LONG-TERM RESPONSIVENESS TO FREE-AIR CO2 ENRICHMENT OF FUNCTIONAL TYPES, SPECIES AND GENOTYPES OF PLANTS FROM FERTILE PERMANENT GRASSLAND

To test inter- and intraspecific variability in the responsiveness to elevated CO2, 9-14 different genotypes of each of 12 perennial species from fertile permanent grassland were grown in Lolium perenne swards under ambient (35 Pa) and elevated (60 Pa) atmospheric partial pressur e of CO2 (pCO(2)) for 3 years in a free air carbon dioxide enrichment (FACE) experiment. The plant species were grouped according to their f unctional types: grasses (L. perenne, L. multiflorum, Arrhenatherum el atius, Dactylis glomerata, Festuca pratensis, Holcus lanatus, Trisetum flavescens), non-legume dicots (Rumex obtusifolius, R. acetosa, Ranun culus friesianus), and legumes (Trifolium repens, T. pratense). Yield (above a cutting height of 4.5 cm) was measured three times per year. The results were as follow. (1) There were highly significant differen ces in the responsiveness to elevated pCO(2) between the three functio nal types; legumes showed the strongest and grasses the weakest yield increase at elevated pCO(2). (2) There were differences in the tempora l development of responsiveness to elevated pCO(2) among the functiona l types. The responsiveness of the legumes declined from the first to the second year, while the responsiveness of the non-legume dicots inc reased over the 3 years. During the growing season, the grasses and th e non-legume dicots showed the strongest response to elevated pCO(2) d uring reproductive growth in the spring. (3) There were no significant genotypic differences in responsiveness to elevated pCO(2). Our resul ts suggest that, due to interspecific differences in the responsivenes s to elevated pCO(2), the species proportion within fertile temperate grassland may change if the increase in pCO(2) continues. Due to the t emporal differences in the responsiveness to elevated pCO(2) among spe cies, complex effects of elevated pCO(2) on competitive interactions i n mixed swards must be expected. The existence of genotypic variabilit y in the responsiveness to elevated pCO(2), on which selection could a ct, was not found under our experimental conditions.