Abstract
Anthropogenic emissions of CO sub(2) and other greenhouse gases have lead to a current atmospheric CO sub(2) concentration that is unprecedented in the recent geological history of the earth. Nearly half of anthropogenic CO sub(2) emissions have been sequestered in the terrestrial biosphere and oceans, slowing the climate change associated with this greenhouse gas. However, the future of this uptake is uncertain. One large sink for atmospheric CO sub(2), the growth of young temperate forests, may decline as these forests mature and undergo successional change in community composition and an expected decline in Net Primary Production (NPP) with increasing age. To examine the effect of elevated atmospheric CO sub(2) concentrations on ecosystem C loss via autotrophic respiration (R sub(a)), I measured rates of fine root respiration at the Duke Free Air CO sub(2) Enrichment (FACE) experiment, the longest running ecosystem level CO sub(2) experiment in a forest. Fine roots were investigated because their respiration was known to comprise a large but poorly quantified portion of total R sub(a). Growth under elevated CO sub(2) increased C release from fine root respiration because of higher amounts of fine root biomass; thus, some of the extra C fixed because of increased photosynthesis under elevated CO sub(2) was immediately respired and not sequestered by the ecosystem. I also investigated C storage more broadly by synthesizing twelve years of research on belowground C and N cycling at Duke FACE in an attempt to mechanistically explain two phenomena: (1) there has been no increase in soil C despite 12 years of increased C inputs to soils under elevated CO sub(2), and (2) the trees have increased soil N acquisition under elevated CO sub(2) and maintained the positive CO sub(2)-induced growth response over a long time period (>10 years). The enhanced rates of NPP under elevated CO sub( 2) increased the flux of C belowground, accelerated the rate of soil organic matter decomposition and increased nitrogen uptake from the soil through a priming mechanism. As a consequence of accelerated rates of soil organic matter decomposition, the aboveground C sink in biomass was maintained, but no additional C was stored in the soil, the longest lived pool of C in aggrading forests. The lack of C buildup in soils makes long-term mitigation of anthropogenic CO sub(2) emissions through sequestration by temperate forest ecosystems less likely, although C storage in biomass contributes to a decadal-scale C sink. In an attempt to understand how NPP and C storage varies at longer timescales in this forest type, I established a chronosequence of 12 forest stands ranging from 15 to 115 years old. These stands spanned the predictable and expected pattern of secondary succession in this region, where early-successional loblolly pines (Pinus taeda) were replaced by shade tolerant hardwoods such as oaks (Quercus spp.), hickories (Cayra spp. ), and a suite of other species. NPP declined strongly with increasing age, from ~1000 gC m super(-2) y super(-1) at 15 years of age to a stable value of ~600 gC m super(-2) y super(-1) at >50 years of age. This decline was driven exclusively by an 80% decline in pine wood production, and partially alleviated by increasing production by mid-successional hardwoods. The decline in pine production was driven by a decline in Gross Primary Production (GPP), not by increasing R sub(a) as was previously thought. The decline in GPP was consistent with increasing hydraulic limitation of leaf-level photosynthetic rates, but not consistent with increasing nitrogen limitation to photosynthetic capacity. Thus, I conclude that old, tall pine trees reduce stomatal conductance more frequently than shorter, young pine trees, which reduces leaf-level photosynthetic rates, GPP, and thus NPP. This suggests that NPP in old forests will be more strongly stimulated through a CO sub(2)-induced increase in GPP rather than the presumed decrease in NPP that would result from warming-induced increases in R sub(a).