Species traits and shoot-root biomass allocation in 20 dry-grassland species

Aims A plant has a limited amount of resources at any time and it allocates them to different structures. In spite of the large number of previous studies on allocation patterns within single species, knowledge of general patterns in species allocation is still very limited. This is because each study was done in different conditions using different methodology, making generalization difficult. We investigate intraspecific above- versus below-ground biomass allocation among individuals across a spectrum of dry-grassland plant species at two different developmental stages and ask whether allocation is age- and species specific, and whether differences among species can be explained by their life-history traits and phylogeny.Methods We collected data on above- and below-ground biomass of seedlings and adult plants of 20 species from a common garden experiment. We analysed data on shoot–root biomass allocation allometrically and studied the relationship between the allometric exponents (slopes on log–log scale), species life-history traits and phylogenetic distances.Important findings We found isometric as well as allometric patterns of biomass allocation in the studied species. Seedlings and adult individuals of more than half of the species differed in their above- versus below-ground biomass allometric exponents. Seedlings and adult individuals of the remaining species differed in their allometric coefficients (intercepts). Annual species generally allocated proportionally more to above- than below-ground biomass as seedlings than as adults, whereas perennial species showed the opposite pattern. Plant life-history traits, such as plant life span, age of first flowering, month in which the species begin flowering and specific leaf area were much more important in explaining differences in shoot–root allometry among species than were phylogenetic relationships. This suggests that allocation patterns vary greatly among closely related species but can be predicted based on species life-history traits.