genetic and geographic distances implies that the populations of P. vachelli along the upper stretch of the Yangtze River are not in regional equilibrium with respect to gene flow and genetic drift (Hutchison and Templeton, 1999). However, the small number of populations used in this study and the relatively large geographic distance (considering the likely dispersal capabilities of this fish) between the sampling 62 locations makes it difficult to determine which of these factors has a greater effect on the observed population structure.

Other possible models of gene flow may better reflect the patterns of genetic variance observed among the populations of P. vachelli in the upper Yangtze River. Tero et al. (2003) presented five models of migration and gene flow applicable to populations that are linearly arranged such as the populations of P. vachelli examined in this study.

Of the models presented there, the distribution of genetic variance detected in this study suggests that the gene flow among the P. vachelli populations in this region may be conforming to the classical metapopulation model. This model allows for free gene flow throughout the entire range of linearly distributed populations and would lead to a lack of population structure among the sampled sites (Tero et al., 2003). The absence of significant genetic differentiation among any of the pairwise comparisons observed here supports the notion that the P. vachelli currently inhabiting this stretch of the Yangtze

River form a single panmictic population and do not represent separate management units (Moritz, 1994; Palsboll et al., 2006).

High Variability and Lack of Population Differentiation

It is possible that the high variability of the microsatellite loci relative to the sample size used in this study may have affected the level of discernible structure among the populations in the sampled stretch of the upper Yangtze River. Although FST has been recommended for use in examining genetic differentiation in fish populations when there is a high number of alleles per locus (O’Connell and Wright, 1997), a negative 63 relationship between microsatellite locus variability and the power of microsatellite data to detect weak population structure has been reported in other fish species (O’Reilly et al., 2004). Thus, although FST is capable of detecting weak population structure in populations with high levels of gene flow and relatively large population sizes, it is possible that the specific loci isolated in this study are too variable to provide resolution of underlying differentiation among the sampled populations. However, measures of genetic distance such as FST are less sensitive the relative variability of loci and population size when compared to other methods for inferring population structure such as assignment tests (Manel et al., 2005). Moreover, within some fish species a large range in the number of alleles per locus has been observed (eg. Salgueiro et al., 2003;

Seeb et al., 2007; see Appendix 3) therefore the high variability of the loci characterized in this study may simply be an artifact of sampling bias. More loci would need to be optimized to explore this possibility.

Another factor which may be obscuring the detection of population structure is size homoplasy amongst the various alleles. Size homoplasy refers to two alleles that are identical in state (size of electromorph) but not by (evolutionary) decent. Although this phenomenon is generally not problematic in evaluating population genetic structure, simulation studies have indicated that with high mutation rates and large population sizes, microsatellite size homoplasy may interfere with the resolution of weak population differentiation (Estoup et al., 2002). Additionally, some experiments have found that depending on the locus, the levels of detectable size homoplasy is not negligible (Adams et al., 2004). As the mutation model of microsatellite evolution is still unclear (Weber 64 and Wong, 1993) and the high variability detected amongst the loci indicates that the population of P. vachelli within the sampled region may be large, it is possible that convergent mutations may be interfering in the detection of any low levels of structure amongst the populations. Moreover, if the levels of mutation are high relative to the amount of migration, calculations of genetic differentiation may not be accurate (Balloux and Lugon-Moulin, 2002). Consequently, it may be necessary to isolate additional loci with lower variability and to include a population that is further geographically separated from those analyzed in this study to verify the lack of population structure among the populations in this stretch of the river.

The presence of some differentiation among the sampling sites is consistent with the large number of private alleles that were detected in each of the populations.

Alternatively, the presence of a high percentage of unique alleles could have arisen through population expansion or sustained large population sizes (Triantafyllidis et al.,

2002) and might be an artifact of the high variability of the microsatellite loci used combined with the low sample size. For studies of population structure using microsatellite loci, samples sizes of 50 or greater are recommended (O’Reilley et al.,

2004) to ensure accuracy in characterizing the distribution of variance and minimizing sampling error (Liu and Cordes, 2004). Therefore future research on the effects of fragmentation on the distribution of genetic variance in P. vachelli within the upper

Yangtze River should consist of larger sample sizes and explore the possibility of alternative markers or loci to guarantee sufficient statistical power. 65

Also noteworthy, is that such high levels of variation and lack of population structure were detected specifically in the upper section of the Yangtze River. Other studies of non-migratory fish with continuous distributions in river basins have generally found lower levels of diversity and higher differentiation in the upstream sections of rivers relative to downstream regions (Hernandez-Martich and Smith, 1997; Yang et al.,

2011). This contrast in the levels of genetic diversity and partitioning of genetic variation in the upper section of the river in comparison to other studies reinforces the notion that there have been historically large populations of P. vachelli within this stretch of the upper Yangtze River and that free gene flow occurs throughout the studied region. A greater understanding of the dispersal capabilities and spawning behaviour of this species is critical for making comprehensive determinations of the population structure of this species. As of yet, little information on the vagility and natural distribution of this species has been published and not much is known regarding the spawning habits of P. vachelli (Wang et al., 2004). As some closely related bagrids in the Yangtze river are known to be semi-migratory (L. Longirostris, Yang et al., 2011) or to have male parental care in terms of nest guarding (P. fulvidraco, Jiang et al., 2008) further knowledge of the natural history of P. vachelli is necessary to interpret the ecological significance of the results presented here.

Limitations of this Study

As fish were purchased from local fisherman, there is some uncertainty as to the exact location that the samples were caught. If fish within sampling sites were caught in separate locations, this could create substructure within the populations and result in 66 deviations from HWE through the Wahlund effect (Carvalho-Costa et al., 2008).

However, deviations from HWE were not consistent across loci therefore the observed deviations are most likely not a result of populations subdivision, admixture, inbreeding, or assortative mating as these would affect the entire genome. Only one locus- population combination showed a deficit in heterozygosity (Pvi 052 in the Song Ji population) for which there was also indication of the presence of a null allele which has been shown to lead to heterozygote deficiency (Brookfield, 1996; Triantafyllidis et al.,

2002). Therefore it is most likely the presence null alleles and the relatively small sample size in comparison to the variability of the microsatellite loci that led to the observed deviations from HWE.

Final Considerations

To maintain the current high levels of genetic variation detected in the populations of darkbarbel catfish along the upper Yangtze River, conservation efforts should be put in place to maintain large population sizes. Future management initiatives should consider the impact of both the high levels of exploitation through fishing as well as the consequences that the construction of hydroelectric installations lacking fish passageways could pose to the continued existence of large and genetically variable populations of this species. Additionally, if breeding programs and intentional translocation of individual fish are going to be continued practices in maintaining wild stocks of this species, efforts should be made to determine both the natural distribution of genetic variation among populations at larger spatial scales within the region and to monitor the genetic composition of domestic strains used in stocking practices. 67

Literature Cited:

Abdul-Muneer, P., Gopalakrishnan, A., Shivanandan, R., Basheer, V., & Ponniah, A. (2011). Genetic variation and phylogenetic relationship between two species of yellow catfish, Horabagrus brachysoma and H. nigricollaris (Teleostei: Horabagridae) based on RAPD and microsatellite markers. Molecular Biology Reports, 38(4), 2225- 2232.

Adams, R. I., Brown, K. M., & Hamilton, M. B. (2004). The impact of microsatellite electromorph size homoplasy on multilocus population structure estimates in a tropical tree (Corythophora alta) and an anadromous fish (morone saxatilis). Molecular Ecology, 13(9), 2579-2588.

An, H., Kim, M., Park, K., Cho, K., Bae, B., Kim, J., et al. (2012). Genetic diversity and population structure in the heavily exploited Korean rockfish, Sebastes schlegeli, in Korea. Journal of the World Aquaculture Society, 43(1), 73-83.

Aspi, J., Roininen, E., Kiiskilä, J., Ruokonen, M., Kojola, I., Bljudnik, L., et al. (2009). Genetic structure of the northwestern Russian wolf populations and gene flow between Russia and Finland. Conservation Genetics, 10(4), 815-826.

Balloux, F., & Lugon-Moulin, N. (2002). The estimation of population differentiation with microsatellite markers. Molecular Ecology, 11(2), 155-165.

Beneteau, C., Mandrak, N., & Heath, D. (2009). The effects of river barriers and range expansion of the population genetic structure and stability in greenside darter (Etheostoma blennioides) populations. Conservation Genetics, 10(2), 477-487.

Bernatchez, L., & Duchesne, P. (2000). Individual-based genotype analysis in studies of parentage and population assignment: How many loci, how many alleles? Canadian Journal of Fisheries and Aquatic Sciences, 57(1), 1-12.

Briggs, J. C. (2005). The biogeography of otophysan fishes (Ostariophysi: Otophysi): A new appraisal. Journal of Biogeography, 32(2), 287-294.

Brookfield, J. F. Y. (1996). A simple new method for estimating null allele frequency from heterozygote deficiency. Molecular Ecology, 5(3), 453-455.

Carlsson, J., Olsén, K. H., Nilsson, J., Øverli, Ø., & Stabell, O. B. (1999). Microsatellites reveal fine-scale genetic structure in stream-living brown trout. Journal of Fish Biology, 55(6), 1290-1303.

Carlsson, J., & Nilsson, J. (2000). Population genetic structure of brown trout (Salmo trutta L.) within a northern boreal forest stream. Hereditas, 132(3), 173-181. 68

Carvalho-Costa, L. F., Hatanaka, T., & Galetti Jr., P. M. (2008). Evidence of lack of population substructuring in the Brazilian freshwater fish Prochilodus costatus. Genetics and Molecular Biology, 31(1), 377-380.

Chen, D., Xiong, F., Wang, K., & Chang, Y. (2009). Status of research on Yangtze fish biology and fisheries. Environmental Biology of Fishes, 85(4), 337-357.

Chen, X., Yan, Y., Fu, R., Dou, X., & Zhang, E. (2008). Sediment transport from the Yangtze River, China, into the sea over the post-Three Gorge Dam period: A discussion. Quaternary International, 186(1), 55-64.

Chen, X., Zong, Y., Zhang, E., Xu, J., & Li, S. (2001). Human impacts on the Changjiang (Yangtze) River Basin, China, with special reference to the impacts on the dry season water discharges into the sea. Geomorphology, 41(2-3), 111-123.

Chistiakov, D. A., Hellemans, B., & Volckaert, F. A. M. (2006). Microsatellites and their genomic distribution, evolution, function and applications: A review with special reference to fish genetics. Aquaculture, 255(1-4), 1-29.

Clemento, A., Anderson, E., Boughton, D., Girman, D., & Garza, J. (2009). Population genetic structure and ancestry of Oncorhynchus mykiss populations above and below dams in south-central California. Conservation Genetics, 10(5), 1321-1336.

Cornuet, J. M., & Luikart, G. (1996). Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics, 144(4), 2001-2014.

Costello, A. B., Down, T. E., Pollard, S. M., Pacas, C. J., Taylor, E. B., & Wallis, G. (2003). The influence of history and contemporary stream hydrology on the evolution of genetic diversity within species: An examination of microsatellite DNA variation in bull trout, Salvelinus confluentis (Pices: Salmonidae). Evolution, 57(2), 328-344.

Dehais, C., Eudeline, R., Berrebi, P., & Argillier, C. (2010). Microgeographic genetic isolation in chub (Cyprinidae: Squalius cephalus) population of the Durance River: Estimating fragmentation by dams. Ecology of Freshwater Fish, 19(2), 267-278.

Deiner, K., Garza, J. C., Coey, R., & Girman, D. J. (2007). Population structure and genetic diversity of trout (Oncorhynchus mykiss) above and below natural and man-made barriers in the Russian River, California. Conservation Genetics, 8(2), 437-454.

DeWoody, J. A., & Avise, J. C. (2000). Microsatellite variation in marine, freshwater and anadromous fishes compared with other animals. Journal of Fish Biology, 56(3), 461- 473. 69

Diogo, R. (2005). Morphological evolution, adaptations, homoplasies, constraints and evolutionary trends; Catfishes as a case study on general phylogeny and macroevolution. Enfield, US: Science publishers, Inc.

Dolores, M., Lassala, P., & Renesto, E. (2007). Reproductive strategies and genetic variability in tropical freshwater fish. Genetics and Molecular Biology, 30(3), 690- 697.

Duan, X. B., Chen, D. Q., Liu, S.P., Chi, C.G., & Yang, R. H. (2002). Studies on status of fishery resources in Three Gorges Reservoir reaches of the Yangtze River. Acta Hydrobiologica Sinca, 26: 6, 605-611. (In Chinese with English abstract).

Duan, Z., and J. Sun. (1999a). Studies on the age and growth of Pelteobagrus vachelli (Richardson). Acta Hydrobiologica Sinica, 06. doi: CNKI:SUN:SSWX.0.1999-06-011 (In Chinese with English abstract).

Duan, Z., and J. Sun. (1999b) Studies on the reproductive biology of Pelteobagrus vachelli (Richardson). Acta Hydrobiologica Sinica, 06. doi: cnki:ISSN:1000-3207.0.1999- 06-010 (In Chinese with English abstract).

Duan, Z., Zhao, K., Peng, Z., Li, J., Diogo, R., Zhao, X., et al. (2009). Comparative phylogeography of the Yellow River schizothoracine fishes (Cyprinidae): Vicariance, expansion, and recent coalescence in response to the Quaternary environmental upheaval in the Tibetan Plateau. Molecular Phylogenetics and Evolution, 53(3), 1025- 1031.

Elmer, K. R., Davilla, J. A., & Lougheed, S. C. (2006). Isolation of simple and compound polymorphic tetranucleotide microsatellites for the neotropical leaflitter frog Eleutherodactylus ockendeni (Leptodactylidae). Molecular Ecology Notes, 6(3), 891- 893.

Estoup, A., Rousset, F., Michalakis, Y., Cornuet, J., Adriamanga, M., & Guyomard, R. (1998). Comparative analysis of microsatellite and allozyme markers: A case study investigating microgeographic differentiation in brown trout (Salmo trutta). Molecular Ecology, 7(3), 339-353.

Estoup, A., Jarne, P., & Cornuet, J. (2002). Homoplasy and mutation model at microsatellite loci and their consequences for population genetics analysis. Molecular Ecology, 11(9), 1591-1604. 70

Excoffier, L., Smouse, P. E., & Quattro, J. M. (1992). Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics, 131(2), 479-491.

Excoffier, L., & Lischer, H. L. (2010). Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10(3), 564-567.

Fan, X., Wei, Q., Chang, J., Rosenthal, H., He, J., Chen, D., et al. (2006). A review on conservation issues in the upper Yangtze River – a last chance for a big challenge: Can Chinese paddlefish (Psephurus gladius), Dabry´s sturgeon, (Acipenser dabryanus) and other fish species still be saved? Journal of Applied Icthyology, 22(Supplemental 1), 32-39.

Ferraris, C. J. (2007). Checklist of catfishes, recent and fossil (Osteichthyes: Siluriformes), and catalogue of siluriform primary types. Zootaxa 1418, 1-628.

Fishbase. Pelteobagrus vachelli http://www.fishbase.org/summary/Pseudobagrus- vachellii.html

Fu, C., Wu, J., Chen, J., Wu, Q., & Lei, G. (2003). Freshwater fish biodiversity in the Yangtze River basin of China: Patterns, threats and conservation. Biodiversity and Conservation, 12(8), 1649-1685.

Fu, S., Zeng, L., Li, X., Pang, X., Cao, Z., Peng, J., et al. (2009). The behavioural, digestive and metabolic characteristics of fishes with different foraging strategies. Journal of Experimental Biology, 212(14), 2296-2302.

Fumagalli, L., Snoj, A., Jesenšek, D., Balloux, F., Jug, T., Duron, O., et al. (2002). Extreme genetic differentiation among the remnant populations of marble trout (Salmo marmoratus) in Slovenia. Molecular Ecology, 11(12), 2711-2716.

Gaggiotti, O. E., Lange, O., Rassmann, K., & Gliddon, C. (1999). A comparison of two indirect methods for estimating average levels of gene flow using microsatellite data. Molecular Ecology, 8(9), 1513-1520.

Gan, L., Ma, X. Z., Zhang, W. B., Chen, Z. Z., Yuan, Y., & Wang, W. (2007). A study on the feeding rhythm and daily feeding rate of larval and juvenile darkbarbel catfish (Pelteobagrus vachelli) in pond. Freshwater Fisheries, 06. doi: CNKI:SUN:DSYY.0.2007- 06-018 (In Chinese with English abstract).

71

Gan, L., Ma, X. Z., Zhang, W. B., Chen, Z.Z., Yuan, Y. & Wang, W. (2008). The growth comparison of larvae and juveniles of Pelteobagrus vachelli and Pelteobagrus fulvidraco. Journal of South China Agricultural University, 03. doi: CNKI:SUN:HNNB.0.2008-03-016 (In Chinese with English abstract).

Garza, J. C., & Williamson, E. G. (2001). Detection of reduction in population size using data from microsatellite loci. Molecular Ecology, 10(2), 305-318.

Glenn, T. C., & Schable, N. A. (2005). Isolating microsatellite DNA loci. Methods in Enzymology, 395, 202-222.

Goudet, J. (1995). FSTAT (version 1.2): A computer program to calculate F-statistics. Journal of Heredity, 86(6), 485-486.

Goudet, J. (2002). FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3.2). [Software] available from http://www2.unil.ch/popgen/softwares/fstat.htm. Updated from Goudet (1995)

Goudet, J., Raymond, M., de-Meeus, T., & Rousset, F. (1996). Testing differentiation in diploid populations. Genetics, 144(4), 1933-1940.

Guo, S. W., & Thompson, E. A. (1992). Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics, 48(2), pp. 361-372.

Hamilton, M. B., Pincus, E. L., Di Fiore, A., & Flesher, R. C. (1999). Universal linker and ligation procedures for construction of genomic DNA libraries enriched for microsatellites. Biotechniques, 27(3), 500-507.

Haponski, A. E., Marth, T. A., & Stepien, C. A. (2007). Genetic divergence across a low- head dam: A preliminary analysis using logperch and greenside darters. Journal of Great Lakes Research, 33, 117-126.

Hardman, M. (2005). The phylogenetic relationships among non-diplomystid catfishes as inferred from mitochondrial cytochrome b sequences; the search for the ictalurid sister taxon (Otophysi: Siluriformes). Molecular Phylogenetics and Evolution, 37(3), 700-720.

Heggenes, J., & Røed, K. H. (2006). Do dams increase genetic diversity in brown trout (Salmo trutta)? microgeographic differentiation in a fragmented river. Ecology of Freshwater Fish, 15(4), 366-375.

Heiner, M., Higgins, J., Li, X., & Baker, B. (2011). Identifying freshwater conservation priorities in the Upper Yangtze River Basin. Freshwater Biology, 56(1), 89-105. 72

Heithaus, M. R., & Laushman, R. H. (1997). Genetic variation and conservation of stream fishes; influence of ecology, life history, and water quality. Canadian Journal of Fisheries and Aquatic Sciences, 54(8), 1822-1836.

Hernandez-Martich, J., & Smith, M. H. (1997). Downstream gene flow and genetic structure of Gambusia holbrooki (eastern mosquitofish) populations. Heredity, 79(3), 295-301.

Hui, X., Denghong, L., Daming, T., & Bofu, G. (2006). Structural and quality changes in the spawning stock of the Chinese sturgeon Acipenser sinensis below the Gezhouba Dam (Yangtze River). Journal of Applied Ichthyology, 22(s1), 111-115.

Hutchinson, W. F., Oosterhout, C. v., Rogers, S. I., & Carvalho, G. R. (2003). Temporal analysis of archived samples indicates marked genetic changes in declining North Sea cod (Gadus morhua). Proceedings: Biological Sciences, 270(1529), 2125-2132.

Hutchison, D. W., & Templeton, A. R. (1999). Correlation of pairwise genetic and geographic distance measures: Inferring the relative influences of gene flow and drift on the distribution of genetic variability. Evolution, 53(6), pp. 1898-1914.

Jiang, P., Yin, H. B., Zhang, Y., Li, D. Y., & Sun, X. W. (2008). Analysis of parentage between male Pelteobagrus fulvidraco and guarded fertilized eggs. Acta Zoologica Sinica, 05. doi: CNKI:SUN:BEAR.0.2008-05-006 (In Chinese with English abstract).

Knaepkens, G., Bervoets, L., Verheyen, E., & Eens, M. (2004). Relationship between population size and genetic diversity in endangered populations of the European bullhead (Cottus gobio): Implications for conservation. Biological Conservation, 115(3), 403-410.

Kruckenhauser, L., Bryant, A., Griffin, S., Amish, S., & Pinsker, W. (2009). Patterns of within and between-colony microsatellite variation in the endangered Vancouver Island marmot (Marmota vancouverensis): Implications for conservation. Conservation Genetics, 10(6), 1759-1772.

Ku, X., Peng, Z., Diogo, R., & He, S. (2007). MtDNA phylogeny provides evidence of generic polyphyleticism for East Asian bagrid catfishes. Hydrobiologia, 579(1), 147- 159.

Laroche, J., Durand, J. D., Bouvet, Y., Guinand, B., & Brohon, B. (1999). Genetic structure and differentiation among populations of two cyprinids, Leuciscus cephalus and Rutilus rutilus, in a large European river. Canadian Journal of Fisheries and Aquatic Sciences, 56(9), 1650-1667. 73

Laroche, J., & Durand, J. D. (2004). Genetic structure of fragmented populations of a threatened endemic percid of the Rhône River: Zingel asper. Heredity, 92(4), 329- 334.

Li, Q., Park, C., Endo, T., & Kijima, A. (2004). Loss of genetic variation at microsatellite loci in hatchery strains of the Pacific abalone (Haliotis discus hannai). Aquaculture, 235(1-4), 207-222.

Liu, K. B. (1988). Quaternary history of the temperate forests of China. Quaternary Science Reviews, 7(1), 1-20.

Liu, Z. J., & Cordes, J. F. (2004). DNA marker technologies and their applications in aquaculture genetics. Aquaculture, 238(1-4), 1-37.

Long, Y., Li, Q., Chen, Y.S., & Diao, X. M. (2005). Technology for artificial propogation of wild Pelteobagrus vachelli in the Jialing River. Journal of Southwest Agricultural University, 05. doi: cnki:ISSN:1000-2642.0.2005-05-034 (In Chinese with English abstract).

Lu, G., Li, S., & Bernatchez, L. (1997). Mitochondrial DNA diversity, population structure, and conservation genetics of four native carps within the Yangtze River, China. Canadian Journal of Fisheries and Aquatic Sciences, 54(1), 47-58.

Ma, X., Bangxi, X., Yindong, W., & Mingxue, W. (2003). Intentionally introduced and transferred fishes in China's inland waters. Asian Fisheries Science, 16, 279-290.

Manel, S., Gaggiotti, O. E., & Waples, R. S. (2005). Assignment methods: Matching biological questions with appropriate techniques. Trends in Ecology and Evolution, 20(3), 136-142.

McLean, J. E., & Taylor, E. B. (2001). Resolution of population structure in a species with high gene flow: Microsatellite variation in the eulachon (Osmeridae: Thaleichthys pacificus). Marine Biology, 139(3), 411-420.

Meldgaard, T., Nielsen, E. E., & Loeschcke, V. (2003). Fragmentation by weirs in a riverine system: A study of genetic variation in time and space among populations of european grayling (thymallus thymallus) in a Danish river system. Conservation Genetics, 4(6), 735-747.

Mo, T. (1991). Anatomy, relationships and systematics of the Bagridae (Teleostei: Siluridei) with a hypothesis of siluroid phylogeny. Koenigstein: Koeltz Scientific Books. 74

Moritz, C. (1994). Defining ‘Evolutionarily significant units’ for conservation. Trends in Ecology & Evolution, 9(10), 373-375.

Nei, M. (1987). Molecular Evolutionary Genetics. Columbia University Press, New York, USA.

Neraas, L. P., & Spruell, P. (2001). Fragmentation of riverine systems: The genetic effects of dams on bull trout (Salvelinus confluentus) in the Clark Fork River system. Molecular Ecology, 10(5), 1153-1164.

Ng, H. H. (2003). Phylogeny and systematics of Bagridae. In Kapoor B.G., G Arratia, M. Chardon, and R. Diogo (Eds), Catfishes. Science Publishers, Enfield, 439 – 463.

O‘Connell, M., & Wright, J. M. (1997). Microsatellite DNA in fishes. Reviews in Fish Biology and Fisheries, 7(3), 331-363.

O'Reilly, P. T., Canino, M. F., Bailey, K. M., & Bentzen, P. (2004). Inverse relationship between FST and microsatellite polymorphism in the marine fish, walleye pollock (Theragra chalcogramma): Implications for resolving weak population structure. Molecular Ecology, 13(7), 1799-1814.

Palsbøll, P. J., Bérubé, M., & Allendorf, F. W. (2007). Identification of management units using population genetic data. Trends in Ecology & Evolution, 22(1), 11-16.

Park, S. D. E. (2001), Trypanotolerance in west African cattle and the population genetic effects of selection [Ph.D. thesis] University of Dublin.

Park, Y., Chang, J., Lek, S., Cao, W., & Brosse, S. (2003). Conservation strategies for endemic fish species threatened by the Three Gorges Dam. Conservation Biology, 17(6), 1748-1758.

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.

Peng, J. L., Cao, Z. D., & Fu, S. J. (2007). Effect of acute low temperature stress on oxygen consumption rate and ventilation frequency in Silurus meridionalis and Pelteobag vachelli. Journal of Anhui Agricultural Sciences, 25. doi: CNKI:SUN:AHNY.0.2007-25-057 (In Chinese with English abstract).

Peng, Z., He, S., Wang, J., Wang, W., & Diogo, R. (2006a). Mitochondrial molecular clocks and the origin of the major Otocephalan clades (Pisces: Teleostei): A new insight. Gene, 370, 113-124. 75

Peng, Z., Ho, S. Y. W., Zhang, Y., & He, S. (2006b). Uplift of the Tibetan Plateau: Evidence from divergence times of glyptosternoid catfishes. Molecular Phylogenetics and Evolution, 39(2), 568-572.

Peng, Z. G., Zhang, Y. G., He, S. P., & Chen, Y. Y. (2005). Phylogeny of Chinese catfishes inferred from mitochondrial cytochrome b sequences. Acta Genetica Sinica, 32 (2): 145-154.

Perdices, A., Cunha, C., & Coelho, M. M. (2004). Phylogenetic structure of Zacco platypus (Teleostei, Cyprinidae) populations on the upper and middle Chang Jiang (=Yangtze) drainage inferred from cytochrome b sequences. Molecular Phylogenetics and Evolution, 31(1), 192-203.

Primmer, C. R., Koshinen, M. T., & Piironen, J. (2000). The one that did not get away: Individual assignment using microsatellite data detects a case of fishing competition fraud. Proceedings: Biological Sciences, 267(1453), 1699-1704.

Pringle, C. M., Freeman, M. C., & Freeman, B. J. (2000). Regional effects of hydrologic alterations on riverine macrobiota in the new world: Tropical-temperate comparisons. Bioscience, 50(9), 807.

Qiao, Y., Tang, X., Brosse, S., & Chang, J. (2006). Chinese sturgeon (Acipenser sinensis) in the Yangtze River: A hydroacoustic assessment of fish location and abundance on the last spawning ground. Journal of Applied Ichthyology, 22(s1), 140-144.

Quan, Y., Sun, X., & Liang, L. (2006). Genetic polymorphism of microsatellite DNA in two populations of northern sheatfish (Silurus soldatovi). Acta Genetica Sinica, 33(10), 908-916.

Raymond, M., & Rousset, F. (1995). An exact test for population differentiation. Evolution, 49(6), 1280-1283.

Reid, S. M., Wilson, C. C., Carl, L. M., & Zorn, T. G. (2008a). Species traits influence the genetic consequences of river fragmentation on two co-occurring redhorse (Moxostoma) species. Canadian Journal of Fisheries & Aquatic Sciences, 65(9), 1892- 1904.

Reid, S. M., Wilson, C. C., Mandrak, N. E., & Carl, L. M. (2008b). Population structure and genetic diversity of black redhorse (Moxostoma duquesnei) in a highly fragmented watershed. Conservation Genetics, 9(3), 531-546.

Rice, W. R. (1989). Analyzing tables of statistical tests. Evolution, 43(1), 223-225. 76

Rousset, F. (2008). GENEPOP’007: A complete re-implementation of the GENEPOP software for Windows and Linux. Molecular Ecology Resources, 8(1), 103-106.

Saitoh, K., Miya, M., Inoue, J. G., Ishiguro, N. B., & Nishida, M. (2003). Mitochondrial genomics of Ostariophysan fishes: Perspectives on phylogeny and biogeography. Journal of Molecular Evolution, 56(4), 464-472.

Sabaj, M. H., Page, L. M., Lundberg, J. G., Ferraris Jr., C. J., Armbruster, J.W . Friel, J. P., & Morris, P. J. (2004). All Catfish Species Inventory Website. Internet address: http://clade.acnatsci.org/allcatfish.

Salgueiro, P., Carvalho, G., Collares-Pereira, M. J., & Coelho, M. M. (2003). Microsatellite analysis of genetic population structure of the endangered cyprinid Anaecypris hispanica in Portugal: Implications for conservation. Biological Conservation, 109(1), 47-56.

Scott, W. B., Crossman, E. J. (1973). Freshwater fishes of Canada. Fisheries research board Canada bulletin, 184. 966 pp.

Seeb, L. W., Antonovich, A., Banks, M. A., Beacham, T. D., Bellinger, M. R., Blankenship, S. M., et al. (2007). Development of a standardized DNA database for Chinook salmon. Fisheries, 32(11), 540-552.

Seutin, G., White, B. N., & Boag, P. T. (1991). Preservation of avian blood and tissue samples for DNA analyses. Canadian Journal of Zoology, 69(1), 82-90.

Silbiger, R. N., Leonard, A. C., Dimsoski, P., Fore, S., Guttman, S. I., Roth, A. C., et al. (2001). Use of molecular markers to study the effects of environmental impacts on genetic diversity in Brown bullhead (Ameirus nebulosus) populations. Environmental Toxicology and Chemistry, 20(11), 2580-2587.

Skalski, G. T., Landis, J. B., Grose, M. J., & Hudman, S. P. (2008). Genetic structure of creek chub, a headwater minnow, in an impounded river system. Transactions of the American Fisheries Society, 137(4), 962-975.

Slatkin, M., & Excoffier, L. (1996). Testing for linkage disequilibrium in genotypic data using the Expectation-Maximization algorithm. Heredity, 76(4), 377.

So, N., Maes, G. E., & Volckaert, F. A. M. (2006). High genetic diversity in cryptic populations of the migratory sutchi catfish Pangasianodon hypophthalmus in the mekong river. Heredity, 96(2), 166-174. 77

Spruell, P., Rieman, B. E., Knudsen, K. L., Utter, F. M., & Allendorf, F. W. (1999). Genetic population structure within streams: Microsatellite analysis of bull trout populations. Ecology of Freshwater Fish, 8(3), 114-121.

Sullivan, J. P., Lundberg, J. G., & Hardman, M. (2006). A phylogenetic analysis of the major groups of catfishes (Teleostei: Siluriformes) using rag1 and rag2 nuclear gene sequences. Molecular Phylogenetics and Evolution, 41(3), 636-662.

Tero, N., Aspi, J., Siikamäki, P., Jäkäläniemi, A., & Tuomi, J. (2003). Genetic structure and gene flow in a metapopulation of an endangered plant species, Silene tatarica. Molecular Ecology, 12(8), 2073-2085.

Thai, B. T., Burridge, C. P., & Austin, C. M. (2007). Genetic diversity of common carp (Cyprinus carpio L.) in Vietnam using four microsatellite loci. Aquaculture, 269(1-4), 174-186.

Triantafyllidis, A., Abatzopoulos, T. J., Leonardos, J., & Guyomard, R. (2002). Microsatellite analysis of the genetic population structure of native and translocated Aristotle's catfish (Silurus aristotelis). Aquatic Living Resources, 15(6), 351-359.

Tzeng, C., Chen, C., Tang, P., & Chiu, T. (2009). Microsatellite and mitochondrial haplotype differentiation in blue mackerel (Scomber australasicus) from the western North Pacific. ICES Journal of Marine Science, 66(5), 816-825.

Usmani, S., Tan, S. G., Siraj, S. S., & Yusoff, K. (2003). Population structure of the Southeast Asian river catfish Mystus nemurus. Animal Genetics, 34(6), 462-464.

Van Oosterhout, C., Hutchinson, W. F., Wills, D. P. M., & Shipley, P. (2004). Micro- checker: Software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes, 4(3), 535-538.

Vazquez-Dominguez, E., Hernandez-Valdes, A., Rojas-Santoyo, A., & Zambrano, L. (2009). Contrasting genetic structure in two codistributed freshwater fish species of highly seasonal systems. Revista Mexicana De Biodiversidad, 80(1), 181-192.

Venturelli, P. P. A., Murphy, C. A., Shuter, B. J., Johnston, T. A., Van Coeverden de Groot, P.J., Boag, P. T., et al. (2010). Maternal influences on population dynamics: Evidence from an exploited freshwater fish. Ecology (Durham), 91(7), 2003-2012.

Wang, J., Shen, T., Ju, J., & Yang, G. (2011a). The complete mitochondrial genome of the Chinese longsnout catfish Leiocassis longirostris (Siluriformes: Bagridae) and a time- calibrated phylogeny of Ostariophysan fishes. Molecular Biology Reports, 38(4), 2507-2516. 78

Wang, W., Bian, W. J., Yu, W. Z. & Shi, Z. D. (2005). Development and behavior ecology of larval and juvenile Pelteobagrus vachelli. Journal of Fisheries of China, 04. doi: cnki:ISSN:1000-0615.0.2005-04-009 (In Chinese with English abstract).

Wang, W., Liu, L. P., Zhang, K. J., Zhang, D. S., Zhao, Q. Z., Zhang, H.G. et al. (2001). Preliminary studies on artificial propogation of Pseudobagrus vachelli. Fisheries Science and Technology Information, 05. doi: cnki:ISSN:1001-1994.0.2001-05-000 (In Chinese with English abstract).

Wang, W., Zhou, X. X., Ma, X. Z. & Li, W. C. (2007). Effect of feeding frequency on the growth and protease activities of Pelteobagrus vachelli. Journal of Shanghai Fisheries University, 03. doi: CNKI:ISSN:1004-7271.0.2007-03-004 (In Chinese with English abstract).

Wang, Y., Yu, X., & Xu, J. (2011b). Decreased trophic position as a function of increasing body size of a benthic omnivorous fish from the largest freshwater lake in China. Environmental Biology of Fishes, 91(4), 505-512.

Wang, Z., Wu, Q., Zhou, J., & Ye, Y. (2004). Geographic distribution of Pelteobagrus fulvidraco and Pelteobagrus vachelli in the Yangtze River based on mitochondrial DNA markers. Biochemical Genetics, 42(11-12), 391-400.

Wang, Z. H., & Xie, X. J. (2003). The chemical compositions and models for predicting energy density in Pelteobagrus vachelli at different seasons. Acta Ecologica Sinica, 01. doi: CNKI:SUN:STXB.0.2003-01-015 (In Chinese with English abstract).

Wang, Z., Zhou, J., Ye, Y., Wei, Q., & Wu, Q. (2006). Genetic structure and low-genetic diversity suggesting the necessity for conservation of the Chinese longsnout catfish, Leiocassis longirostris (Pisces: Bagriidae). Environmental Biology of Fishes, 75(4), 455-463.

Ward, R. D., Woodwark, M., & Skibinski, D. O. F. (1994). A comparison of genetic diversity levels in marine, freshwater, and anadromous fishes. Journal of Fish Biology, 44(2), 213-232.

Ward, R. D., & Grewe, P. M. (1994). Appraisal of molecular genetic techniques in fisheries. Reviews in Fish Biology and Fisheries, 4(3), 300-325.

Weber, J. L., & Wong, C. (1993). Mutation of human short tandem repeats. Human Molecular Genetics, 2(8), 1123-1128.

Weir, B.S. (1996). Genetic Data Analysis II. Sinauer Associates, Sunderland, Massachusetts. 79

Weir, B. S., & Cockerham, C. C. (1984). Estimating F-statistics for the analysis of population structure. Evolution, 38(6), pp. 1358-1370.

Whiteley, A. R., Spruell, P., & Allendorf, F. W. (2006a). Can common species provide valuable information for conservation? Molecular Ecology, 15(10), 2767-2786.

Whiteley, A. R., Spruell, P., Rieman, B. E., & Allendorf, F. W. (2006b). Fine-scale genetic structure of bull trout at the southern limit of their distribution. Transactions of the American Fisheries Society, 135(5), 1238-1253.

Wilson, A. J., Gíslason, D., Skúlason, S., Snorrason, S. S., Adams, C. E., Alexander, G., et al. (2004). Population genetic structure of arctic charr, Salvelinus alpinus from northwest Europe on large and small spatial scales. Molecular Ecology, 13(5), 1129- 1142.

Wirgin, I. I., & Waldman, J. R. (1994). What DNA can do for you. Fisheries, 19(7), 16-27.

Withler, R. E., Rundle, T., & Beacham, T. D. (2007). Genetic identification of wild and domesticated strains of Chinook salmon (Oncorhynchus tshawytscha) in southern British Columbia, Canada. Aquaculture, 272, S161-S171.

Withler, R. E., Candy, J. R., Beacham, T. D., & Miller, K. M. (2004). Forensic DNA analysis of Pacific salmonid samples for species and stock identification. Environmental Biology of Fishes, 69(1-4), 275-285.

Wu, J., Huang, J., & Han, X. (2003). Response to "Three-Gorges Dam: Risk to ancient fish". Science, 302, 1149-1150.

WWF. (2004). Rivers at risk: Dams and the future of freshwater ecosystems.

Xiao, F. (2006) Three Gorges revisited. Chinese National Geographic, 4. Retrieved from http://www.internationalrivers.org/files/attached-files/fanxiao.pdf

Xiao-Gu, Z., Jin-Gou, T., & Bang-Xi, X. (2006). Applications of microsatellite markers in studies of genetics and breeding of fish. Chinese Journal of Agricultural Biotechnology, 3(2), 83-87.

Xiao, T.Y., Sheng, L. Z., Su, J. M., Zhang, X. W., Chen, K. J. & Wang, X. Q. (2002). Study on morphology, growth and reproduction of Pelteobagrus vacleii. Journal of Hunan Agricultural University, 04. doi: cnki:ISSN:1007-1032.0.2002-04-016 (In Chinese with English abstract).

Xie, P. (2003). Three-Gorges Dam: Risk to ancient fish. Science, 302, 1149. 80

Yamamoto, S., Morita, K., Koizumi, I., & Maekawa, K. (2004). Genetic differentiation of white‐spotted charr (Salvelinus leucomaenis) populations after habitat fragmentation: Spatial–temporal changes in gene frequencies. Conservation Genetics, 5(4), 529-538.

Yamazaki, Y., Yamano, A., & Oura, K. (2011). Recent microscale disturbance of gene flow in threatened fluvial lamprey, Lethenteron sp. N, living in a paddy water system. Conservation Genetics, 12(5), 1373-1377.

Yang, D., Kynard, B., Wei, Q., Chen, X., Zheng, W., & Du, H. (2006). Distribution and movement of Chinese sturgeon, Acipenser sinensis, on the spawning ground located below the Gezhouba Dam during spawning seasons. Journal of Applied Ichthyology, 22(s1), 145-151.

Yang, G., Xiao, M., Yu, Y., & Xu, S. (2011). Genetic variation at mtDNA and microsatellite loci in Chinese longsnout catfish (Leiocassis longirostris). Molecular Biology Reports, 1-13

Yang, J. (2006). Studies on the composition and growing characteristics in the reproductive population of Pelteobagrus vachelli. Journal of Chongqing Three Gorges University, 03. doi: cnki:ISSN:1009-8135.0.2006-03-025 (In Chinese with English abstract).

Yonghui, Y., Baiping, Z., Xiaoding, M., & Peng, M. (2006). Large-scale hydroelectric projects and mountain development on the upper Yangtze River. Mountain Research and Development, 26(2), 109-114.

Zeng, W., Li, Z., Ye, S., Xie, S., Liu, J., Zhang, T., et al. (2010). Effects of stocking density on growth and skin color of juvenile darkbarbel catfish Pelteobagrus vachelli (Richardson). Journal of Applied Ichthyology, 26(6), 925-929.

Zhang, F., & Tan, D. (2010). Genetic diversity in population of largemouth bronze gudeon (Coreius guichenoti sauvage et darby) from Yangtze River determined by microsatellite DNA analysis. Genes & Genetic Systems, 85, 351-357.

Zhang, H., Wei, Q. W., Du, H., & Li, L. X. (2011). Present status and risk for extinction of the Dabry’s sturgeon (Acipenser dabryanus) in the Yangtze River watershed: A concern for intensified rehabilitation needs. Journal of Applied Ichthyology, 27(2), 181-185.

Zhang, S., Wang, D., & Zhang, Y. (2003). Mitochondrial DNA variation, effective female population size and population history of the endangered Chinese sturgeon, Acipenser sinensis. Conservation Genetics, 4(6), 673-683. 81

Zhao, J., Zhang, D. D., Zhu, X. P., Zheng, G. M., & Chen, K. C. (2011). Microsatellite variation of mud carp (Cirrhina molitorella Valenciennes, 1844) populations from the Pearl River Basin in China. Journal of Applied Ichthyology, 27(6), 1303-1306.

Zheng, K., Zhu, X., Han, D., Yang, Y., Lei, W., & Xie, S. (2010). Effects of dietary lipid levels on growth, survival and lipid metabolism during early ontogeny of Pelteobagrus vachelli larvae. Aquaculture, 299(1-4), 121-127.

Zhu, Y. P., Cao, Z. D., & Fu, S. J. (2010). Muscle lactate and blood glucose levels of juvenile darkbarbel catfish (Pelteobagrus vachelli Richardson) in response to exhaustive exercise. Journal of Chongqing Normal University (Natural Science), 02. doi: CNKI:SUN:CQSF.0.2010-02-006 (In Chinese with English abstract).

82

Appendix 1: a)

0.30 Pvi 004

0.20 Songji Luzhou

0.10 Yibin Frequency

0.00

174 178 184 188 192 198 202 208 212 216 162 176 180 182 186 190 194 196 200 204 206 210 214 220 240 Allele Size b) 0.25 Pvi 026 0.20 Songji 0.15 Luzhou 0.10 Yibin Frequency 0.05

0.00

235 243 245 251 257 265 271 273 277 279 241 247 249 253 255 261 263 267 269 275 283 Allele Size c) 0.25

0.20 Pvi 038 Songji 0.15 Luzhou 0.10

Frequency 0.05 Yibin

0.00

181 183 195 197 199 201 211 213 215 217 229 231 233 235 185 187 189 191 193 203 205 207 209 219 221 223 225 227 247 Allele Size d) 0.400 Pvi 045 0.300 Songji 0.200 Luzhou Yibin Frequency 0.100

0.000

212 214 216 218 220 222 224 226 228 232 234 236 240 254 Allele Size

83 e) 0.30

Pvi 052 Songji 0.20 Luzhou

0.10 Yibin Frequency

0.00

232 236 246 248 258 260 262 276 278 280 238 240 242 244 250 252 254 256 264 266 270 272 282 Allele Size f)

0.50 Pvi 055 0.40 Songji 0.30 Luzhou 0.20 Yibin Frequency 0.10

0.00

105 109 113 119 123 127 101 103 107 111 115 117 121 125 129 Allele Size g) 0.25

0.20 Pvi 066 Songji 0.15 Luzhou 0.10

Frequency 0.05 Yibin

0.00

324 328 332 336 340 344 348 352 356 364 370 322 326 330 334 338 342 346 350 354 360 362 368 Allele Size h) 0.60

0.50 Pvi 114

0.40 Songji 0.30 Luzhou 0.20 Frequency Yibin 0.10

0.00

184 188 190 192 220 222 224 258 178 180 182 194 208 212 216 218 226 230 234 240 252 Allele Size

84

Frequency of allele sizes for each locus within each of the populations where Population 5 represents Song Ji, population 6 represents Luzhou, and population 7 represents Yibin: a) Pvi 004, b) Pvi 026, c) Pvi 038, d) Pvi 045, e) Pvi 052, f) Pvi 055, g) Pvi 066, h) Pvi 114.

85

Appendix 2:

Principle Co-ordinates

0.6

0.4

0.2 Co-ord. 3 Co-ord. 0.0

0.8 -0.2 0.6 0.4 0.2 -0.4 0.0 0.6 0.4 -0.2 Co-ord. 1 Songji 0.2 -0.4 0.0 Luzhou Co-ord. 2 -0.2 -0.6 Yibin -0.4

3 Dimensional scatter plot of the 61 P. vachelli samples along the first three axes of the principal components analysis where ● represents the individuals from Songji, ○ Luzhou, and Yibin. 86

Appendix 3: Number of microsatellite loci studied and ranges in number of alleles per locus of various species in riverine and marine environments Taxa Common name Species # of Loci Range in # of alleles Reference Marine Blue mackerel Scomber australasicus 7 19 - 45 Tzeng et al., 2009 Chub makerel Scomber japonicus 7 25 - 44 Tzeng et al., 2009 Korean rockfish Sebastes schlegeli 8 3 - 33 An et al., 2012

Anadromous Chinook salmon Oncorhynchus tshawytscha 13 9 - 74 Seeb et al., 2007 Eulachon Thaleichthys pacificus 5 3 – 10 McLean and Taylor, 2001 Marble trout Salmo marmoratus 14 1 - 12 Fumagalli et al., 2002 White-spotted char Salvelinus leucomaenis 5 2 - 24 Yamamoto et al., 2004 European grayling Thymallus thymallus 8 4 - 10 Meldgaard et al., 2003 Steelhead Oncorhynchus mykiss 24 3 – 49 Clemento et al., 2008

Potadromodous Curimbata-pioa Prochilodus costatus 6 2 – 16 Carvalho-Costa et al., 2008 Saramugo Anaecypris hispanica 5 1 – 38 Salgueiro et al., 2003 Bull trout Salvelinus confluentus 7 1 – 4 Spruell et al., 1999 Bull trout Salvelinus confluentus 7 2 – 17 Costello et al., 2003 Mud carp Cirrhina molitorella 11 4 – 18 Zhao et al., 2011 Common carp Cyprinus carpio L. 4 10 – 23 Thai et al., 2007 Chub Squalius cephalus 5 8 – 14 Dehais et al., 2010 Black redhorse Moxostoma duquesnei 8 12 – 33 Reid et al., 2008b Mountain whitefish Prosopium williamsoni 6 7 – 61 Whiteley et al., 2006a Bull trout Salvelinus confluentus 8 2 – 8 Neraas and Spruell, 2001 Brown trout Salmo trutta 9 2 – 27 Heggenes and Roed, 2006 Greenside darter Etheostoma blennioides 9 7 – 70 Beneteau et al., 2009 Creek chub Semotilus atromaculatus 19 1 – 14 Skalski et al., 2008 Largemouth bronze Coreius guichenoti 11 5 – 19 Zhang and Tan, 2010 gudgeon

Non-migratory Brown trout Salmo trutta 5 2 – 5 Carlsson et al., 1999 Northern Sheatfish Silurus soldatovi 24 3 – 23 Quan et al., 2006 Mangrove Molly Poecilia orri 3 3 – 15 Vazquez-Dominguez et al., 2009 Mosquito fish Gambusia yucatana 3 6 – 18 Vazquez-Dominguez et al., 2009 Yellow catfish Horabagrus nigricollaris 5 1 – 8 Abdul Muneer et al., 2011 Yellow catfish Horabagrus brachysoma 5 1 – 8 Abdul Muneer et al., 2011 Aristotle catfish Silurus aristotelis 8 18 – 24 Triantafyllidis et al., 2002 Longsnout catfish Leiocassis longirostris 20 6 – 29 Yang et al., 2011 Southeast Asian river catfish Mystus nemurus 20 2 – 11 Usmani et al., 2003 87

Appendix 4: Summary table of observed and expected heterozygosity per population in fish species with varying migratory patterns

Taxa Common name Species HO HE # of loci Reference Marine Korean rockfish Sebastes schlegeli 0.737 – 0.806 0.794-0.821 8 An et al., 2012

Anadromous Eulachon Thaleichthys pacificus mean 0.28 mean 0.29 5 McLean and Taylor, 2001 Marble trout Salmo marmoratus 0.046 – 0.233 0.039 – 0.219 14 Fumagalli et al., 2002 White-spotted char Salvelinus leucomaenis 0.29 – 0.69 0.35 – 0.71 5 Yamamoto et al., 2004 Steelhead Oncorhynchus mykiss 0.573 – 0.712 0.597 – 0.723 24 Clemento et al., 2008

Potadromodous Curimbata-pioa Prochilodus costatus 0.420 – 0.480 0.650 – 0.660 6 Carvalho-Costa et al., 2008 Greenside darter Etheostoma lennioides 0.56 -0.74 0.62 – 0.74 9 Beneteau et al., 2009 Saramugo Anaecypris hispanica 0.44 - 0.79 0.59 - 0.78 5 Salgueiro et al., 2003 Bull trout Salvelinus confluentus ----- 0.330 – 0.426 7 Spruell et al., 1999 Bull trout Salvelinus confluentus 0.00-0.78 0.05-0.37 7 Costello et al., 2003 Sutchi catfish Pangasianodon hypophthalmus mean 0.734 mean 0.757 7 So et al., 2006 Mud carp Cirrhina molitorella 0.5105 – 0.6182 0.7081 – 0.7727 11 Zhao et al., 2011 River redhorse Moxostoma carinatum 0.61 – 0.80 ----- 11 Reid et al., 2008a Shorthead redhorse Moxostoma macrolepidotum 0.68 – 0.82 ----- 11 Reid et al., 2008a Black redhorse Moxostoma duquesnei 0.59 – 0.67 0.71 – 0.75 8 Reid et al., 2008b Brown trout Salmo trutta 0.57 – 0.65 0.58 – 0.66 9 Heggenes and Roed, 2006 Creek chub Semotilus atromaculatus 0.424 – 0.566 0.417 – 0.572 19 Skalski et al., 2008 Common carp Cyprinus carpio L. 0.40 - 0.82 0.53 - 0.83 4 Thai et al., 2007 Mountain whitefish Prosopium williamsoni ----- 0.000 – 0.538 6 Whiteley et al., 2006a Largemouth bronze Coreius guichenoti 0.8182 – 0.8700 0.8047 - 0.8339 11 Zhang and Tan, 2010 gudgeon

Non-migratory Northern sheatfish Silurus soldatovi mean 0.435 mean 0.758 24 Quan et al., 2006 Rhone streber Zingel asper 0.68 – 0.92 0.69 – 0.86 5 Laroche and Durand, 2004 Mangrove Molly Poecilia orri 0.000 – 0.222 0.000 – 0.814 3 Vazquez-Dominguez et al., 2009 Mosquito fish Gambusia yucatana 0.159 – 0.600 0.422 – 0.800 3 Vazquez-Dominguez et al., 2009 Aristotle catfish Silurus aristotelis 0.532 – 0.703 0.562 – 0.734 8 Triantafyllidis et al., 2002 Longsnout catfish Leiocassis longirostris 0.8205 – 0.8697 0.7757 – 0.8300 20 Yang et al., 2011 Southeast Asian river Mystus nemurus 0.4418 – 0.5730 0.4350 – 0.5410 20 Usmani et al., 2003 catfish