Genotypic variation in the response of two perennial grass species to elevated carbon dioxide

Shoot and reproductive biomass of genotypes of Bromus erectus and Dactylis glomerata grown in competition at ambient and elevated CO₂ were examined for 2 consecutive years in order to test whether genetic variation in those traits exists and whether it is maintained over time. At the species level, a positive CO₂ response of shoot biomass of both species was only found in the first year of treatment. At the genotype level, no significant CO₂2genotype interaction was found at any single harvest either for vegetative or reproductive biomass of either species. Analysis over time, however, indicated that there is a potential for evolutionary adaptation only for D. glomerata: (1) repeated measures ANOVA detected a marginally significant CO₂2genotype2time interaction for shoot biomass, because the range of the genotypes CO₂ response increased over time; (2) genotypes that displayed the highest response during the first year under elevated CO₂ also showed the highest response the second year. Null (B. erectus) or weak (D. glomerata) selective potentials of elevated CO₂ were detected in this experiment, but short time series could underestimate this potential with perennial species.