Plant response to nutrients differs among traits and depends on species' nutrient requirements

Intraspecific trait variations in response to nutrient availability are expected to depend on (1) the category of traits considered, and (2) species ecology, with species requiring high nutrient levels expected to be more plastic. However, there are few comparisons of trait responses considering simultaneously (a) above-ground traits approximating ecological strategies, (b) root traits involved in nutrient acquisition, and (c) traits integrating the whole plant, and including multiple species.

We studied 17 annual species coming from two contrasted environments in the same rangeland of southern France. Plants were grown in a common garden under two fertilization treatments.

We evidenced no effect of origin on trait values, suggesting little or no differentiation according to the environment of origin. Among the 14 traits measured, whole-plant traits, in particular plant nitrogen content, plant dry mass and root mass fraction, showed strong plastic responses to fertilization, whereas the response was weak or even absent for above- and below-ground organ-level traits related to ecological strategies, suggesting that they play a secondary role in plant responses to nutrient availability. Finally, species' ecological preferences (i.e. their nutrient requirements), predicted the plasticity in plant nitrogen content per mass, whereas species position along the acquisition-conservation trade-off (approximated by leaf traits) predicted plasticity in plant dry mass. This cautions against the systematic use of leaf traits as a proxy of species ecology and functioning.

Our results challenge the assumption that leaf traits universally reflect plant responses to nutrient availability. They advocate a better characterization of traits directly involved in nutrient acquisition and underscore the importance of considering how trait-trait and trait-environment relationships may depend on the group of species considered. These findings offer an avenue for more accurate predictions of plant responses to nutrient gradients in natural and managed ecosystems.