Get access

Sources of increased N uptake in forest trees growing under elevated CO2: results of a large-scale 15N study


Correspondence: Kirsten S. Hofmockel, tel. 515 294 2589, fax 515 294 1337, e-mail:


Nitrogen availability in terrestrial ecosystems strongly influences plant productivity and nutrient cycling in response to increasing atmospheric carbon dioxide (CO2). Elevated CO2 has consistently stimulated forest productivity at the Duke Forest free-air CO2 enrichment experiment throughout the decade-long experiment. It remains unclear how the N cycle has changed with elevated CO2 to support this increased productivity. Using natural-abundance measures of N isotopes together with an ecosystem-scale 15N tracer experiment, we quantified the cycling of 15N in plant and soil pools under ambient and elevated CO2 over three growing seasons to determine how elevated CO2 changed N cycling between plants, soil, and microorganisms. After measuring natural-abundance 15N differences in ambient and CO2-fumigated plots, we applied inorganic 15N tracers and quantified the redistribution of 15N for three subsequent growing seasons. The natural abundance of leaf litter was enriched under elevated compared to ambient CO2, consistent with deeper rooting and enhanced N mineralization. After tracer application, 15N was initially retained in the organic and mineral soil horizons. Recovery of 15N in plant biomass was 3.5 ± 0.5% in the canopy, 1.7 ± 0.2% in roots and 1.7 ± 0.2% in branches. After two growing seasons, 15N recoveries in biomass and soil pools were not significantly different between CO2 treatments, despite greater total N uptake under elevated CO2. After the third growing season, 15N recovery in trees was significantly higher in elevated compared to ambient CO2. Natural-abundance 15N and tracer results, taken together, suggest that trees growing under elevated CO2 acquired additional soil N resources to support increased plant growth. Our study provides an integrated understanding of elevated CO2 effects on N cycling in the Duke Forest and provides a basis for inferring how C and N cycling in this forest may respond to elevated CO2 beyond the decadal time scale.