Keywords:Windthrow, Eddy covariance, Vegertation recovery
Forest ecosystems are expected as a major sink of atmospheric carbon dioxide (CO2), whereas their ability to absorb CO2 is severely perturbed by disturbances, such as deforestation, fires, windthrow etc. Many studies which adopted the chrono-sequence approach reported that such severe disturbances often changed forest to be a CO2 source. However, there are few studies that directly measure CO2 flux for a long term (more than 10 years) above a disturbed forest site during an early stage of secondary succession after severe disturbance.A flux site of a larch plantation in Tomakomai, Hokkaido, Japan was struck by a typhoon in September 2004. Because of wind storm, about 90% of trees fell down. The fallen trees were removed by heavy machinery from the site, through which the soil surface and understory species were also disturbed. After the operations of timber transport, secondary succession progressed naturally in the ex-forest site. We recommenced flux measurement in August 2005. CO2 flux has been measured by the eddy covariance technique with an open-path CO2 / H2O analyzer (LI7500, Licor) during a snow-free period from mid-April to mid-November. Cumulative net ecosystem CO2 exchange (NEE) during the snow-free period was positive every year until 2015, whereas it showed a negative relationship, which indicates that the CO2 source strength of the ecosystem decreased. This negative relationship was caused by the increase of gross primary production (GPP) or ecosystem photosynthesis, which corresponded to vegetation recovery through secondary succession.