Seasonal variation in forest-floor CO2 exchange in a Swedish coniferous forest

Over a period of 2 years, CO2 exchange at the forest floor was studied at two stands in a mixed Scots pine and Norway spruce forest in central Sweden. Forest-floor efflux was measured with a transparent, open-chamber system, which means that during the day, both photosynthetic uptake by forest-floor vegetation and forest-floor respiration affect the efflux. Measured daily average net flux rates, F-n, ranged from between 0.3 and 1.7 mumol m(-2) s(-1) in winter, to maxima of 14.2 and 11.0 mumol m(-2) s(-1) in the summer at the southwest and north stand, respectively. Two-month periods of collected nocturnal data were used to estimate the seasonal variation in respiration at 0degreesC (R-0)and the temperature sensitivity (k) of an exponential function describing respiration (R). The parameters showed significant variation and similar patterns in the two stands; R-0 had a peak from May through August, whereas k was greatest at the turn of February-March. At that time k was very high, ca. 0.9 at both stands, but decreased to <0.08 in the summer. Subtraction of estimated R from F gave carbon assimilation by the field- and bottom-layer vegetation, F-g. The temperature dependence of Fg also varied significantly during the period May-August, whereas the light response did not; light-saturation occurred at 40-80 mumol m(-2) s(-1). At 20 degreesC, maximum photosynthesis was estimated to be 5.2 and 2.7 mumol m(-2) (ground) s(-1) in the southwest and north stand, respectively. Annual. forest-floor respiration was estimated to be 1.41-1.57 and 0.96-0.99 kg C m(-2) in the southwest and north stand, respectively. Carbon assimilation by forest-floor vegetation, as estimated from a fitted empirical function, reduced the efflux by 12-16%, which resulted in a net loss of 1.25-1.36 kg C m(-2) in the southwest stand and 0.81-0.86 kg C m(-2) in the north stand. The study shows that the seasonal variation in temperature dependence can be extremely large. This variation, which is affected by other factors than temperature, is very important if fluxes are to be studied at a shorter time-scale or when developing physical models. (C) 2002 Elsevier Science B.V All rights reserved.