Herein we tested the repeatability of RAD-seq phylogeographic construction by creating a second, independent phylogeography of the pitcher-plant mosquito, Wyeomyia smithii. We sampled 25 populations drawn from different localities nearby previous collection sites and used these new data to construct a second, independent phylogeography to test the reproducibility of phylogenetic patterns. Our previous phylogeography was based on 3,741 phylogenetically informative markers from 21 populations and rooted with mitochondrial COI. The present phylogeography was based on 16,858 informative markers and rooted with RAD-seq. We found correspondence between clades at the extremes of W. smithii's distribution; however, there were several discrepancies between the trees, including the refugium that gave rise to all post-glacial populations. We observed that combining all 46 populations resolved these discrepancies and, equally importantly, that extensive taxon sampling in areas of historical importance is more valuable than increasing the number of informative sites in establishing an accurate, robust phylogeography. This thesis includes unpublished co-authored material.

The upland (Pseudacris feriarum) and midland (P. triseriata) chorus frogs are closely related cryptic species that are best distinguished genetically. The distribution of these species within the Commonwealth of Kentucky has previously been defined by only a handful of genetic samples, making delineation of range limits for each species difficult. Accurate understanding of species distributions, and the genetic structure within them, are vitally important for conservation management of amphibian species. In this study, I have collected genetic samples from across the putative ranges of P. triseriata and P. feriarum in Kentucky and used next-generation sequencing technology to generate more fine-scale estimates of species ranges. The genetic data generated in this study support the delineation of two species in Kentucky, and the species assignments of all individuals and populations are in general concordance with the previously hypothesized species distributions. However, I have identified two previously unrecognized contact zones for these species and revealed areas of hybridization. By delineating species distributions and identifying potentially important regions of genetic admixture, this study will be informative to future conservation management and conservation genetic research of chorus frogs in Kentucky.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第22286号 / 理博第4600号 / 新制||理||1660(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 工藤 洋, 教授 田村 実, 准教授 高山 浩司 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

Phylogenetics is crucial in the study of evolutionary processes and the determination of appropriate conservation units, and often these efforts are complicated by hybridization and introgression. Aliciella subsection Subnuda consists of seven species of herbaceous plants occurring in Utah and the Four Corners region of North America. Previous molecular and morphological work left relationships in the subsection unresolved. Here, we use comparative DNA sequencing of ITS and cpDNA regions and RAD-seq data to clarify phylogenetic relationships and examine the role of hybridization in the subsection. We construct haplotype and nucleotype networks from the cpDNA and ITS sequence matrices and compare nuclear and chloroplast phylogenies to identify multiple putative chloroplast capture events. The RAD-seq maximum likelihood phylogeny robustly resolves relationships between six clades, supportive of merging of two species. We employ STRUCTURE and HyDe on the RAD-seq data to evaluate the influence of hybridization within the subsection. The HyDe results provide evidence of hybridization among and between all species in the subsection. Our study robustly resolves relationships in Aliciella subsection Subnuda and provides a framework for discussing its speciation despite a history of hybridization and introgression.

La génomique des populations rend possible la mise en évidence de traces de sélection dans le génome. Les travaux effectués considèrent en général une échelle de temps longue (~ 10³ générations). En comparaison, peu d’intérêt a été porté aux études expérimentales de court terme (~ 10 générations). De telles expériences sont pourtant susceptibles de nous renseigner sur la base génétique de caractères complexes. Nous proposons une méthode de vraisemblance basée sur un modèle de Wright-Fisher pour détecter la sélection à partir d’échantillons génétiques temporels acquis sur une période de dix générations. Nous montrons par simulation que notre méthode permet de différencier les signaux dus à la combinaison de la sélection et de la dérive génétique de ceux dus à la dérive seule. Nous montrons également par simulation qu’il est possible d’estimer le coefficient de sélection appliqué à un locus testé. De plus, nous illustrons l’intérêt de notre méthode pour la détection de marqueurs candidats à la sélection au travers de deux études génomiques sur données réelles, chez le diable de Tasmanie (Sarcophilus harrisii) et chez la truite arc-en-ciel (Oncorhynchus mykiss). Ces applications mettent en évidence des régions génomiques candidates pour des phénotypes complexes dans des contextes différents. Dans l’ensemble, nos résultats montrent qu’il est possible de détecter des gènes sujets à une sélection directionnelle intense à partir d’échantillons génétiques temporels, même si la sélection est de courte durée et si les populations examinées ont un faible effectif. / Population genomics makes it possible to detect traces of selection in the genome. Studies in this field have mainly focused on long time scale (~ 10³ generations). In comparison, short-term experimental studies (~ 10 generations) have attracted much less interest. Such experiments are, however, likely to inform us about the genetic basis of complex characters. We propose a likelihood method based on a Wright-Fisher model to detect selection from genetic temporal samples collected over ten generations. We show through simulation that our method can disentangle signals due to the combination of genetic drift and selection to those due to drift alone. We also show through simulation that it is possible to estimate the selection coefficient applied to a tested locus. In addition, we illustrate the interest of our method for the detection of candidate markers for selection through two genome scans performed on real data, in the Tasmanian devil (Sarcophilus harrisii) and in the rainbow trout (Oncorhynchus mykiss). These practical applications highlight candidate genomic regions for complex phenotypes in different contexts. Collectively, our results show the possibility of detecting genes submitted to strong directional selection from genetic time-series, even if selection is applied on a short time period and if the examined populations are small.

Caligid copepods, also called sea lice, are ectoparasites of marine fish, with Lepeophtheirus salmonis (Krøyer, 1837) emerging as a problem for mariculture of Atlantic salmon (Salmo salar Linnaeus, 1758) in the northern hemisphere. Annual costs of sea lice to global salmon farming was estimated to be in excess of €300 million in 2006, with the majority of this accounted for through expenses accrued from chemical treatments. Only a limited range of anti-sea louse drugs are available and licensed for the treatment of fish, and the continued use of only a few compounds creates a situation potentially favouring the development of drug resistance. Emamectin benzoate (EMB) is currently used as a salmon delousing agent, being employed as a 0.2 % in-feed pre-mix (SLICE®). Atlantic salmon farmers have reported increased incidence of reduced L. salmonis sensitivity to SLICE®, which has highlighted the requirement for further research into the molecular mechanisms controlling salmon louse resistance to EMB. Genomic and transcriptomic research concerning L. salmonis drug resistance mechanisms has not often been reported, with previous transcriptomic studies using candidate gene approaches and genetic studies focussing on population genetics. Drug resistance in ecdysozoan invertebrates is associated with a variety of molecular mechanisms including target site mutations and changes in the expression of components in drug detoxification pathways. The research reported in this thesis was aimed at the exploration of mechanisms employed by L. salmonis to reduce the toxicity of EMB exposure, following a transcriptomic approach that utilised custom oligonucleotide (oligo) microarrays and a genetic approach that utilised Restriction-site associated DNA sequencing (RAD-seq) to identify Single Nucleotide Polymorphism (SNP) markers. An EMB-resistant (PT) and drug-susceptible (S) L. salmonis laboratory-maintained strain were to be used as a model for this research, as these two strains differ in EMB susceptibility (~ 7-fold) and show stable susceptibility profiles through multiple generations, suggesting that this drug resistance phenotype may be a heritable trait. Sequence resources available for salmon lice are limited as an annotated L. salmonis genome is currently under construction. Therefore, a significant amount of this study involved creating new resources to facilitate the analysis of EMB susceptibility. Suppression subtractive hybridisation (SSH) was used to enrich for transcripts that were differentially expressed between strains PT and S, which provided sufficient target sequence for the development of 15K oligo microarrays when combined with sequences assembled from existing L. salmonis ESTs. Additionally, transcripts were generated through sequencing a pooled sample representing key developmental stages of the L. salmonis life cycle, which were later used in the construction of a 44K oligo microarray. The toxicity of EMB and other avermectins (AVMs) against ecdysozoan invertebrates is reported to be based mainly on their interaction with ligand-gated ion channels (LGIC), specifically glutamate-gated chloride channels (GluCl). However, -aminobutyric acid (GABA)-gated chloride channels (GABA-Cls) are also believed to be targeted by AVMs and neuronal acetylcholine receptors (nAChRs) can be allosterically modulated by the AVM compound ivermectin. Transcriptional responses in PT and S salmon lice were investigated using custom 15K L. salmonis oligo microarrays. In the absence of EMB exposure, 359 targets differed in transcript abundance between the two strains. GABA-Cl and nAChR subunits showed significantly lower transcript levels in PT compared to S lice, which was estimated at ~1.4-fold for GABA-Cl and ~2.8-fold for nAChR using RT-qPCR, suggesting their involvement in AVM toxicity in caligids. Although, salmon lice from the PT strain showed few transcriptional responses following acute exposure (1 or 3 h) to 200 µg L-1 of EMB, a drug concentration tolerated by PT lice, but toxic for S lice. RAD-seq analysis of both genders from L. salmonis strains S and PT identified 15 RAD-markers that show complete association with salmon louse strain, although these preliminary results will need further analysis to confirm marker association with reduced EMB susceptibility. Additionally, RAD marker Lsa101901 showed complete association with sex for all individuals analysed, being heterozygous in females and homozygous in males. Using an allele-specific PCR assay, this SNP association pattern was further confirmed for three unrelated salmon louse strains. Marker Lsa101901 was located in the coding region of the prohibitin-2 gene, which showed a sex-dependent differential expression, with mRNA levels determined by RT-qPCR about 1.8-fold higher in adult female than adult male salmon lice. In conclusion, the identification of decreased transcript abundances for LGIC subunits in EMB-resistant salmon lice, and polymorphic SNP markers showing complete association with L. salmonis strains S or PT, provides suitable candidates for further investigation into their association with reduced EMB susceptibility. Further analysis will also be required to confirm whether EMB-induced mechanisms are not associated with reduced EMB susceptibility in L. salmonis. Additionally, the identification of sex-linked SNP Lsa101901 suggests that sex determination in the salmon louse is genetic and follows a female heterozygous system, with marker Lsa101901 providing a tool to determine the genetic sex of salmon lice. Improved knowledge of L. salmonis biology and the mechanisms potentially involved in EMB resistance, obtained during this study, may provide molecular markers that contribute to successful monitoring and management of this commercially important parasite of Atlantic salmon.

The aim of the current thesis was the analysis of the genetics of sex determination of farmed fish with sexual dimorphism, using Next Generation Sequencing. Three different species of farmed fish with sex-determining systems of varying complexity were studied. Both full-sibs and more distantly related specimens of Atlantic halibut (Hippoglossus hippoglossus), Nile tilapia (Oreochromis niloticus) and European sea bass (Dicentrarchus labrax) were used for this study. Application of Restriction-site Associated DNA sequencing (RAD-seq) and double digest Restriction-site Associated DNA sequencing (ddRAD-seq), two related techniques based on next generation sequencing, allowed the identification of thousands of Single Nucleotide Polymorphisms (SNPs; > 3,000) for each of the above species. The first SNP-based genetic maps for the above species were constructed during the current study. The first evidence concerning the location of the sex-determining region of Atlantic halibut is provided in this study. In the case of Nile tilapia both novel sex-determining regions and fine mapping of the major sex-determining region are presented. In the study of European sea bass evidence concerning the absence of a major sex-determining gene was provided. Indications of putative sex-determining regions in this species are also provided. The results of the current thesis help to broaden current knowledge concerning sex determination in three important farmed fish. In addition the results of the current thesis have practical applications as well, towards the production of mono-sex stocks of those species for the aquaculture industry.

京都大学 / 新制・課程博士 / 博士(農学) / 甲第24658号 / 農博第2541号 / 新制||農||1098(附属図書館) / 学位論文||R5||N5439(農学部図書室) / 京都大学大学院農学研究科森林科学専攻 / (主査)教授 井鷺 裕司, 教授 北島 薫, 教授 柴田 昌三 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DGAM

Tesis por compendio / [ES] Un objetivo fundamental en los programas de conservación es mantener la diversidad genética y la estrategia de gestión más eficiente para lograrlo es aplicar el método de Contribuciones Óptimas. Este método optimiza las contribuciones de los candidatos a reproductores minimizando el parentesco global ponderada, lo que conduce a niveles más altos de diversidad genética, medida como heterocigosis esperada, y a un control efectivo del aumento de consanguinidad. El parámetro fundamental de este método es la matriz de parentesco. Esta matriz se ha obtenido tradicionalmente a partir del pedigrí, pero la disponibilidad actual de genotipos para un gran número de polimorfismos de un solo nucleótido (SNP) nos permite estimarla con una mayor precisión. Sin embargo, se han propuesto muchas medidas de parentesco genómico y se desconoce qué medida es la más apropiada para minimizar la pérdida de diversidad genética. Por lo tanto, el objetivo general de esta tesis fue investigar la eficiencia de diferentes matrices genómicas de parentesco en la gestión de poblaciones en programas de conservación, cuando se aplica el método de Contribuciones Óptimas. Las matrices comparadas fueron aquellas basadas en: i) la proporción de alelos compartidos por dos individuos (CSIM); ii) las desviaciones del número observado de alelos compartidos por dos individuos respecto del número esperado (CL&H); iii) la matriz de relaciones genómicas obtenida a través el método 1 de VanRaden (CVR1); iv) la matriz de relaciones genómicas obtenida a través el método 2 de VanRaden (CVR2); v) la matriz de relaciones genómicas obtenida a través el método de Yang (CYAN); y vi) segmentos idénticos por descendencia (CSEG). Los resultados para una sola generación, usando miles de SNP genotipados en individuos de una población cultivada de rodaballo, mostraron grandes diferencias en la magnitud de los seis coeficientes de parentesco. Las correlaciones entre los diferentes coeficientes variaron mucho, siendo las más bajas aquellas entre CSIM, CL&H o CSEG y CVR2 o CYAN. La gestión que utiliza matrices basadas en la proporción de alelos o segmentos compartidos (CSIM, CL&H y CSEG) retuvieron una mayor diversidad que aquella que utiliza matrices de relaciones genómicas (CVR1, CVR2 y CYAN). Como era de esperar, maximizar la heterocigosis llevó los alelos hacia frecuencias intermedias. Sin embargo, alejar las frecuencias alélicas de las frecuencias iniciales puede ser indeseable, ya que se pueden perder adaptaciones particulares al medio. Se utilizaron simulaciones estocásticas para investigar la eficiencia de CL&H y CVR2 en el manejo de poblaciones no divididas a lo largo de 50 generaciones y ambas matrices se compararon tanto en términos de la diversidad genética como en términos de los cambios asociados en las frecuencias alélicas. EL uso de CL&H en el método de Contribuciones Óptimas llevó a una mayor diversidad genética pero también en un mayor cambio de frecuencias alélicas que el uso de CVR2. Las diferencias entre estrategias fueron menores cuando sólo se usaron SNP con una frecuencia del alelo menos común (MAF) por encima de un umbral particular (MAF > 0.05 y MAF > 0.25) para calcular CL&H y CVR2 así como cuando se aplicó el método de Contribuciones Óptimas en poblaciones con censo más pequeños (se pasó de N = 100 a N = 20). La evaluación de CL&H y CVR2 se extendió a poblaciones subdivididas, también a través de simulaciones por ordenador. En poblaciones subdivididas, la diversidad genética se distribuye en dos componentes: dentro y entre subpoblaciones. Cuando se otorga un mayor peso al componente dentro de subpoblaciones, es posible restringir los niveles de consanguinidad. Bajo este escenario, la utilización de CL&H resultó ser la mejor opción para gestionar este tipo de poblaciones, ya que mantuvo una mayor diversidad global, condujo a una menor consanguinidad y a cambios en las frecuencias similares a los observados cuando se utilizó CVR2. / [CA] Un objectiu fonamental en els programes de conservació és mantindré la diversitat genètica i l'estratègia de gestió més eficient per aconseguir-ho és aplicar el mètode de contribucions òptimes. Aquest mètode optimitza les contribucions dels candidats a reproductors minimitzant el parentiu global ponderat, cosa que condueix a nivells més alts de diversitat genètica, mesurada com a heterozigosi esperada, i a un control efectiu de l'augment de consanguinitat. El paràmetre fonamental d'aquest mètode és la matriu de parentiu. Aquesta matriu s'ha obtingut tradicionalment a partir del pedigrí, però la disponibilitat actual de genotips per a un gran nombre de polimorfismes d'un sol nucleòtid (SNP) ens permet estimar-la amb més precisió. Tot i això, s'han proposat moltes mesures de parentiu genòmic i es desconeix quina mesura és la més apropiada per minimitzar la pèrdua de diversitat genètica. Per tant, l'objectiu general d'aquesta tesi va ser investigar l'eficiència de diferents matrius genòmiques de parentiu en la gestió de poblacions en programes de conservació, quan s'aplica el mètode de Contribucions Òptimes. Les matrius comparades van ser aquelles basades en: i) la proporció d'al·lels compartits per dos individus (CSIM); ii) les desviacions del nombre observat d'al·lels compartits per dos individus respecte del nombre esperat (CL&H); iii) la matriu de relacions genòmiques obtinguda a través del mètode 1 de VanRaden (CVR1); iv) la matriu de relacions genòmiques obtinguda a través del mètode 2 de VanRaden (CVR2); v) la matriu de relacions genòmiques obtinguda a través del mètode de Yang (CYAN); i vi) segments idèntics per descendència (CSEG). Els resultats per a una sola generació, usant milers d'SNP genotipats en individus d'una població cultivada de turbot, van mostrar grans diferències en la magnitud dels sis coeficients de parentiu. Les correlacions entre els coeficients van variar molt, sent les més baixes aquelles entre CSIM, CL&H o CSEG i CVR2 o CYAN. La gestió que utilitza matrius basades en la proporció d'al·lels o segments compartits (CSIM, CL&H i CSEG) van retindré una diversitat més gran que aquella que utilitza matrius de relacions genòmiques (CVR1, CVR2 i CYAN). Com calia esperar, la màxima heterozigosi va portar els al·lels cap a freqüències intermèdies. No obstant això, allunyar les freqüències al·lèliques de les freqüències inicials pot ser indesitjable, ja que es poden perdre adaptacions particulars al medi. Es van fer servir simulacions estocàstiques per investigar l'eficiència de CL&H i CVR2 en el maneig de poblacions no dividides al llarg de 50 generacions i les dues matrius es van comparar tant en termes de la diversitat genètica com en termes dels canvis associats a les freqüències al·lèliques. L'ús de CL&H en el mètode de Contribucions Òptimes va resultar en una major diversitat genètica però també en un canvi més gran de freqüències al·lèliques que l'ús de CVR2. Les diferències entre estratègies van ser menors quan només es van fer servir SNP amb una freqüència de l'al·lel menys comú (MAF) per sobre d'un llindar particular (MAF > 0.05 i MAF > 0.25) per calcular CL&H i CVR2 així com quan es va aplicar el mètode de Contribucions Òptimes en poblacions amb cens més petit (es va passar de N = 100 a N = 20). L'avaluació de CL&H i CVR2 es va estendre a poblacions subdividides, també a través de simulacions per ordinador. En poblacions subdividides, la diversitat genètica es distribueix en dos components: dins i entre subpoblacions. Quan s'atorga un pes més gran al component dins de subpoblacions, és possible restringir els nivells de consanguinitat. Sota aquest escenari, la utilització de CL&H va resultar ser la millor opció per gestionar aquest tipus de poblacions, ja que va mantindré una diversitat global més gran, va conduir a una menor consanguinitat i a canvis en les freqüències similars als observats quan es va utilitzar CVR2. / [EN] A main objective in conservation programs is to maintain genetic diversity, and the most efficient management strategy to achieve it is to apply the Optimal Contributions method. This method optimizes the contributions of breeding candidates by minimizing the global weighted coancestry. This leads to the highest levels of genetic diversity, when measured as expected heterozygosity, and to an effective control of the increase of inbreeding. The fundamental parameter of the method is the coancestry matrix which, traditionally, has been obtained from pedigree data. The current availability of genome-wide information allows us to estimate coancestries with higher precision. However, many different genomic coancestry measures have been proposed and it is unknown which measure is more efficient to minimize the loss of genetic diversity. Thus, the general aim of this thesis was to investigate the efficiency of different genomic coancestry matrices in the management of conserved populations when the Optimal Contributions method is applied to maximize genetic diversity. The matrices compared were those based on: i) the proportion of shared alleles (CSIM); ii) deviations of the observed number of alleles shared by two individuals from the expected number (CL&H); iii) the realized relationship matrix obtained by VanRaden's method 1 (CVR1); iv) the realized relationship matrix obtained by VanRaden's method 2 (CVR2); v) the realized relationship matrix obtained by Yang¿s method (CYAN); and vi) identical by descent segments (CSEG). Results for a single generation using thousands of SNP genotyped in individuals from a farm turbot population, showed large differences in the magnitude of the six coancestry coefficients. Moreover, pairwise correlations were those between coefficients greatly varied (especially for self-coancestry). The lowest correlations between CSIM, CL&H or CSEG and CVR2 or CYAN. Management with matrices based on the proportion of shared alleles or on segments (CSIM, CL&H and CSEG) retained higher variability than those based on realized genomic relationship matrices (CVR1, CVR2 and CYAN). As expected, maximizing heterozygosity pushed alleles toward intermediate frequencies. However, moving allele frequencies away from initial frequencies may be undesirable as particular adaptations to the environment can be lost. Stochastic simulations were used to investigate the efficiency of CL&H and CVR2 in the management of an undivided population across 50 generations and both matrices were compared not only in terms of the genetic diversity maintained but also in terms of the associated changes in allele frequencies across generations. The use of CL&H in the Optimal Contribution method resulted in a higher genetic diversity but also in a higher change of allele frequencies than the use of CVR2. The differences between strategies were reduced when only SNPs with a minimum allele frequency (MAF) above a particular threshold (MAF > 0.05 and MAF > 0.25) were used to compute CL&H and CVR2 as well as when the Optimal Contributions method was applied in populations of smaller sizes (N = 20 vs N = 100). The evaluation of CL&H and CVR2 was extended to subdivided populations, also via computer simulations. When populations are subdivided into different breeding groups, it is possible to give different weights to the within- and between-subpopulation components of genetic diversity. When a higher weight is given to the within-subpopulation component, the levels of inbreeding can be restricted. In this scenario, the use of CL&H was the best option for managing subdivided populations as it maintained more global diversity, led to less inbreeding and to changes in frequencies similar to those observed when using CVR2 when a large weight was given to the within-subpopulation term. / Esta tesis doctoral se realizó en el Instituto de Investigación y Tecnología Agraria y Alimentaria (INIA-CSIC) de Madrid. El trabajo expuesto en el capítulo 2 se realizó en parte en el Instituto Roslin, de la Universidad de Edimburgo durante una estancia predoctoral. Los trabajos expuestos en la tesis han sido financiados con una beca predoctoral FPI (BES-2017-081070) y a través de proyectos del Plan Estatal de I+D+i del Ministerio de Ciencia e Innovación (proyectos CGL2016-75904-C2-2-P y PID2020- 114426GB-C22) y de la Unión Europea (‘European Union‘s Seventh Framework Program, KBBE.2013.1.2-10, under Grant Agreement No. 613611, FISHBOOST project’ y ‘European Commission Horizon 2020, Framework Programme through Grant Agreement No. 727315, MedAID project’). / Morales González, E. (2023). Management of Genetic Diversity in Conservation Programs Using Genomic Coancestry [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/195231 / Compendio

The Atlantic cod (Gadus morhua L.) fishery is of great economic importance to the Faroese economy. There are two separately managed cod stocks around the Faroe Islands, the Faroe Plateau and the Faroe Bank cod. Both have experienced dramatic decreases in size and informed management decisions are vital for both stock viability and exploitation. The stocks are geographically isolated by an 800 m deep channel and water temperatures are on average 1 – 2 ºC higher on the Faroe Bank than on the Faroe Plateau. There are clear phenotypic differences between the stocks; in particular, the markedly higher growth rate for the Faroe Bank cod has caught public and scientific attention. There is continuing debate regarding the relative importance of genetics and environmental contributions to the contrasting phenotypes. Analyses of reproductive parameters (field data and experimental captive spawnings) as well as analyses of microsatellite and single nucleotide polymorphism (SNP) markers were undertaken to better resolve the issue. Field data as well as data from experimental captive spawnings provided evidence of reproductive differences between Faroe Plateau and Faroe Bank cod. Peak spawning occurred earlier on the Faroe Plateau than on the Faroe Bank and this difference in timing of spawning was maintained in captivity. In particular, differences in sizes of eggs (average diameters of 1.40 and 1.30 mm for Faroe Plateau and Faroe Bank cod eggs, respectively) and indirect evidence of greater volumes spawned by the Faroe Bank females suggested stock differences with respect to egg size – egg number trade-off. It was hypothesised that the strategy adopted by cod on the Faroe Bank, with a higher number of smaller eggs, evolved in response to a more hostile environment (bare seabed and higher exposure to predators) experienced by early life stages in this area. Experimental captive spawnings with Faroe Bank cod showed a large interfamily skew in survival rates of cod eggs and fry. Egg size was identified as a useful indicator of survival rates in the egg stage, but egg survival rates could not be used to predict viability in later developmental stages, thus highlighting the importance of employing some sort of genetic monitoring of cod fry to ensure sufficient family representation in the progeny. While no tank effect was evident concerning fry survival, a significant tank effect was identified concerning body sizes of fry. Microsatellite data were analysed using large sample sizes of Faroe Plateau and Faroe Bank cod with the Faroe Plateau divided into two locations, Faroe Plateau North-East and Faroe Plateau West (cod from each of the two were known to belong to separate spawning grounds). Two Norwegian coastal cod samples were included as outlier populations. While no genetic differentiation was detected between the two Faroe Plateau locations, these analyses revealed a detectable, albeit relatively modest, degree of genetic differentiation between cod from the Faroe Plateau and the Faroe Bank (FST = 0.0014 and 0.0018; DJost_EST = 0.0027 and 0.0048; P < 0.0001 and P < 0.001 for the Faroe Plateau North-East – Faroe Bank and the Faroe Plateau West – Faroe Bank comparisons). These values were several times smaller than those between Faroese and Norwegian coastal cod (pairwise FST and DJost_EST values in the range of 0.0061 – 0.0137 and 0.0158 – 0.0386, respectively). Despite recent reductions in census population sizes for Faroe Plateau and, particularly, Faroe Bank cod, genetic diversity estimates were comparable to the ones observed for Norwegian coastal cod and there was no evidence of significant genetic bottlenecks. Lastly, data for one of the markers (Gmo132) indicated genotype-dependent vertical distribution of cod (as investigated for Faroe Plateau North-East cod). Contrary to some previously published studies, analysis of SNPs of two candidate genes for adaptive divergence, the hemoglobin gene Hb-ß1 and the transferrin gene Tf1, failed to detect differentiation between samples of Faroe Plateau and Faroe Bank cod analysed in this thesis. Of 3533 novel SNPs simultaneously discovered and genotyped by restriction-site associated DNA (RAD) sequencing, 58 showed evidence of genetic differentiation between Faroe Plateau North-East and Faroe Bank cod (P < 0.05). No single locus was fixed for different alleles between Faroe Plateau and Faroe Bank cod. A set of eight informative SNPs (FST values between Faroe Plateau and Faroe Bank samples > 0.25; P < 0.0005) were selected for validation in larger samples, that included cod from both Faroe Plateau areas and the Faroe Bank as well as Norwegian coastal and White Sea cod. Six out of the eight loci amplified successfully with a PCR-based method and there was 100 % concordance between genotypes of individuals screened by both techniques. Due to ascertainment bias, the SNPs should only be applied with caution in a broader geographical context. Nonetheless, these SNPs did confirm the genetic substructure suggested for Faroese cod by microsatellite analyses. While no genetic differentiation was evident between the two Faroe Plateau locations, significant genetic differentiation was evident between Faroe Plateau and Faroe Bank cod at five of the SNPs (FST values in the range of 0.0383 – 0.1914). This panel of five SNPs could confidently be used to trace groups of Faroe Plateau and Faroe Bank cod to their population of origin. In conclusion, multiple lines of evidence demonstrate that Faroe Plateau and Faroe Bank cod are truly two genetically distinct populations. While the findings contribute to a broader understanding of the biology and the genetics of Faroe Plateau and Faroe Bank cod, the novel SNPs developed may provide a valuable resource for potential future demands of i.e. genetic stock identification methods.