Proper management and accounting of forest carbon requires good knowledge of how disturbances and climate affect the carbon dynamics of different stand types. We have investigated such relationships by measuring, over a 5-year period (2003–2007), the net ecosystem productivity (NEP), gross ecosystem productivity (GEP) and ecosystem respiration (ER) of 26 forest sites in Canada using the eddy covariance technique. The study included black spruce, jack pine, Douglas-fir, aspen, boreal mixedwood and white pine forest ecosystems ranging in age from 1- to 153-years. The dataset included six chronosequences (one afforested plantation, three harvested and two burned).
Following planting, the afforested white pine stands quickly became carbon sinks and offset initial carbon losses after 4 years. Depending on forest type, the other forest stands were carbon sources for 10–18 years following a disturbance, offset initial carbon losses after 19–47 years, and showed net total gains ranging from 38 to 86 Mg C ha−1 at 80 years. Peak NEP ranged from 0.9 to 2.9 Mg C ha−1 year−1 at ages of 35–55 years except for the afforested white pine where it was 6.9 Mg C ha−1 year−1 at 15–20 years. Stepwise regression and Pearson correlation analyses indicated that the GEP and ER of mature stands (>70 years old) were driven mainly by climate, while fluxes of young stands (<19 years old) were driven by both leaf area index and climate.
Although stand age of the afforested white pine plantations did not affect the GEP growing season lengths, the growing season length of the other forests increased with age until about 20 years and this coincided with the switch from carbon source to sink. With the exception of the afforested white pine, peak GEP/ER ratios of the youngest sites occurred later in the growing season compared to older sites. The strong influence of stand age on the seasonal dynamics of GEP fluxes needs to be considered to avoid confounding the impacts of climate change with those of disturbance. These age-related seasonality effects are continental in scope and should be important in interpreting the time series of atmospheric CO2 concentration measurements at regional and global scales.