Relationship between functional traits and genome size variation of angiosperms with different life forms

Aims: The genome size between species, especially in angiosperms, can be extremely diverse. Here, we compiled genome size data for 11,215 angiosperm species from 2,226 genera and 245 families to explore the relationships between four functional traits (i.e. seed mass, maximum plant height, leaf nitrogen and phosphorus concentrations) with genome size in angiosperms from different life forms (i.e. annual herbs, perennial herbs, and woody plants). Method: We used the 1C-value of DNA content as a measurement for genome size. Genome sizes were obtained from the latest version of Kew Plant DNA C-values Database and Genome Size in Asteraceae Database (GSAD). We also complemented our taxon sampling with data from the literature over the past 10 years. We obtained life form and functional trait from Flora of China, Flora of North America and the Seed Information Database (SID). We used the most recent updated time-calibrated phylogeny published by Smith and Brown in 2018, and pruned it to the 6,612 species from our species list. We used two indices (i.e. Blomberg’s K and Pagel’s λ) to test for the prescence of a phylogenetic signal for the evolution of angiosperm genome size. We performed a standardized major axes (SMA) Model II and focused on the relationships between genome size and the four functional traits. We also conducted a principal components analysis (PCA) to explore trade-offs between functional traits and genome size in angiosperms with different life forms. Results: The genome size for most angiosperms was small and few species had large genomes. The median value of angiosperm genome size was 1.58 pg with perennial herbs having the largest median genome size (2.5 pg), followed by annual herbs (1.55 pg), and then woody species (1.14 pg). Variation of the genome size was greatest in perennial herbs distributed over a wider range than woody species and then annual herbs. Tests for phylogenetic signals with genome size indicated that evolution was non-random. The value for Blomberg’s K was 0.031 (P < 0.001) and the value for Pagel’s λ was 0.943 (P < 0.001. There was also a significant difference between functional traits and genome size among the three different life forms. Our results from the standardized major axes regression found that there was a significant relationship between seed mass with genome size in herbs but not woody plants. However, the relationship of maximum plant height was significant with genome size in woody plants but not herbs. There were no significant correlations between leaf nitrogen or phosphorus concentration with genome size except for leaf nitrogen concentration in woody plants. When looking at the relationship between four functional traits with genome size, we fund a negative correlation between seed mass and maximum plant height with genome size, and saw no significant correlation with leaf nitrogen or phosphorus concentration which is consistent with the SMA results. Conclusion: Our study highlights that the correlation between functional traits and genome size vary between herbaceous and woody species and suggests that trade-offs between genome size, life forms and functional traits might play an essential role in ecological adaptation and evolution of angiosperms.