Inferring ecosystem-level rates of gross primary productivity, respiration, and evapotranspiration with automatic light-dark measurement chambers

<p>In experimental ecosystem ecology, plot sizes are most often too small to apply eddy flux techniques and estimation of ecosystem gas exchange rates relies on various chamber measurement technologies. Furthermore, drained areas often results in increased growth of trees which complicates application of eddy flux measurements on small plots.</p><p>We combined ECO<sub>2</sub>FluX ecosystem-level automatic chambers (prenart.dk) with an LI-8100/LI-8150 multiplexer systems (licor.com) in a range of Danish and Norwegian ecosystems experiments spanning agriculture, grassland/heathlands and peatland ecosystems. The automatic closed, none-steady state chambers each cover an area of 3,117 cm<sup>2</sup> (63 cm diameter), are 80 cm tall (volume: 250L), and are capable of switching automatically between transparent and darkened mode, enabling separation of light-sensitive and light-indifferent processes in the ecosystems covered. For CO<sub>2</sub> fluxes, net exchange (NEE) was estimated as the flux in transparent mode, ecosystem respiration (R<sub>E</sub>) in darkened mode, while Gross Ecosystem Productivity (GPP) was estimated as NEE – R<sub>E</sub>.</p><p>Chambers were set up to measure gas concentrations every second using enclosure times of 4-5 minutes, first in light mode and 10-30 minutes later in dark mode, with 3-48 repetitions per day. The longest time series spans 5 years of measurements and contain >60,000 point measurements.</p><p>In this presentation, we will present an analysis of the ability of the light-dark chamber data to infer ecosystem-level rates of gross primary productivity, respiration, net CO<sub>2</sub> exchange, and evapotranspiration. In the two Norwegian peatland sites, flux measurements may be compared directly with eddy flux measurements. We also compare the rates of the direct estimates of GPP from the light-dark chamber measurements to estimates inferred from using the light (NEE) measurements only followed by applying methodologies normally used for eddy flux measurements. This comparison may help constrain potential biases in both the closed chamber and eddy flux techniques. Finally, we investigate the ability of using such closed chambers to estimate ecosystem evapotranspiration rates at the plot scale. Such application may be useful for estimating the effects on evapotranspiration in field-scale experiments manipulating the ecosystem water balance either directly or indirectly.</p>