CARBON DIOXIDE AND METHANE FLUXES BY AFOREST SOIL UNDER LABORATORY-CONTROLLEDMOISTURE AND TEMPERATURE CONDITIONSRICHARD D. BOWDEN,1* KATHLEEN M. NEWKIRK2and GINA M. RULLO11Allegheny College, Department of Environmental Science, Meadville, PA 16335, U.S.A. and2TheEcosystems Center, Marine Biological Laboratory, Woods Hole, MA 02543, U.S.A.(Accepted 17 September 1997)SummaryÐCarbon dioxide and methane are important greenhouse gases whose exchange rates betweensoils and the atmosphere are controlled strongly by soil temperature and moisture. We made a labora-tory investigation to quantify the relative importance of soil moisture and temperature on ¯uxes ofCO2and CH4between forest soils and the atmosphere. Forest ¯oor and mineral soil material were col-lected from a mixed hardwood forest at the Harvard Forest Long-Term Ecological Research Site (MA)and were incubated in the laboratory under a range of moisture (air-dry to nearly saturated) and tem-perature conditions (5±258C). Carbon dioxide emissions increased exponentially with increasing tem-perature in forest ¯oor material, with emissions reduced at the lowest and highest soil moisturecontents. The forest ¯oor Q10of 2.03 (from 15±258C) suggests that CO2emissions were controlled pri-marily by soil biological activity. Forest ¯oor CO2emissions were predicted with a multiple polynomialregression model (r2=0.88) of temperature and moisture, but the ®t predicting mineral soil respiration wasweaker (r2=0.59). Methane uptake was controlled strongly by soil moisture, with reduced ¯uxes underconditions of very low or very high soil moisture contents. A multiple polynomial model accuratelydescribed CH4uptake by mineral soil material (r2=0.81), but only weakly (r2=0.45) predicteduptake by forest ¯oor material. The mineral soil Q10of 1.11 for CH4uptake indicates that methaneuptake is controlled primarily by physical processes. Our work suggests that inclusion of both moistureand temperature can improve predictions of soil CO2and CH4exchanges between soils and the atmos-phere. Additionally, global change models need to consider interactions of temperature and moisture inevaluating eects of global climate change on trace gas ¯uxes. # 1998 Elsevier Science Ltd. All rightsreservedINTRODUCTIONCarbon dioxide and methane are greenhouse gaseswhose atmospheric concentrations have beenincreasing over the last several centuries (Prather etal., 1996; Schimel et al., 1996). Methane is ofincreasing concern because it has a radiative poten-tial 26 times more eective than CO2(Lilieveld andCrutzen, 1992), and because it in¯uences the oxi-dative potential of the atmosphere (Thompson,1992).Soils are major global sources and sinks of CO2and CH4and thus play an important role in regu-lating atmospheric concentrations of these gases(Mooney et al., 1987; Melillo et al. , 1989). Soil res-piration by terrestrial ecosystems contributes 50±75  109tCyÿ1(Kicklighter et al., 1994) to the at-mosphere, which is an order of magnitude greaterthan the annual increase in atmospheric carbon(Watson et al., 1990). Globally, consumption of at-mospheric methane by both methanotrophic andnitrifying bacteria in aerobic soil (Bedard andKnowles, 1989) is 1.5±4.5  107tChaÿ1yÿ1, repre-senting 1 to 8% of the oxidative potential of the at-mosphere (Born et al., 1990; Schutz et al., 1990;Prather et al., 1996). Of this total, temperate ever-green and deciduous forests consume up to 37%(Steudler et al., 1989).Understanding the dynamics of CO2and CH4exchanges between soils and the atmosphere is criti-cal to construction of regional trace gas models andto prediction of trace gas ¯uxes under models ofglobal climate change. Numerous ®eld studies intemperate forests have shown a strong temporalpattern in CO2and CH4¯uxes (e.g. Castro et al.,1994; Peterjohn et al., 1994; Castro et al., 1995)that corresponds strongly with seasonal changes insoil moisture and temperature. Temperature is con-sidered the primary predictor of CO2euxes, but,not surprisingly, moisture also in¯uences soil respir-ation rates (e.g. Wiant, 1967; Groman et al.,1992). Moisture usually exerts strong control overCH4uptake rates, although inclusion of both moist-ure and temperature in models can increase predic-tive capabilities. Lessard et al. (1994) suggested thatthe strong relationship between moisture and CH4uptake may mask relationships between tempera-Soil Biol. Biochem. Vol. 30, No. 12, pp. 1591±1597, 1998# 1998 Elsevier Science Ltd. All rights reservedPrinted in Great Britain0038-0717/98 $19.00 + 0.00PII: S0038-0717(97)00228-9*Author for correspondence.1591ture and uptake, thus it has been dicult to deter-mine the relative importance of these factors.Using ®eld studies to determine the relative im-portance of moisture and temperature in controlling¯ux rates is dicult because soil temperature andmoisture usually covary seasonally in temperateecosystems. Soil temperatures are usually highest bylate summer, but strong evapotranspiration poten-tials usually reduce soil water even if precipitationstays relatively constant. Thus, it is not a straight-forward exercise to determine if maximum rates ofsoil respiration or CH4uptake in late summer, forexample, are due to high temperatures, lower soilmoisture, or an interaction of both factors.Understanding the controls of moisture and tem-perature is important because there is increasingeort to construct regional trace gas models andaccurate models cannot be produced if controllingfactors are not well understood. To overcome limi-tations of ®eld-based data in determining the im-portance of soil moisture and temperature, we did afull-factorial laboratory study to evaluate the rela-tive contribution of each factor, as well as inter-actions, on CO2emissions and CH4uptake by awell-drained temperate forest soil.MATERIALS AND METHODSSite descriptionSoil material was collected at the Harvard ForestLong-Term Ecological Research Site in north-cen-tral Massachusetts, U.S.A., from an approximately50-y-old mixed hardwood stand established follow-ing abandonment from pastures in the early 1900's(Magill et al., 1997). Total biomass in the stand was111 t haÿ1, and was dominated by black oak(Quercus velutina Lam.), which comprises 76% ofthe total basal area. Remaining vegetation includesblack birch (Betula lenta L.), paper birch (Betulapapyrifera March.) and red maple (Acer rubrum L.).Soils at the site are of the Canton series