Across the angiosperms, we see a range in genome size of over four orders of magnitude. Species may lose or gain large chromosomal segments or smaller transposable elements, may undergo gene loss or gain, or experience polyploidy. These events can all lead to different trends in the genome size evolution of different clades. However, the external drivers of these changes are multifactorial and difficult to prise apart.
In a new article published in Annals of Botany, lead authors N. Ivalú Cacho and Patrick J. McIntyre and colleagues attempted to determine what inter-related factors are responsible for genome size evolution in a diverse clade of the Brassicaceae. The researchers ascertained the genome size of 47 species of the genus Streptanthus and its close relatives, comparing this data to previously gathered evidence on the species’ habitat, life history, soil preference, geographic range, and biochemistry. Modelling was used to identify which variables could best explain the trends seen across the clade.
The study species showed variation in both genome size and rate of evolution across subclades. In addition to a moderate to strong phylogenetic signal, the two factors found to be most responsible for genome size were climate seasonality and the production of glucosinolate compounds. None of the ‘usual suspects’ – life history, range size, or soil biology – appear to have an important influence on genome size for this group.
In the case of climate seasonality, larger genomes were associated with more seasonal environments. This may be because “highly seasonal environments where conditions of temperature and precipitation relatively more favourable to cell division were limited to short windows, may generally select for rapid growth through cell expansion, temporally decoupled from cell division,” write the authors. Glucosinolate compounds, which function as herbivore deterrents, were produced in larger amounts by species with smaller genomes, though the reason behind the correlation is not yet well understood.
Despite finding two significant factors in genome size variation in this clade, the best models produced still explained only 20% of the total variation, suggesting there are important factors yet to be taken into account.