Genomes bear the scars of their history. These scars can give important clues to how adaptation and diversity can arise during evolution. Of particular interest are genomic islands - regions of the genome that exhibit exceptionally low or high degrees of divergence. One hypothesis is that these islands derive from a limited repertoire of alleles that have been repeatedly deployed during evolution. Alternatively, they may derive from novel functional variants that have arisen independently. We aim to test these hypotheses by studying genomic islands in Antirrhinum, where they have been previously identified at the self- incompatibility and flower colour loci. By studying allelic variations at these loci in both a phylogenetic and genomic context, we aim to establish how alleles and traits have evolved in different lineages. This should allow us to determine how islands arise, persist and decay, identifying the processes and genes that shape natural diversity and provide the raw material for plant breeding.