Evolutionary genomics analysis of plants aims to reveal and help us to understand the history of genome evolution that plants have undergone. So far, many specific topics and questions of genome evolution have been studied and answered. However, there are still a large number of questions to which the answers are unknown or not clear. In this dissertation, I focus on two specific topics of evolutionary genomics: (1) genome size evolution following genomic rearrangements in plants; (2) ancestral genome reconstruction in legumes.

Using a model of two wild peanut relatives in which one genome experienced large rearrangements, we find that the main determinant in genome size reduction is a set of inversions which experienced subsequent net sequence removal in the inverted regions. We observe a general pattern in which sequence is lost more rapidly at newly distal (telomeric) regions than it is gained at newly proximal (pericentromeric) regions – resulting in net sequence loss in the inverted regions. The major driver of this process is recombination, determined by the chromosomal location. Any type of genomic rearrangement that exposes proximal regions to higher recombination rates can cause genome size reduction by this mechanism. Sequence loss in those regions was primarily due to removal of transposable elements. Illegitimate recombination is likely the major mechanism responsible for the sequence removal, rather than unequal intrastrand recombination. We also measure the relative rate of genome size reduction in these two Arachis diploids. We also test our model in other plant species and find that it applies in all cases examined, suggesting our model is widely applicable.

Inversions occurring in tetraploid cultivated peanut after the polyploidization event provide us an excellent opportunity to examine the model of genome size reduction following genomic rearrangements in polyploidy. It is also a good opportunity to understand the genome size reduction process at its early stage, since the inversions are quite recent (likely younger than 10,000 years). We observe that the model of genome size reduction still holds in the recently-derived tetraploid peanut as it does in the much earlier-diverging diploid progenitors. We find that the genome size reduction process starts with differences in very long sequence deletions and then spreads to mid-length sequence deletions later. We measure the relative rate of size reduction of the inverted region in tetraploid peanut, finding that it is higher than the rates calculated in our previous study between Arachis diploids. We argue this is because the rate of size reduction is more rapid in the early generations after the inversion.

We describe the reconstruction of a hypothetical ancestral genome for the papilionoid legumes, in order to help us better understand the evolutionary histories of these legumes. We use a novel method for identifying informative markers, to reconstruct the ancestral genomes for selected legume species, including Glycine max, which has a recent exclusive WGD event. We infer that the reconstructed most recent common ancestor of all selected legume species (all within the Papilionoideae) has 9 chromosomes. The model then predicts that chromosome numbers reduced to 8 in Medicago truncatula and Cicer arietinum separately, through two separate single fusion events. In Lotus japonicus, a series of rearrangement events is the major cause of the chromosome number reduction to 6. We infer that the chromosome number increased mostly independently in Cajanus cajan, Glycine max, Phaseolus vulgaris and Vigna radiata. In Arachis (A. duranensis and A. ipaensis), there was an increase in chromosome number prior to their divergence. The chromosome structural evolution described here is consistent with the phylogenetic distribution of a large collection of chromosome counts in the legumes.