Effect of competition on the responses of grasses and legumes to elevated atmospheric CO₂ along a nitrogen gradient: differences between isolated plants, monocultures and multi‐species mixtures

The responses to CO 2 of perennial grasses ( Danthonia richardsonii and Phalaris aquatica ) and legumes ( Lotus pedunculatus and Trifolium repens ) were compared under controlled conditions for isolated plants, monoculture stands and mixed‐species stands along a N gradient to test whether: plant–plant interactions between species in mixed stands changed with concentration of CO 2 ; responses to CO 2 of species in mixtures could be related to their responses as single stands; responses of mixtures to CO 2 could be related to the responses of individual species to CO 2 and to competition. Plants were grown for 60 d in sand, using nutrient solutions (six nitrate concentrations from 0.25 to 16 mM NO 3 ), at ambient ( c . 357 μl l −1 ) or elevated CO 2 ( c . 712 μl l −1 ). Species dominance in the mixtures depended more on the range of N than of CO 2 concentration provided: T. repens and L. pedunculatus dominated at low concentrations of N; L. pedunculatus and P. aquatica performed better at high concentrations. Responses of species in mixtures to CO 2 were related to their responses in monocultures but not to those of isolated plants. Species biomass proportions in mixtures under ambient CO 2 determined the outcome of mixture responses to CO 2 more than of individual species responses to CO 2 . These results emphasize the influence of plant–plant interactions on community responses to CO 2 , since mixture behaviour under elevated CO 2 could not be scaled‐up from responses by isolated plants in this experiment.