A free-air enrichment system for exposing tall forest vegetation to elevated atmospheric CO2


Correspondence: George R. Hendrey, fax + 1/516-344-2060, e-mail


A free-air CO2 enrichment (FACE) system was designed to permit the experimental exposure of tall vegetation such as stands of forest trees to elevated atmospheric CO2 concentrations ([CO2]a) without enclosures that alter tree microenvironment. We describe a prototype FACE system currently in operation in forest plots in a maturing loblolly pine (Pinus taeda L.) stand in North Carolina, USA. The system uses feedback control technology to control [CO2] in a 26 m diameter forest plot that is over 10 m tall, while monitoring the 3D plot volume to characterize the whole-stand CO2 regime achieved during enrichment. In the second summer season of operation of the FACE system, atmospheric CO2 enrichment was conducted in the forest during all daylight hours for 96.7% of the scheduled running time from 23 May to 14 October with a preset target [CO2] of 550 μmol mol–1, ≈ 200 μmol mol–1 above ambient [CO2]. The system provided spatial and temporal control of [CO2] similar to that reported for open-top chambers over trees, but without enclosing the vegetation. The daily average daytime [CO2] within the upper forest canopy at the centre of the FACE plot was 552 ± 9 μmol mol–1 (mean ± SD). The FACE system maintained 1-minute average [CO2] to within ± 110 μmol mol–1 of the target [CO2] for 92% of the operating time. Deviations of [CO2] outside of this range were short-lived (most lasting < 60 s) and rare, with fewer than 4 excursion events of a minute or longer per day. Acceptable spatial control of [CO2] by the system was achieved, with over 90% of the entire canopy volume within ± 10% of the target [CO2] over the exposure season. CO2 consumption by the FACE system was much higher than for open-top chambers on an absolute basis, but similar to that of open-top chambers and branch bag chambers on a per unit volume basis. CO2 consumption by the FACE system was strongly related to windspeed, averaging 50 g CO2 m–3 h–1 for the stand for an average windspeed of 1.5 m s–1 during summer. The [CO2] control results show that the free-air approach is a tractable way to study long-term and short-term alterations in trace gases, even within entire tall forest ecosystems. The FACE approach permits the study of a wide range of forest stand and ecosystem processes under manipulated [CO2]a that were previously impossible or intractable to study in true forest ecosystems.