Yield response of Lolium perenne swards to free air CO2 enrichment increased over six years in a high N input system on fertile soil

After a step increase in the atmospheric partial pressure of CO2 (pCO(2)), the availability of mineral N may be insufficient to meet the plant's incre ased demand for N. Over time, however, the ecosystem may adapt to the new c onditions, and a new equilibrium may be established in the fluxes of C and N. This would result in a higher dry mass (DM) yield response of the plants to elevated pCO(2). The effect of elevated atmospheric pCO(2) (60 Pa pCO(2)) was studied in Lol ium perenne L. swards with two N fertilization treatments (14 and 56 g m(-2 ) y(-1)) in a six-year FACE (Free Air Carbon dioxide Enrichment) experiment . In the high N treatment, the input of N with fertilizer considerably exce eded the export of N with the harvested plant material in both CO2 treatmen ts leading to an apparent net input of N into the ecosystem. Accordingly, t he proportion of harvested N derived from N-15 labelled fertilizer N, appli ed throughout the experiment (< 6 years), increased over the years. Under t hese high N conditions, the annual DM yield response of the Lolium perenne sward to elevated pCO(2) increased (from 7% in 1993 to 25% in 1998). In par allel, the response of N yield to elevated pCO(2) increased, and the initia lly negative effect of elevated pCO(2) on specific leaf area (SLA) disappea red. The high N input system seemed to overcome in part an initially limiti ng effect of N on the yield response to elevated pCO(2) within a few years. In contrast, there was no apparent net input of N into the ecosystem in th e low N treatment, because N fertilization just compensated the export of N with the harvested plant material. Accordingly, the proportion of harveste d N yield, derived from fertilizer N, which was applied throughout the expe riment, remained low. At low N, the availability of mineral N strongly limi ted plant growth and yield production in both CO2 treatments; the low yield s of DM and N, the low concentration of N in the plant material, and the lo w SLA reflected this. Although the plants grew under the same environmental conditions and the same management treatment as plants in the high N treat ment, the response of DM yields to elevated pCO(2) in the low N treatment r emained weak throughout the experiment (5% in 1993 and 9% in 1998). The res ults are discussed in the context of the sizes of the different N pools in the soil, the allocation of N within the plant and the possible effects on temporal immobilization, and the availability of mineral N for yield produc tion as affected by elevated pCO(2) and N fertilization.