Species habitat preferences and root trait variation across 65 temperate perennial forbs

While we know a lot about variation of root traits across large set of species, knowledge on differences in root traits among species with different ecological optima, simultaneously considering species life span and phylogeny, is limited. We also do not know if inter-specific differences in root traits measured in one environment apply in another environment. Such knowledge is crucial to predict species responses to future environments. Using 65 species cultivated under uniform conditions, we studied effects of species habitat preference, describing under which conditions the species naturally occur, on root morphological and chemical traits and allocation to roots while also considering species life span, phenology at harvest and phylogeny. In a subset of species, we explored if species rankings in values of different traits depend on specific substrate of growth. Inter-specific trait differences were strongly linked to species habitat preferences. The best predictor was indicator value for soil disturbance with roots of species preferring disturbed habitats having higher specific root length and lower diameter, suggesting low collaboration with mutualists. While life span and phylogeny also determined trait values, their inclusion into models did not change effects of habitat preferences. The patterns are thus not a result of species niche conservatism, but contemporary species adaptations. Species ranking in different substrates was more consistent for root morphology than for root chemistry and root/shoot ratio. Root trait variation is driven by species habitat preferences indicating that inter-specific root trait variation is a result of species adaptations to different environments. Interestingly, disturbance indicator value was a better predictor of root trait variation than other, more commonly, considered habitat characteristics. Inter-specific differentiation in root morphology is consistent among substrates and can thus be compared across studies, but root chemistry and allocation data have to be used with caution.