Trifolium repens L. and Lolium perenne L. were grown in monocultures and bi-species mixture in a Free Air Carbon Dioxide Enrichment (FACE) experiment at elevated (60 Pa) and ambient (35 Pa) CO2 partial pressure (pCO2) for three years. The effects of defoliation frequencies (4 and 7 cuts in 1993; 4 and 8 cuts in 1994/95) and nitrogen fertilization (10 and 42 g m–2 y–1 N in 1993; 14 and 56 g m–2 y–1 N in 1994/95) on the growth response to pCO2 were investigated.
There were significant interspecific differences in the CO2 responses during the first two years, while in the third growing season, these interspecific differences disappeared. Yield of T. repens in monocultures increased in the first two years by 20% when grown at elevated pCO2. This CO2 response was independent of defoliation frequency and nitrogen fertilization. In the third year, the CO2 response of T. repens declined to 11%. In contrast, yield of L. perenne monocultures increased by only 7% on average over three years at elevated pCO2. The yield response of L. perenne to CO2 changed according to defoliation frequency and nitrogen fertilization, mainly in the second and third year. The ratio of root/yield of L. perenne increased under elevated pCO2, low N fertilizer rate, and frequent defoliation, but it remained unchanged in T. repens. We suggest that the more abundant root growth of L. perenne was related to increased N limitation under elevated pCO2.
The consequence of these interspecific differences in the CO2 response was a higher proportion of T. repens in the mixed swards at elevated pCO2. This was evident in all combinations of defoliation and nitrogen treatments. However, the proportion of the species was more strongly affected by N fertilization and defoliation frequency than by elevated pCO2. Based on these results, we conclude that the species proportion in managed grassland may change as the CO2 concentration increases. However, an adapted management could, at least partially, counteract such CO2 induced changes in the proportion of the species. Since the availability of mineral N in the soil may be important for the species’ responses to elevated pCO2, more long-term studies, particularly of processes in the soil, are required to predict the entire ecosystem response.