Developments in Next-Generation Sequencing and Bioinformatics for Ecological Genetics

• Main Author: Fox, Graeme
• Format: Thesis
• Language: English
• Published: 2019
• Subjects: Homarus gammarus
• Online Access: https://e-space.mmu.ac.uk/626361/2/GFox_PhD_thesis.pdf

This thesis investigates the applications of next-generation sequencing to ecological studies by interrogating the power of high-throughput sequence data and developing resources to better understand the capabilities and limitations. In Chapter two, I use microsatellite detection methods to develop new markers for Raja undulata and show that after the production of the first captive generation, this small population shows no evidence of inbreeding depression or effects of genetic clustering by aquarium. I demonstrate the population has retained high genetic diversity throughout and highlight the importance of genetic management of ex situ populations. In Chapter three, I develop a novel in silico microsatellite marker design method. This new method allows the automated removal of markers likely to show elevated rates of null alleles, allelic dropout or cryptic fragment length altering mutations which invalidate the assumptions of mutation at a microsatellite locus. Furthermore, the method enables the automatic selection of likely polymorphic loci, thus removing many of the inefficiencies of marker design. In Chapter four, I perform parallel microsatellite and single nucleotide polymorphism (SNP) analysis to compare the application and relative power of each marker type in the analysis of the population structure of the larval dispersing decapod, (Homarus gammarus). Neither marker detects any genetic structuring in the fisheries of the UK and Ireland implying that genetic mixing is extremely high. SNP analysis is the preferred method due to quicker generation of data and results. In Chapter five, I conduct an investigation into the biases involved when selecting a metabarcoding marker for analysis of plant communities in mixed pollen samples collected from honey bee hives. I find high rates of false-positive identifications and 8 highly contrasting descriptions of plant communities, indicating low confidence in the data generated by each individual marker. I conclude that for plant metabarcoding, multiple ...