Growing season and spatial variations of carbon fluxes of Arctic and boreal ecosystems in Alaska (USA)

• Published in: Ecological applications 2013-12, Vol.23 (8), p.1798-1816
• Main Authors: Ueyama, Masahito, Iwata, Hiroki, Harazono, Yoshinobu, Euskirchen, Eugénie S, Oechel, Walter C, Zona, Donatella
• Format: Article
• Language: English
• Subjects: Alaska; Arctic Regions; boreal forest; Boreal forests; Carbon - chemistry; Carbon - metabolism; Carbon Cycle; Carbon Dioxide - chemistry; Carbon Dioxide - metabolism; Climate models; CO2 fluxes; Deciduous forests; ecosystem respiration; eddy covariance; Environmental Monitoring; fluxes; Forest ecosystems; Forest fires; gross primary productivity; Growing seasons; Marine ecosystems; Net ecosystem exchange; Plants - classification; Plants - metabolism; Seasons; Terrestrial ecosystems; Time Factors; Trajectory of the Arctic; tundra; Tundras
• ISSN: 1051-0761; 1939-5582
• DOI: 10.1890/11-0875.1

To better understand the spatial and temporal dynamics of CO 2 exchange between Arctic ecosystems and the atmosphere, we synthesized CO 2 flux data, measured in eight Arctic tundra and five boreal ecosystems across Alaska (USA) and identified growing season and spatial variations of the fluxes and environmental controlling factors. For the period examined, all of the boreal and seven of the eight Arctic tundra ecosystems acted as CO 2 sinks during the growing season. Seasonal patterns of the CO 2 fluxes were mostly determined by air temperature, except ecosystem respiration (RE) of tundra. For the tundra ecosystems, the spatial variation of gross primary productivity (GPP) and net CO 2 sink strength were explained by growing season length, whereas RE increased with growing degree days. For boreal ecosystems, the spatial variation of net CO 2 sink strength was mostly determined by recovery of GPP from fire disturbance. Satellite-derived leaf area index (LAI) was a better index to explain the spatial variations of GPP and NEE of the ecosystems in Alaska than were the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI). Multiple regression models using growing degree days, growing season length, and satellite-derived LAI explained much of the spatial variation in GPP and net CO 2 exchange among the tundra and boreal ecosystems. The high sensitivity of the sink strength to growing season length indicated that the tundra ecosystem could increase CO 2 sink strength under expected future warming, whereas ecosystem compositions associated with fire disturbance could play a major role in carbon release from boreal ecosystems.