Reciprocal exchange of eukaryotic genetic material during meiotic crossing over is a major source of genetic variation in sexually reproducing species. Crossover events are not distributed randomly along chromosomes and some regions of the genome, like the centromeres, rarely undergo recombination. Modifying crossover levels and distributions via genetic engineering may provide effective tools for plant breeders to accelerate strain improvement. Despite low recombination rates and their evolutionary conserved function as kinetochore assembly loci, centromeres exhibit some of the highest levels of variation within eukaryotic genomes. Discovery of centromere structure has been hindered by the challenging process of genomic assembly of repetitive regions, as many species contain megabase-long arrays of centromeric tandem repeats. While new long-read DNA sequencing technologies allow for more accurate assembly across the centromeres, methods for their annotation are also required.

In this thesis, I present the development of Tandem Repeat Annotation and Structural Hierarchy (TRASH) software that facilitates analysis of tandem arrays, including centromeric satellite arrays, without prior knowledge of repeat families present in an assembly. I benchmarked TRASH against other software and found it to advance on the current annotation and analysis methods. I used TRASH to analyse in depth the centromeric structures of multiple accessions of metacentric Arabidopsis thaliana, Arabidopsis lyrata, Brassica oleracea and holocentric Rhynchospora genus species.

I also present progress towards investigation of the HEI10 meiotic E3 ligase and its role in modulation of crossover levels. Specifically, quantification of the dosage effect of HEI10 multi-copy lines of tomato and Arabidopsis on the crossover recombination landscape.

Together this work contributes to a better understanding of plant centromeric regions and meiotic recombination modulation. It also provides a novel bioinformatics tool for centromere sequence analysis and tandem repeat identification to the scientific community,