Partitioning European grassland net ecosystem CO2 exchange into gross primary productivity and ecosystem respiration using light response function analysis

Tower CO2 flux measurements from 20 European grasslands in the EUROGRASSFLUX data set covering a wide range of environmental and management conditions were analyzed with respect to their ecophysiological characteristics and CO2 exchange (gross primary production, P(g), and ecosystem respiration, R(e)) using light-response function analysis. Photosynthetically active radiation (Q) and top-soil temperature (T(s)) were identified as key factors controlling CO2 exchange between grasslands and the atmosphere at the 30-min scale. A nonrectangular hyperbolic light-response model P(Q) and modified nonrectangular hyperbolic light-temperature-response model P(Q, T(s)) proved to be flexible tools for modeling CO2 exchange in the light. At night, it was not possible to establish robust instantaneous relationships between CO2 evolution rate rn and environmental drivers, though under certain conditions, a significant relationship r(n)=r(0) e(k)T(T)s was found using observation windows 7-14 days wide. Principal light-response parameters--apparent quantum yield, saturated gross photosynthesis, daytime ecosystem respiration, and gross ecological light-use efficiency, epsilon= P(g)/Q, display patterns of seasonal dynamics which can be formalized and used for modeling. Maximums of these parameters were found in intensively managed grasslands of Atlantic climate. Extensively used semi-natural grasslands of southern and central Europe have much lower production, respiration, and light-use efficiency, while temperate and mountain grasslands of central Europe ranged between these two extremes. Parameters from light-temperature-response analysis of tower data are in agreement with values obtained using closed chambers and free-air CO2 enrichment. Correlations between light-response and productivity parameters provides the possibility to use the easier to measure parameters to estimate the parameters that are more difficult to measure. Gross primary production (P(g)) of European grasslands ranges from 1700 g CO2 m-2 year-1 in dry semi-natural pastures to 6900 g CO2 m-2 year-1 in intensively managed Atlantic grasslands. Ecosystem respiration (R(e)) is in the range 1800 < R(e) < 6000 g CO2 m-2 year-1. Annual net ecosystem CO2 exchange (NEE) varies from significant net uptake (>2400 g CO2 m-2 year-1) to significant release (<-600 g CO2 m-2 year-1), though in 15 out of 19 cases grasslands performed as net CO2 sinks. The carbon source was associated with organic rich soils, grazing, and heat stress. Comparison of P(g), R(e), and NEE for tower sites with the same characteristics from previously published papers obtained with other methods did not reveal significant discrepancies. Preliminary results indicate relationships of grassland P(g) and R(e) to macroclimatic factors (precipitation and temperature), but these relationships cannot be reduced to simple monofactorial models.