Repeated fuel reduction burns have little long-term impact on soil greenhouse gas exchange in a dry sclerophyll eucalypt forest

Agricultural and Forest Meteorology 201 (2015) 17–25

Contents lists available at ScienceDirect

Agricultural and Forest Meteorology journal homepage: www.elsevier.com/locate/agrformet

Repeated fuel reduction burns have little long-term impact on soil greenhouse gas exchange in a dry sclerophyll eucalypt forest Benedikt J. Fest a,∗ , Stephen J. Livesley b , Joseph C. von Fischer c , Stefan K. Arndt a a b c

Department of Forest and Ecosystem Science, The University of Melbourne, Melbourne, Victoria, Australia Department of Resource Management and Geography, The University of Melbourne, Melbourne, Victoria, Australia Department of Biology and Graduate Degree Program in Ecology, Colorado State University, Fort Collins, CO, USA

a r t i c l e

i n f o

Article history: Received 7 April 2014 Received in revised form 23 October 2014 Accepted 1 November 2014 Keywords: Prescribed burning Eucalypt forest Soil methane exchange (CH4 ) Soil respiration (CO2 ) Long-term effect Burning frequency

a b s t r a c t Fuel reduction burning is a widespread management tool in fire-tolerant forest systems to mitigate wildfire risk, but has the potential to impact soil greenhouse gas exchange processes. Soil disturbance often alters soil carbon dioxide (CO2 ) and methane (CH4 ) flux; however, the influence of repeated fuel reduction burning upon these flux processes long-term is still not well understood. In this study we measure soil CH4 flux, soil methanotrophic activity and soil CO2 flux in all seasons from March 2009 to February 2011 in three different fire frequency treatments applied to a dry sclerophyll eucalypt forest (Victoria, Australia) for the last 27 years. The low-intensity fire treatments are forest burnt in autumn (i) every 3 years, (ii) every 10 years, and (iii) not burned (since before 1985). Mean soil CO2 emissions were greater in the burnt as compared to un-burnt treatments. In contrast, soil CH4 oxidation did not show a response to repeated burning and there was no statistical difference in soil CH4 flux among treatments. Furthermore, we did not detect changes in the relationships of soil CH4 flux or soil CO2 flux and key environmental controls. Our results indicate that low intensity fuel reduction burns have no cumulative negative impact on biogeochemical processes related to soil respiration or soil CH4 oxidation. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Changes in our climate system are likely to increase the risk of fire events in fire prone ecosystems. In South Eastern Australia increased temperatures and decreases in rainfall are likely to lead to an increase in the fire danger days (Clarke et al., 2011; Hasson et al., 2009) and therefore increase the risk of large-scale catastrophic fires. One management option that potentially reduces the risk of catastrophic fires is fuel reduction burning, where the fine fuel loads are regularly reduced to limit the intensity and severity of future wildfires. Recently, the yearly target for fuel reduction burns on forested public land in Victoria, Australia has been increased from around 100,000 ha per year to around 380,000 ha per year (Parliament of Victoria, 2010), in response to some severe and catastrophic wildfires. This increase in the area of forest burnt will lead to more frequent burning in any given area as compared to the historic forest fire management over the last 30–40 years (Parliament

∗ Corresponding author at: Department of Forest and Ecosystem Science The University of Melbourne 500 Yarra Boulevard, Richmond, 3121 Victoria, Australia. Tel.: +61 3 9035 6967; mobile: +61 407 342 305. E-mail address: [email protected] (B.J. Fest). http://dx.doi.org/10.1016/j.agrformet.2014.11.006 0168-1923/© 2014 Elsevier B.V. All rights reserved.

of Victoria, 2008). However, we have very little information on how increased frequent fuel reduction burning will impact long-term forest soil greenhouse gas exchange. Well drained upland soils in temperate climates are the primary biotic sink for atmospheric methane (CH4 ) as a result of CH4 oxidation by methanotrophic bacteria. Soils in temperate forest ecosystems are estimated to contribute around 50% to this CH4 sink (Dutaur and Verchot, 2007; IPCC, 2007). Major factors regulating and influencing CH4 uptake (negative soil-atmosphere CH4 exchange; FCH4 ) by forest soils include soil moisture, soil temperature, soil pH, soil inorganic nitrogen levels and soil physical structure (Bodelier and Laanbroek, 2004; Smith et al., 2003). A combination of environmental (soil moisture content and soil temperature) and physical (bulk density and soil porosity) factors determine soil diffusivity, a major regulator of CH4 uptake rate (Dorr et al., 1993; King, 1997; von Fischer et al., 2009). Temperate forest ecosystems also sequester, store and release carbon (C) and are therefore of major importance to the global C cycle (IPCC, 2007; Keith et al., 2009; Martin et al., 2007; Schlesinger and Andrews, 2000). In the global terrestrial C cycle, soil CO2 efflux (FCO2 ) is the second largest flux after gross primary productivity (Raich and Schlesinger, 1992). Soil CO2 efflux is the product of autotrophic respiration (plant roots and associated mycorrhizal

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

2. Material and methods 2.1. Site description The study was based within the Victorian ‘Fire Effect Study Areas’ (FESA) the longest ongoing Australian study to investigate the ecological impacts of low intensity fuel reduction fires in dry sclerophyll forests (Tolhurst et al., 1992; Tolhurst, 2003; Tolhurst and Kelly, 2003). The three areas included in this study (Barkstead, Blakeville and Musk Creek) are located in the Wombat State Forest (approximate distance between sites = 10 km) north-west of Melbourne close to the township of Daylesford (Table 1). The forest is dominated by Eucalyptus obliqua (L. Hér), Eucalyptus rubida (H. Deane & Maiden) and Eucalyptus radiata (Sieber ex DC) with an approximate overstorey height of 25–28 m and a basal area ranging from 29 to 43 m2 ha−1 . The climate is classified as between cool-temperate and Mediterranean, as it has cold, wet winters and warm, dry summers (Fig. 1). The annual rainfall range is

200

40

30

100

20

50

10

o

150

C]

Monthly Rainfall Mean Max Temp Mean Min Temp

0

Air temperature [

fungi) and heterotrophic respiration (soil microorganism and soil fauna) (Raich and Schlesinger, 1992). Factors regulating FCO2 are soil temperature, soil water content, soil organic C, fine root density, and nutrient availability (Lloyd and Taylor, 1994; Schlesinger, 1984; Smith et al., 2003). Depending on their intensity, fuel reduction burns have the potential to alter some, if not all, of the factors regulating soilatmosphere exchange of CH4 and CO2 over different timescales (Certini, 2005; Close et al., 2011; Inbar et al., 2014; Raison, 1980; Raison et al., 1986; Smith et al., 2008; Switzer et al., 2012; Williams et al., 2012; Wüthrich et al., 2002). A shift in forest burning policy therefore raises concerns that frequent burning of forest ecosystems may lead to changes in the forest soil source and/or sink strength for greenhouse gases. Studies on the effect of fuel reduction burning on soilatmosphere CH4 and CO2 flux in different ecosystems have so far focused on short- and medium-term responses following single experimental burns. The results from these studies range from an increase in soil CH4 uptake (Jaatinen et al., 2004), temporary or long-term increases in soil CO2 efflux (Jia et al., 2012; Tufekcioglu et al., 2010; Wüthrich et al., 2002), a decrease in soil CH4 uptake (Prieme and Christensen, 1999), a decrease in soil CO2 efflux (Kim et al., 2011; Ryu et al., 2009), or no detectable change (Concilio et al., 2005, 2006; Kim et al., 2011; Meyer et al., 1997). However, the studies above considered the short-term effects of a single burning event and there is currently a lack of studies into the cumulative, long-term effects of repeated fuel reduction burning on soil atmosphere CH4 and CO2 flux. Changes in soil factors such as nutrient availability, pH, soil moisture, soil bulk density and soil organic C content have been linked to burning frequency (Campbell et al., 2008; Scharenbroch et al., 2012; Williams et al., 2012) and it is therefore possible that a long-term increase in the frequency of planned burning may have significant long-term impacts on the greenhouse gas fluxes of CH4 and CO2 between the soil and the atmosphere within our forest systems. In our study we measured soil CH4 flux, methanotrophic activity and soil CO2 flux at forest sites that have been burned at various frequencies of low intensity fire over the last 27 years (planned burning every 3 years, planned burning every 10 years, and unburnt). Our main objectives were to (i) measure any cumulative long-term effects of planned fire regimes on soil CH4 flux, methanotrophic activity and soil CO2 flux in a dry sclerophyll Eucalyptus forest system–the predominant forest type in Victoria (Australia) and (ii) to investigate if the relationship between CH4 flux, or soil CO2 flux, and key environmental/edaphic controls has changed as a consequence of the fire treatment.

Precipitation [mm]

18

0 Jan Feb Mar Apr May Jun

Jul Aug Sep Oct Nov Dec

Month Fig. 1. Climate data 1987–2011 for the Wombat forest in South Eastern Australia. Data were compiled from weather stations at the three FESA’s and surrounding Bureau of Meteorology weather stations.

814–901 mm (1987–2011) and the mean monthly minimum temperatures range from 2 ◦ C (July/August) to 12 ◦ C (January/February; Table 1) with daily summer temperatures often exceeding 35 ◦ C. Mean monthly maximum temperatures range from 7 ◦ C (July) to 24 ◦ C (January/February). The soils in the southern part of the Wombat State Forest derive from weathered sandstone and shale. The surface texture is fine sandy loam to clay loam. Soils are classified as Acidic-mottled, Dystrophic, Yellow Dermosol after Australian soil classification (Robinson et al., 2003). Bulk density at the study sites averaged around 0.75 g cm−3 and pH was 4.8. Following experimental burning treatments are applied in a fully replicated design across all FESAs: - Low intensity burn approximately every 3 years in autumn (AH). - Low intensity burn approximately every 10 years in autumn (AL). - control, fire exclusion (CONTROL). Treatments were established in 1985 and are ongoing, with only a short interruption ‘no burn’ period between 1998 and 2002. The burning treatments at the three chosen sites were last applied between March 2003 and October 2008. More detailed information about treatment history and frequency is presented in Table 1. 2.2. Experimental design Three plots (upslope, mid-slope, bottom of slope), each 25 m2 , were established in each treatment (CONTROL, AH, AL) at each of the three FESA sites (Barkstead, Blakeville and Musk Creek). Beginning in March 2009, we sampled trace gas fluxes with five manual chambers in each plot. When possible we sampled all plots on consecutive days (3–5 days per sampling cycle). We sampled trace gas fluxes seasonally between March 2009 and February 2011, a total of 135 chamber flux measurements per sampling cycle. 2.3. Flux measurements The manual closed chamber method (Hutchinson and Mosier, 1981) was used to quantify the spatial and seasonal variation in soil-atmosphere exchange of CO2 and CH4 . Manual chambers were made of dark PVC inspection pipe fittings (diameter 15 cm, height 15 cm, volume 2.8 L, basal area 0.018 m2 ) with a PVC screw-onlid incorporating a butyl-rubber septum and a rubber O-ring to form a gas tight seal. Chambers were fitted with low voltage fans to ensure good headspace mixing. PVC anchors (height 100 mm), for

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

19

Table 1 Summary of the environment, stand characteristics, soil characteristics, and the fire history for the 3 FESAs in central Victoria, Australia. FESA

Latitude/Longitude Elevation (metre above sea level)a Slope (degree)a Aspect (degree)a Mean annual rainfall (mm)b Mean monthly max temp. (◦ C)b Mean monthly min temp. (◦ C)b Tree mean basal area (m2 ha−1 )a Tree mean height (m)a Last thinning (year)c Last wildfire (year): CONTROLc Total experimental area (ha)c Mean fire interval (years): ALd Mean fire interval (years): AHd

Barkstead

Blakeville

Musk Creek

37◦ 29 S, 144◦ 05 E 635–650 0–4 120–315 901 8–23 2–10 31 28 1979 1931 19 9 (3, 1987–2007) 4 (5, 1987–2007)

37◦ 31 S, 144◦ 05 E 590–665 1–13 130–295 871 9–24 2–10 43 26 1964 1935 81 9.5 (3, 1987–2008) 3 (6, 1987–2007)

37◦ 28 S, 144◦ 10 E 620–720 1–16 40–310 856 10–24 3–11 29 25 1974 1974 78 16 (2, 1987–2004) 4 (6, 1987–2008)

a

This study’s plots were in close vicinity to the measurement plots of Bennett et al. (2013). From automated weather stations within 4 km of each FESA (Barkstead 1986–1999, all others 1986–2002 and 2007–2010). Tolhurst et al. (1992). d Mean interval in years between successive prescribed fires during the experimental period (values in brackets indicate the number of prescribed fires, and the years of first and last prescribed fire per FESA); ‘AH’ autumn High frequency, ‘AL’ autumn Low frequency. e This study n = 9 per site [1SE]. b c

the manual chamber tops to slide on to during measurements, were installed at least three months before the first flux measurements and were left in place over the study period. The litter layer was cut around the circumference of each anchor and the anchors were inserted 40 mm into the soil surface, protruding 60 mm above the soil. Ground vegetation was sparse and so was avoided inside the chambers. The internal height of each manual chamber was measured on each of the six measurement campaigns so as to calculate the headspace volume of each chamber. To enable flux measurements, chamber tops were pushed over the PVC anchors, lids were attached and twist-sealed and four 20 mL headspace gas samples were taken at 10 min intervals starting 3 min after closure using a 20 mL syringe (TerumoTM USA) fitted with a one-way stopcock. 2.4. Gas diffusivity and methanotrophic activity measurements We quantified the separate physical and biological controls of methane uptake using the approach of von Fischer et al. (2009). This approach is based on the premise that methane uptake rates are a function of the soil’s capacity to conduct diffusive gas movement (“diffusivity”, D) and a first-order reaction rate constant referred to as “methanotroph activity” (Mu). To determine soil gas diffusivity, we spiked the chamber headspace with a small volume (2 mL) of the non-reactive tracer gas sulfur hexafluoride (SF6) immediately after lid closure. This established an initial chamber headspace concentration of approximately 6 ppb SF6. Headspace gas samples were then collected as before, at 0, 10, 20 and 30 min after lid closure. We calculated soil diffusivity from these SF6 measures by finding the diffusivity value that gave the best least squared error to the equation of Rolston et al. (1991), after correcting for soil airfilled porosity. An appropriate molecular weight correction was applied to yield a methane diffusivity value. Under the premise that methane uptake is a function of soil diffusivity and methanotroph activity, we used the approach of von Fischer et al. (2009) to find the methanotroph activity value (Mu) that best predicted the chamber headspace methane concentrations, given the measured diffusivity value for that same chamber. 2.5. Gas sample analysis and flux calculation Gas samples were transferred to, and stored in, pressurised 12 mL Exetainers® (Labco Ltd., UK) before being analysed within 7 days using gas chromatography (GC) (Shimadzu GC17A) to

determine CH4 concentrations using a flame ionisation detector (FID) and CO2 concentrations through the addition of a Methaniser (SRI Instruments, USA) before the FID. SF6 concentrations were determined using an electron capture detector (ECD) in the same GC run. Previous tests demonstrated no detectable change in concentrations with storage over this 7 day time period. CH4 fluxes (FCH4 ) and CO2 fluxes (FCO2 ) were calculated by fitting linear regressions to the headspace concentration change of the respective gases with time. Visual screening and statistical analysis showed that a linear regression was best suited to describe temporal changes in chamber head space concentrations over the relatively short incubation period of 30 min. In cases where the R2 of the linear regression model between CO2 concentration and time was <0.9 both the CO2 and CH4 data were excluded from further analysis. The minimum detectable limit (MDL) for this GC/manual chamber system was calculated as: MDL = 2 × SD × V/(A × t), where SD is the standard deviation of the gas concentration in ambient air, V is the volume (litres) of the chamber, A is the area of soil under the chamber (m2 ) and t is the time for incubation (hours). This was then calculated to a flux rate (mg/␮g CO2 /CH4 –C m−2 h−1 ) using the ideal gas law and atom/molecular mass. The MDL CH4 was 2.4 ␮g CH4 –C m−2 h−1 and the MDL for CO2 was 10 mg CO2 –C m−2 h−1 . Fluxes below the MDL of the system were set to zero for statistical analysis. 2.6. Soil and litter properties Periodically surface litter was collected at three randomly selected locations at each plot using a hoop of 0.48 m diameter (0.18 m2 ), the litter was sieved (5 mm) and then air dried (35 ◦ C for 48 h) to determine standing litter dry-mass per m2 (LDM). On all measurement campaigns, composite soil samples (three 0–50 mm samples) were collected from within each plot, sieved (2 mm) and sub-sampled for 1 M KCl extraction (1:4, soil:KCl) and gravimetric water content (GWC) determination (105 ◦ C for 48 h). KCl extracts were shaken for 1 h and then filtered (Whatman 42) and the filtrate frozen prior to analysis for nitrate (NO3 − ) and ammonium (NH4 + ) concentration using an auto-analyser (SFA, TechniconTM ). Concurrent to each chamber measurement soil temperature (TS ) at a depth of 50 mm was measured next to each chamber with a handheld Cole-Palmer® stainless steel temperature probe. Soil moisture (MS ) was measured with a handheld impedance probe

20

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

(Theta Probe ML2, Delta-T Devices, Cambridge, UK) at three points in the top 6 cm next to each individual chamber. Over the course of the measurement period 90 volumetric soil cores (0–50 mm, Ø 72 mm) were sampled at each site to determine volumetric water content (VWC) and soil bulk density (BD). The data were used to establish site dependent calibration curves between the impedance probe readings and volumetric water content (Kaleita et al., 2005). These bulk density and volumetric water content data and their relationship to impedance probe readings were further used to calculate soil porosity (ϕ), air filled porosity (ϕair ) and percentage water filled pore space (%WFPS) for each plot and measuring campaign according to Loveday and Commonwealth Bureau of Soils (Great Britain) (1973). At the end of the study, a composite soil sample from three soil cores was collected at 0–50 mm at each plot, air dried, sieved (2 mm) and analysed for soil particle size analysis through dispersion, suspension, settling and sequential hydrometer readings (Ashworth et al., 2001). A sub-sample of each air-dried soil was analysed for pH (1:5, soil:water) and for total C and N content using an elemental analyser (LECO® ). 2.7. Data analysis We used linear mixed model (LMM) procedures in Genstat 14.0 (VSN International, UK) to test for main effects and interactions of burning frequency (CONTROL, AH, AL), measuring event and site (Barkstead, Blakeville and Musk Creek) and their effect on soil CH4 flux, soil microbial activity, soil CO2 flux and measured soil environmental variables. Main effects were considered significant if p ≤ 0.05, and interactions were considered significant at p ≤ 0.01. The random model represented the nested design structure of the experiment with burning treatments blocked within site within each measuring event. Visual inspection of residual histograms, fitted-value plots and half normal plots showed that the data fulfilled the assumption of normal distribution. To investigate seasonal relationships between measured soil environmental variables and soil CH4 flux, soil methanotrophic activity and CO2 flux we used multiple linear regression procedures and stepwise linear regression procedures (SPSS 20, IBM, USA) treating individual measuring events and sites and burn treatments as replicates. We first ran stepwise liner regression procedures as an explorative tool to identify significant predictors and predictor combinations and then tested these using simple or multiple linear regression models. Of the four predictors expressing soil moisture levels (volumetric moisture content, air filled porosity, water filled pore space and gravimetric moisture content) we report the model with the highest coefficient of determination including only one of these predictors. We transformed data when necessary to reduce heteroscedasticity for linear regression analysis. Specific soil environmental variables, only measured once (soil particle size, pH, EC, total carbon, total N, C/N, bulk density), were compared between treatments using a separate LMM procedure in Genstat 14.0. The treatment dependent soil temperature response of soil CO2 flux was further analysed by fitting two different temperature functions to the measured data. An Arrhenius type function (Eq. (1)) as described by Lloyd and Taylor (1994) and the Q10 function (Eq. (2)), both of which are widely used describing the exponential FCO2 to temperature relationship:



FCO2 = Rref × FT





1 1 − (Tref −T0 ) (TS − T0 ) T − T  S ref with FT = Rref × Q10 × exp 10

FCO2 = Rref × FT with FT = exp E0 ×

(1) (2)

where TS is the measured soil temp (K), Tref is the reference temperature (set to the mean of the dataset = 286.5 K) and T0 = 227.13 K

after Lloyd and Taylor (1994). All other model parameters were fitted by the non-linear regression feature in SPSS 20. Rref is the soil respiration rate at the reference temperature. E0 is an exponential parameter affecting the temperature sensitivity of FCO2 . In a second step we multiplicatively added a Gompertz type soil moisture function (Eq. (3)) to account for soil moisture sensitivity of FCO2 following the approach of Subke et al. (2003): FCO2 = Rref × FT (Eq. (1) or (2)) × FSWC with FSWC = e × exp(−e × exp((a − b × VWCS )))

(3)

where a and b are both data specific constants. 2.8. Normalisation of soil respiration As soil respiration in the three burning treatments per site was measured at different times of the day (11 am–1 pm), temperature differences amongst the treatment measured first and the treatment measured last at a given day differed by up to 1.5 ◦ C. To correct for any systematic error associated with these temperature differences, soil respiration rates were normalised to the mean daily soil temperature using Q10 -values obtained from Eq. (3) for each site and treatment. This involved a slight modification of the Q10 -function itself: Rnorm = R × Q10 × exp((Tdaymean − TS )/10), where R was the measured soil respiration rate (␮mol CO2 m−2 s−1 ), Ts the measured soil temperature (◦ C) and Tdaymean the average daily soil temperature (◦ C) during measurement period. This correction did not alter the seasonal pattern and had marginal impact upon the magnitude of soil respiratory fluxes. 3. Results The forest soil at all sites and under all treatments acted as a CH4 sink and a CO2 source over the measurement timeframe (Fig. 2), and the inorganic nitrogen pool in the soil was low and generally dominated by ammonium (Table 2). The LMM analysis showed that two way interactions between site and measuring campaign were significant for soil temperature (F10, 108 = 21.46; p < 0.001), soil volumetric water content (F10 ,108 = 9.33; p < 0.001), soil air filled porosity (F10 ,108 = 9.4; p < 0.001), soil gravimetric water content (F10 ,108 = 4.01; p < 0.001), soil ammonium content (F10 ,108 = 2.62; p = 0.007) and soil nitrate content (F10 ,108 = 2.66; p = 0.006). Furthermore, significant two way interactions between site and treatment existed for soil temperature (F4 ,108 = 23.25; p < 0.001) and soil ammonium content (F4 ,108 = 5.88; p < 0.001). These variables were therefore included alone and in combination as covariates in the LMM to determine treatment effects on soil CH4 flux soil methanothrophic activity and soil CO2 flux (see Sections 3.1, 3.2 and 3.3). Furthermore, the LMM analysis showed a significant three way interaction between site, treatment and measuring campaign for soil temperature (F20 ,108 = 6.46, p < 0.001). 3.1. Mean effect of burning frequency on soil CH4 flux Mean CH4 flux rates (plot level) varied between −87.0 to −10.1 ␮g CH4 –C m−2 h−1 , from which we calculated predicted treatment means and standard errors (SE) via the LMM procedure (Fig. 2). Treatment mean FCH4 values varied little across treatments and there was no significant difference in FCH4 between the control, the high and the low burning frequency regardless whether soil temperature (TS ) was included as a covariate (F2 , 88 = 1.59; p = 0.209) or not (F2 , 89 = 1.5; p = 0.229). Furthermore, the introduction of soil ammonium as a covariate to the fixed model did not change the significance level of the treatment effect (F2 ,88 = 1.23; p = 0.296) the same was observed when soil temperature and soil

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

21

Table 2 Predicted treatment means and standard errors [1SE] of measured soil variables for the different burning frequency categories. Capitals in superscripts (AB ) mark significant differences between treatments AH (autumn burn every 3 years), AL (autumn burn every 10 years) and CONTROL (long-term unburnt) as indicated by post hoc pair wise comparison (LSD ≤ 5%). GWC (gravimetric soil water content), ϕair (air filled porosity), VWC (volumetric soil water content), NH4 + (soil ammonium concentration), NO3 − (soil nitrate concentration). Treatment AH Mean

AL SE

Mean

CONTROL SE

Mean

SE

Soil variables (0–50 mm) ␸air (cm3 cm−3 ) VWC (cm3 cm−3 ) WFPS (%) GWC (g g−1 ) NO3 − (mg kgsoil −1 of N) NH4 + (mg kgsoil −1 of N)

0.427 0.225 33.36 0.236A,B 0.016 2.37

[0.02] [0.01] [1.39] [0.01] [0.01] [0.82]

0.426 0.226 33.30 0.262A 0.019 2.86

[0.02] [0.01] [1.39] [0.01] [0.01] [0.78]

0.435 0.218 31.77 0.243B 0.017 2.53

[0.02] [0.01] [1.39] [0.01] [0.01] [0.72]

Clay (%) Silt (%) Sand (%) Total nitrogen (N, %) Total carbon (C, %) C/N ratio EC (mS) pH Bulk density (g cm−3 ) Soil porosity (cm3 cm−3 )

20.23 19.41 60.33 0.17 4.74 28.14 0.07 4.80 0.86 0.68

[1.16] [3.31] [3.46] [0.02] [0.11] [2.11] [0.01] [0.14] [0.05] [0.02]

21.86 25.09 53.06 0.19 5.50 29.30 0.07 4.88 0.85 0.68

[1.57] [0.56] [1.40] [0.03] [0.59] [0.86] [0.00] [0.06] [0.04] [0.02]

20.00 25.58 54.42 0.18 5.29 28.67 0.08 4.72 0.83 0.69

[1.26] [1.51] [1.03] [0.01] [0.50] [2.09] [0.00] [0.06] [0.03] [0.01]

361.26

[39.4]

362.15

[39.4]

437.76

[39.4]

Litter DW (g m−2 )

ammonium content were both added to the model as covariates (F2 ,87 = 1.25; p = 0.290). 3.2. Mean effect of burning frequency on soil methanotrophic activity Mean methanotrophic activity (Mu) varied between 0.0001 and 0.1 min−1 , from which we calculated predicted treatment means and standard errors (SE) via the LMM procedure (Fig. 2). The burn treatment effect was not significant on Mu (F2 , 89 = 2.77; p = 0.068). The significance level of the main effect did not change regardless of whether soil temperature or soil ammonium were added as single covariates or in combination to the fixed model (data not shown). 3.3. Mean effect of burning frequency on soil CO2 flux Mean soil CO2 flux (FCO2 ) varied between 38.6 and 252.9 mg CO2 –C m−2 h−1 at the plot level, from which we calculated treatment means and standard errors (SE) using the LMM procedure (Fig. 2). There was a significant effect of burning frequency on FCO2 (F2 ,107 = 3.84; p = 0.024), with FCO2 in AL being significantly greater than that in the CONTROL. The significance level of the main effect did not change regardless of whether soil temperature or soil ammonium were added as single covariates or in combination (data not shown). 3.4. Mean effect of burning frequency and disturbance on soil structural and environmental variables Gravimetric water content (GWC) was significantly higher in AL than the CONTROL (F2 ,108 = 3.61, p = 0.03). Soil bulk density, soil porosity, and other measured structural soil parameters were not significantly different among the three fire frequency treatments (Table 2). We did not observe a significant treatment effect on soil volumetric water content or ϕair . However, there was a trend towards lower soil moisture levels in the CONTROL compared to the two burning treatments (Table 2). Furthermore, soil CH4 diffusivity did not differ significantly among treatments (Fig. 2). Soil NH4 + and soil NO3 − contents were in general very low and did not differ with regards to burning treatment (Table 2).

3.5. Soil environmental drivers of CH4 flux Treatment specific linear regression analyses showed that ϕair explained 68% of the variability in FCH4 for treatment AH (F = 28.1, p < 0.001), 84% of the variably in FCH4 for treatment AL (F = 76.7, p < 0.001), and 84% of the variably in FCH4 for the CONTROL (F = 74.7, p < 0.001) (Table 3). Adding TS as a predictor to the linear regression model did not improve the outcome for any treatment category and TS was not a significant coefficient. The slopes of the linear regression models did not differ significantly among the treatments. 3.6. Soil environmental drivers of soil CO2 efflux Treatment specific linear regression analyses showed that TS explained 42% of the variability in FCO2 for treatment AH (F = 12.7, p = 0.003), 36% of the variably in FCO2 for treatment AL (F = 10.6, p = 0.005), and 46% of the variably in FCO2 for the CONTROL (F = 15.8, p = 0.001) (Table 3). GWC alone explained 29% of the variability in FCO2 for treatment AH (F = 7.7, p = 0.014), 43% of the variably in FCO2 for treatment AL (F = 14.3, p = 0.002) and 27% of the variably in FCO2 for treatment CONTROL (F = 7.2, p = 0.016) (Table 3). Combined, TS and GWC explained 70% of the variability in FCO2 for treatment AH (F = 20.3, p < 0.001), 77% for treatment AL (F = 29.4, p < 0.001), and 80% for the CONTROL (F = 35.7, p < 0.001) (Table 3). Adding more predictors to the linear regression models did not improve their predictive power. Further statistical analysis showed that the slopes of the linear regression models including either TS or GWC or both predictors together did not differ significantly among the fire frequency treatments (data not shown). Table 4 shows the result of applying different soil temperature and soil moisture functions. The Lloyd and Taylor equation generally performed slightly better in explaining the FCO2 to soil temperature relationship than the Q10 equation, with a coefficient of determination of 0.45 for AH, 0.39 for AL, and 0.48 for the CONTROL. When Eq. (3) was added to the model the coefficient of determination for the Lloyd and Taylor model increased to 0.82 for AH, 0.86 for AL, and 0.90 for the CONTROL, indicating a strong relationship between FCO2 and VWC for all treatments. E0 values showed a similar trend to Q10 values as they were least in the

22

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

Table 3 Parameters and coefficients of determination (Adj. R2 ) of treatment separated linear regression models explaining seasonal variability in soil CH4 flux (FCH4 ) and soil CO2 flux (FCO2 ). Unstandardised and standardised coefficients ˇ (in parenthesis); SD refers to standard deviation of parameter; level of significance as indicated by ANOVA (*≤0.05, **≤0.01, ***≤0.001). Predictors: TS (Soil temperature) ϕair (soil air filled porosity) and GWC (gravimetric water content). Treatment

Dependent variable

CONTROL

FCH4 (SD = 12.55)

FCO2 (SD = 35.4)

AL

FCH4 (SD = 14.20)

FCO2 (SD = 33.6)

AH

FCH4 (SD = 14.66)

FCO2 (SD = 33.4)

Constant −26.771* 0.196 5.611

TS (SD = 4.1) −0.944 (−0.327) – −0.484 (−0.159)

ϕair (SD = 0.11) – −98.69*** (−0.923) −95.49*** (−0.893)

Adj. R2 0.038 0.840*** 0.856***

Constant 26.858 53.763* −28.603

TS (SD = 4.1) 6.034*** (0.706) – 6.135*** (0.717)

GWC (SD = 0.09) – 219.763* (0.559) 225.549*** (0.574)

Adj. R2 0.466*** 0.269** 0.804***

Constant −30.445 6.478 7.408

TS (SD = 3.8) −0.395 (−0.109) – −0.073 (−0.02)

ϕair (SD = 0.12) – −105.44*** (−0.925) −105.217*** (−0.923)

Adj. R2 −0.064 0.844*** 0.832***

Constant 46.032 61.952** −1.446

TS (SD = 3.8) 5.589** (0.632) – 5.065*** (0.572)

GWC (SD = 0.09) – 231.026** (0.687) 213.019*** (0.633)

Adj. R2 0.361** 0.439** 0.770***

Constant −33.425* 1.847 6.919

TS (SD = 3.9) −0.290 (−0.084) – −0.367 (−0.107)

ϕair (SD = 0.12) – −100.025*** (−0.835) −100.370*** (−0.838)

Adj. R2 −0.076 0.673*** 0.656***

Constant 43.010 69.826** 3.250

TS (SD = 3.9) 5.726** (0.678) – 5.385*** (0.638)

GWC (SD = 0.09) – 199.267** (0.583) 182.783*** (0.534)

Adj. R2 0.424** 0.295** 0.707***

Table 4 Parameters ± SE and coefficients of determination (R2 ) for soil temperature and soil moisture functions to explain seasonal variability in CO2 flux (FCO2 ) observed across all measuring campaigns and planned burning treatments. Rref (Reference Respiration, mg CO2 –C m−2 h−1 ) at 13.3 ◦ C, E0 (Activation energy in Kelvin), a and b are dimensionless. Coefficient of determination adjusted for n = 54 for overall, n = 18 for AH, AL. Moisture function: ‘None’ indicates that the used model relied only on the temperature dependency of soil CO2 flux; ‘Gompertz’ indicates that an additional Gompertz function was included in the model to account for moisture dependency of soil CO2 flux (Subke et al., 2003). Temperature function

Moisture function

Lloyd and Taylor (1994)

None

Rref E0 R2

104.45 ± 6.45 196.53 ± 55.3 0.48

118.47 ± 6.62 159.71 ± 54.51 0.39

116.936 ± 6.764 169.558 ± 52.43 0.45

Gompertz

Rref E0 a b R2

131.49 ± 10.07 232.02 ± 26.18 0.891 ± 0.47 11.106 ± 4.08 0.90

137.31 ± 8.33 193.13 ± 32.10 0.887 ± 0.65 13.265 ± 5.75 0.86

130.65 ± 6.77 201.11 ± 33.9 1.61 ± 1.09 20.61 ± 9.48 0.82

None

Rref Q10 R2

103.10 ± 7.09 1.71 ± 0.27 0.47

117.45 ± 7.27 1.54 ± 0.26 0.37

115.67 ± 6.99 1.61 ± 0.26 0.43

Gompertz

Rref Q10 a b R2

134.47 ± 11.79 1.90 ± 0.15 0.84 ± 0.47 10.46 ± 4.15 0.89

139.28 ± 9.09 1.71 ± 0.16 0.85 ± 0.65 12.79 ± 5.79 0.85

133.00 ± 7.59 1.75 ± 0.17 1.35 ± 1.00 18.25 ± 8.67 0.80

Q10

Treatment CONTROL

high burning frequency treatment and greatest in the CONTROL (Table 4). 4. Discussion 4.1. Soil CH4 exchange and methanotrophic activity The absence of any significant effect from planned burning on FCH4 is consistent with some previous reported results (Kim et al., 2011; Meyer et al., 1997). However, our study is the first to investigate and report that frequent fuel reduction burning had no long-term or cumulative effect on forest soil FCH4 . Furthermore, our

AL

AH

results show that repeated fuel reduction burning also had no longterm effect on soil variables that influence soil diffusivity such as air filled porosity, or soil bulk density, soil porosity and clay content (Table 2). The soils at all three sites were sinks for atmospheric CH4 . The mean soil CH4 uptake rates reported here are within the range reported for other temperate eucalypt forest studies in Australia (Dalal et al., 2008; Fest et al., 2009; Livesley et al., 2009; Meyer et al., 1997) and are comparable with similar studies worldwide (Borken and Brumme, 1997; Butterbach-Bahl, 2002; Butterbach-Bahl et al., 2002a,b; Price et al., 2003). Burning treatment did not significantly affect mean methanotrophic activity or methane diffusivity (Fig. 2).

160

A

140

b

ab

a

120 100

-20

60

B -30

-1

Methanotrophic Activity [min ]

-40

0.06

-50

CH4 flux ug [CH 4 -C m -2 h-1 ]

80

C

0.05

0.04

0.03

0.02

2.0

D 1.5

1.0

0.5 CONTROL

AL

CH4 diffusivity [cm 2 min -1 ]

CO2 flux [mg CO 2-C m -2 h-1 ]

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

AH

Burning treatment Fig. 2. Linear mixed model treatment means and error bars (1xSE) for soil respiration CO2 flux (A), CH4 flux (B), methanotrophic activity (C) and soil CH4 diffusivity (D). Letters above bars (ab) mark significant differences between treatments AH (autumn burn every 3 years), AL (autumn burn every 10 years) and CONTROL (longterm unburnt) as indicated by post hoc pair wise comparison (LSD ≤ 5%).

This indicates that the burning treatments had no major, long-term negative implications for soil methanotrophic bacteria communities, even though in other studies short-term losses of up to 50% topsoil microbial biomass has been reported following low intensity planned burning (Campbell et al., 2008). The regression analysis showed that FCH4 was highly influenced by soil moisture levels in this forest system. Soil moisture alone, expressed as ϕair , explained up to 92% of the seasonal variability in FCH4 across the treatments. The finding that the relationship between ϕair and FCH4 did not change as a result of the fire treatment further confirms the absence of a long-term effect from prescribed burning on soil FCH4 . 4.2. Soil CO2 flux Our study is the first to report a cumulative effect from frequent planned burning which increases soil CO2 efflux in cool temperate eucalypt forests. The mean soil CO2 flux values are within

23

the reported range for other temperate eucalypt forest systems in Australia (Ellis, 1969; Fest et al., 2009; Keith and Wong, 2006; Kirschbaum et al., 2007; Livesley et al., 2009; Meyer et al., 1997) and are comparable with temperate forest systems worldwide (Raich and Schlesinger, 1992; Schlesinger, 1977; Smith et al., 2010; Subke et al., 2003). Mean soil CO2 flux (FCO2 ) under the two burning treatments was around 9–10% (AH) and 13–14% (AL) higher than the CONTROL treatment. Our results of higher soil CO2 flux rates in the two burning treatments agree with those reported by Knapp et al. (1998), Tufekcioglu et al. (2010), Xu and Wan (2008) and Jia et al. (2012). Possible reasons for the increase in soil CO2 flux rates after repeated burning are: (1) an increase in decomposition rates due to higher inputs of easily decomposable compounds from dead plant residues within, and at, the soil surface (Wüthrich et al., 2002), (2) higher surface soil temperatures caused by decreased plant canopy cover resulting in higher biological activity in the soil, (3) a long-term reduction of available soil nutrients leads to an increase in C allocation below ground and therefore the root to shoot ratio of plants (Abbott and Loneragan, 1983; Brassard et al., 2009) or (4) a change in vegetation density or growth and therefore root biomass and root respiration. There were no short-term increases in soil CO2 flux directly after a fire event in similar forest systems (Meyer et al., 1997). Similarly, Wüthrich et al. (2002) observed no increase in soil CO2 flux immediately after lower intensity burns as compared to when soils experienced fires of greater intensity. Therefore, changes in heterotrophic soil respiration are unlikely to explain the observed differences between soil CO2 flux in burning treatments and the control treatment. As the planned burns in our study were of a low intensity, a long-term increase in soil temperature is rather unlikely since overstorey tree species did not die or lose canopy cover (Chatto et al., 2003; Tolhurst et al., 1992). One possible cause of the long-term increase in soil CO2 flux rates in the burning treatments could be an increase in fine root biomass, either in response to soil nutrient depletion (Abbott and Loneragan, 1983) or simply from an increase in vegetation density. Other studies on the long-term effect of repeated fuel reduction burning have reported decreases in nitrogen availability, soil organic nitrogen and lower microbial biomass (Ojima et al., 1990, 1994; Wilson et al., 2002). A study into soil C, N and P levels at the FESA study sites indicated that P did not change with burning treatment, but that the soil C/N ratio increased even though total soil C and N decreased with increased burning frequency (Hopmans, 2003). Trees in forests of the northern hemisphere (Brassard et al., 2009) allocate more C belowground, and reduce shoot and foliar growth, when soil nutrients become limiting. This is partially supported in the FESA study by the observation that aboveground growth of large trees was reduced in some of the burning treatments as compared to that in the CONTROL treatments (Bennett et al., 2013). However, additional research into fine root mass and dynamics would be needed to verify this theory. Furthermore, the CO2 flux in the AL treatment was 121.3 ± 3.7 mg CO2 –C m2 h−1 as compared to 107.8 ± 3.6 mg CO2 –C m2 h−1 in the CONTROL, so whilst this was statistically significant, the long-term impact may be rather minor. Generally, soil CO2 flux is regulated by soil temperature and moisture levels in this forest system, with soil temperature able to explain 40% of seasonal variability and soil temperature and moisture combined able to explain up to 90% of the variation in soil CO2 flux. The result that relationship between Ts , GWC and FCO2 did not change significantly in response to the fire treatment further

24

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25

supports the conclusion that prescribed burning has no long-term effect on soil microbial processes.

5. Conclusion Repeated fuel reduction burning had no long-term cumulative effect on soil CH4 uptake. Our results show that frequent planned burning has a small cumulative effect on soil CO2 flux in temperate eucalypt forest systems and that frequently burned sites had slightly greater soil CO2 effluxes. Overall our results indicate that repeated low intensity fuel reduction burning as a management tool to mitigate wildfire risks has no cumulative negative longterm impact on the biochemical processes related to soil respiration or soil CH4 oxidation. Hence, while there is evidence that burning events can alter soil biogeochemical processes and soil environmental variables, our results indicate that frequent, low intensity fuel reduction burns have little effect on dry sclerophyll eucalypt forest soils. However, frequent low intensity fuel reduction burning can reduce carbon stocks in standing and dead biomass (Bennett et al., 2013) and also in coarse woody debris carbon stocks and attributes (Aponte et al., 2014). Hence, increases in the frequency of low intensity fuel reduction burns may not alter biogeochemical processes in the long-term but can still change important forest attributes.

Acknowledgements The study was supported by funding from the Terrestrial Ecosystem Research Network (TERN) Australian Supersite Network, the TERN OzFlux Network, the Australian Research Council (ARC) grants LE0882936 and DP120101735 and the Integrated Forest Ecosystem Research (iFER) program, funded by the Victorian Department of Environment and Primary Industries (DEPI). We would like to thank Dr. Kevin Tolhurst for his support in allowing us access to the FESA sites and background information. We also would like to thank the many internship students, especially from the Institut Polytechnique LaSalle Beauvais, Nina Hinko-Najera and Catherine Nield-Fest who helped us with field data collection and in the laboratory. A special thanks also to Dr. Ian Gordon and Rachel Sore for their advice in statistical analyses.

References Abbott, I., Loneragan, O., 1983. Influence of fire on growth rate, mortality and butt damage in Mediterranean forest of Western Australia. For. Ecol. Manage. 6 (2), 139. Aponte, C., Tolhurst, K.G., Bennet, L.T., 2014. Repeated prescribed fires decrease stocks and change attributes of coarse woody debris in a temperate eucalypt forest. Ecol. Appl. 24, 976–989. Ashworth, J., Keyes, D., Kirk, R., Lessard, R., 2001. Standard procedure in the hydrometer method for particle size analysis. Commun. Soil Sci. Plant Anal. 32 (5–6), 633–642. Bennett, L.T., Aponte, C., Tolhurst, K.G., Löw, M., Baker, T.G., 2013. Decreases in standing tree-based carbon stocks associated with repeated prescribed fires in a temperate mixed-species eucalypt forest. For. Ecol. Manage. 306, 243–255. Bodelier, P.L.E., Laanbroek, H.J., 2004. Nitrogen as a regulatory factor of methane oxidation in soils and sediments. FEMS Microbiol. Ecol. 47 (3), 265–277. Borken, W., Brumme, R., 1997. Liming practice in temperate forest ecosystems and the effects on CO2 , N2 O and CH4 fluxes. Soil Use Manage. 13 (4), 251–257. Brassard, B.W., Chen, H.Y.H., Bergeron, Y., 2009. Influence of environmental variability on root dynamics in northern forests. Crit. Rev. Plant Sci. 28 (3), 179–197. Butterbach-Bahl, K., 2002. CH4 . In: Gasche, R., Papen, H., Rennenberg, H. (Eds.), Trace Gas Exchange in Forest Ecosystems. Kluwer Academic Publishers, Dordrecht, pp. 141–156. Butterbach-Bahl, K., Breuer, L., Gasche, R., Willibald, G., Papen, H., 2002a. Exchange of trace gases between soils and the atmosphere in Scots pine forest ecosystems of the northeastern German lowlands 1. Fluxes of N2 O, NO/NO2 and CH4 at forest sites with different N-deposition. For. Ecol. Manage. 167 (1–3), 123–134. Butterbach-Bahl, K., Rothe, A., Papen, H., 2002b. Effect of tree distance on N2 O and CH4 -fluxes from soils in temperate forest ecosystems. Plant Soil 240 (1), 91–103.

Campbell, D.C., Cameron, M.C., Bastias, A.B., Chen, C., Cairney, W.G.J., 2008. Long term repeated burning in a wet sclerophyll forest reduces fungal and bacterial biomass and responses to carbon substrates. Soil Biol. Biochem. 40, 2246–2252. Certini, G., 2005. Effects of fire on properties of forest soils: a review. Oecologia 143 (1), 1–10. Chatto, K., Kellas, J.D., Bell, T.L., 2003. Effects of repeated low-intensity fire on tree growth and bark in a mixed eucalypt foothill forest in south-eastern Australia. Research report no. 66. Deptartment of Sustainability and Environment, East Melbourne, Victoria, Australia. Clarke, H.G., Smith, P.L., Pitman, A.J., 2011. Regional signatures of future fire weather over eastern Australia from global climate models. Int. J. Wildland Fire 20 (4), 550–562. Close, D.C., Davidson, N.J., Swanborough, P.W., Corkrey, R., 2011. Does low-intensity surface fire increase water- and nutrient-availability to overstorey Eucalyptus gomphocephala? Plant Soil 349 (1–2), 203–214. Concilio, A., et al., 2005. Soil respiration response to prescribed burning and thinning in mixed-conifer and hardwood forests. Can. J. For. Res. 35 (7), 1581–1591. Concilio, A., Ma, S., Ryu, S.-R., North, M., Chen, J., 2006. Soil respiration response to experimental disturbances over 3 years. For. Ecol. Manage. 228 (1–3), 82– 90. Dalal, R.C., Allen, D.E., Livesley, S.J., Richards, G., 2008. Magnitude and biophysical regulators of methane emission and consumption in the Australian agricultural, forest, and submerged landscapes: a review. Plant Soil 309 (1–2), 43–76. Dorr, H., Katruff, L., Levin, I., 1993. Soil texture parameterization of the methane uptake in aerated soils. Chemosphere 26 (1–4), 697–713. Dutaur, L., Verchot, L.V., 2007. A global inventory of the soil CH4 sink. Glob. Biogeochem. Cycles 21, GB4013. Ellis, R.C., 1969. The respiration of the soil beneath some Eucalyptus forest stands as related to the productivity of the stands. Aust. J. Soil Res. 7, 349–357. Fest, B.J., Livesley, S.J., Drösler, M., van Gorsel, E., Arndt, S.K., 2009. Soil-atmosphere greenhouse gas exchange in a cool, temperate Eucalyptus delegatensis forest in south-eastern Australia. Agric. For. Meteorol. 149 (3–4), 393–406. Hasson, A., Mills, G., Timbal, B., Walsh, K., 2009. Assessing the impact of climate change on extreme fire weather events over southeastern Australia. Clim. Res. 39 (2), 159–172. Hopmans, P., 2003. Effects of repeated low-intensity fire on carbon, nitrogen and phosphorus in the soils of a mixed eucalypt foothill forest in south-eastern Australia. Research report no. 60. Deptartment of Sustainability and Environment, East Melbourne, Victoria, Australia. Hutchinson, G.L., Mosier, A.R., 1981. Improved soil cover method for field measuremet of nitrous-oxide fluxes. Soil Sci. Soc. Am. J. 45 (2), 311–316. Inbar, A., Lado, M., Sternberg, M., Tenau, H., Ben-Hur, M., 2014. Forest fire effects on soil chemical and physicochemical properties, infiltration, runoff, and erosion in a semiarid Mediterranean region. Geoderma 221–222, 131–138. IPCC., 2007. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: The Scientific Basis. Contribution of Working Group I to the Fourth Assessment Report of the intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom/New York, NY, USA, p. 996. Jaatinen, K., Knief, C., Dunfield, P.F., Yrjala, K., Fritze, H., 2004. Methanotrophic bacteria in boreal forest soil after fire. FEMS Microbiol. Ecol. 50 (3), 195–202. Jia, X., Shao, M.a, Wei, X., 2012. Responses of soil respiration to N addition, burning and clipping in temperate semiarid grassland in northern China. Agric. For. Meteorol. 166–167, 32–40. Kaleita, A.L., Heitman, J.L., Logsdon, S.D., 2005. Field calibration of the theta probe for Des Moines lobe soils. Appl. Eng. Agric. 21 (5), 865–870. Keith, H., Wong, S.C., 2006. Measurement of soil CO2 efflux using soda lime absorption: both quantitative and reliable. Soil Biol. Biochem. 38 (5), 1121–1131. Keith, H., Mackey, B.G., Lindenmayer, D.B., 2009. Re-evaluation of forest biomass carbon stocks and lessons from the world’s most carbon-dense forests. Proc. Natl. Acad. Sci. U. S. A. 106 (28), 11635–11640. Kim, Y.S., et al., 2011. Greenhouse gas emissions after a prescribed fire in white birch-dwarf bamboo stands in northern Japan, focusing on the role of charcoal. Eur. J. For. Res. 130 (6), 1031–1044. King, G.M., 1997. Responses of atmospheric methane consumption by soils to global climate change. Glob. Change Biol. 3 (4), 351–362. Kirschbaum, M.U.F., et al., 2007. Modelling net ecosystem carbon and water exchange of a temperate Eucalyptus delegatensis forest using multiple constraints. Agric. For. Meteorol. 145 (1/2), 48. Knapp, A.K., Conard, S.L., Blair, J.M., 1998. Determinants of soil CO2 flux from a subhumid grassland: effect of fire and fire history. Ecol. Appl. 8 (3), 760–770. Livesley, S.J., et al., 2009. Soil-atmosphere exchange of greenhouse gases in a Eucalyptus marginata woodland, a clover-grass pasture, and Pinus radiata and Eucalyptus globulus plantations. Glob. Change Biol. 15 (2), 425–440. Lloyd, J., Taylor, J.A., 1994. On the temeprature dependence of soil respiration. Funct. Ecol. 8 (3), 315–323. Loveday, J., Commonwealth Bureau of Soils (Great Britain), 1973. Methods for Analysis of Irrigated Soils. Commonwealth Agricultural Bureaux, Farnham Royal, Buckinghamshire, pp. 208. Martin, D., Beringer, J., Hutley, L.B., McHugh, I., 2007. Carbon cycling in a mountain ash forest: analysis of below ground respiration. Agric. For. Meteorol. 147 (1–2), 58–70. Meyer, C.P., et al., 1997. The enhanced emission of greenhouse gases from soil following prescribed burning in a southern eucalyptus forest. Final report to the National Greenhouse Gas Inventory Committee. CSIRO, Division of Atmospheric Research, Aspendale, Victoria, pp. 1–66.

B.J. Fest et al. / Agricultural and Forest Meteorology 201 (2015) 17–25 Ojima, D.S., Parton, W.J., Schimel, D.S., Owensby, C.E., 1990. Simulated Impacts of Annual Burning on Prairie Ecosystems, Fire in North American Prairies. Oklahoma Press, Norman, Okla, pp. 118–132. Ojima, D.S., Schimel, D.S., Parton, W.J., Owensby, C.E., 1994. Long- and short-term effects of fire on nitrogen cycling in tallgrass prairie. Biogeochemistry 24 (2), 67–84. Parliament of Victoria, 2008. Inquiry into the impact of public land management practices on bushfires in Victoria and management practices on bushfires in Victoria: report of the Natural Resources Committee on the inquiry into the impact of public land management practices on bushfires in Victoria/Environment and Natural Resources Committee. Parliamentary paper no. 116, session 2006–2008. Government Printer, Melbourne. Parliament of Victoria, 2010. 2009 Victorian Bushfires Royal Commission: Final Report. The Commission, Melbourne, Victoria, Australia. Price, S.J., et al., 2003. Pristine New Zealand forest soil is a strong methane sink. Glob. Change Biol. 10 (1), 16–26. Prieme, A., Christensen, S., 1999. Methane uptake by a selection of soils in Ghana with different land use. J. Geophys. Res. Atmos. 104 (D19), 23617–23622. Raich, J.W., Schlesinger, W.H., 1992. The global carbon-dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus Ser. B: Chem. Phys. Meteorol. 44 (2), 81–99. Raison, R.J., 1980. A review of the role of fire in nutrient cycling in Australian native forests, and of methodology for studying the fire-nutrient interaction. Aust. J. Ecol. 5, 15–21. Raison, R.J., Woods, P.V., Jakobsen, B.F., Bary, G.A.V., 1986. Soil temperatures during and following low-intensity prescribed burning in a Eucalyptus pauciflora forest. Aust. J. Soil Res. 24 (1), 33–47. Robinson, N., et al., 2003. A Land Resource Assessment of the Corangamite Region. Primary Industries Research Victoria, Bendigo, Victoria. Rolston, D.E., Glauz, R.D., Grundmann, G.L., Louie, D.T., 1991. Evaluation of an insitu mehtod for measurement of gas diffusivity in surface soils. Soil Sci. Soc. Am. J. 55 (6), 1536–1542. Ryu, S.-R., Concilio, A., Chen, J., North, M., Ma, S., 2009. Prescribed burning and mechanical thinning effects on belowground conditions and soil respiration in a mixed-conifer forest California. For. Ecol. Manage. 257, 1324–1332. Scharenbroch, B.C., Nix, B., Jacobs, K.A., Bowles, M.L., 2012. Two decades of lowseverity prescribed fire increases soil nutrient availability in a Midwestern USA oak (Quercus) forest. Geoderma 183, 80–91. Schlesinger, W.H., 1977. Carbon balance in terrestrial detritus. Ann. Rev. Ecol. Evol. Syst. 8, 51–81. Schlesinger, W.H., 1984. Soil organic matter: a source of atmospheric CO2 . In: The Role of Terrestrial Vegetation in the Global Carbon Cycles: Measurement by Remote Sensing. John Wiley & Sons, New York, pp. 111–127. Schlesinger, W.H., Andrews, J.A., 2000. Soil respiration and the global carbon cycle. Biogeochemistry 48 (1), 7–20.

25

Smith, K.A., et al., 2003. Exchange of greenhouse gases between soil and atmosphere: interactions of soil physical factors and biological processes. Eur. J. Soil Sci. 54 (4), 779–791. Smith, N.R., Kishchuk, B.E., Mohn, W.W., 2008. Effects of wildfire and harvest disturbances on forest soil bacterial communities. Appl. Environ. Microbiol. 74 (1), 216–224. Smith, D.R., et al., 2010. Soil surface CO2 flux increases with successional time in a fire scar chronosequence of Canadian boreal jack pine forest. Biogeosciences 7 (5), 1375–1381. Subke, J.-A., Reichstein, M., Tenhunen, J.D., 2003. Explaining temporal variation in soil CO2 efflux in a mature spruce forest in Southern Germany. Soil Biol. Biochem. 35, 1467–1483. Switzer, J.M., Hope, G.D., Grayston, S.J., Prescott, C.E., 2012. Changes in soil chemical and biological properties after thinning and prescribed fire for ecosystem restoration in a Rocky Mountain Douglas-fir forest. For. Ecol. Manage. 275, 1–13. Tolhurst, K.G., 2003. Effects of repeated low-intensity fire on the understorey of a mixed eucalypt foothill forest in south-eastern Australia. Research report no. 58. Deptartment of Sustainability and Environment, East Melbourne, Victoria, Australia. Tolhurst, K.G., Kelly, N., 2003. Effects of repeated low-intensity fire on fuel dynamics in a mixed eucalypt foothill forest in south-eastern Australia. Research report no. 59. Deptartment of Sustainability and Environment, East Melbourne, Victoria, Australia. Tolhurst, K.G., Flinn, D.W., Loyn, R.H., Wilson, A.A.G., Foletta, I.J., 1992. The Ecological Effects of Fuel Reduction Burning in a Dry Sclerophyll Forest: A Summary of Principal Research Findings and Their Management Implications. Department of Conservation and Environment, Kew, Victoria, Australia. Tufekcioglu, A., Kucuk, M., Bilmis, T., Altun, L., Yilmaz, M., 2010. Soil respiration and root biomass responses to burning in Calabrian pine (Pinus brutia) stands in Edirne, Turkey. J. Environ. Biol. 31 (1/2), 15–19. von Fischer, J.C., Butters, G., Duchateau, P.C., Thelwell, R.J., Siller, R., 2009. In situ measures of methanotroph activity in upland soils: a reaction-diffusion model and field observation of water stress. J. Geophys. Res. Biogeosci. 114 (G1015). Williams, R.J., Hallgren, S.W., Wilson, G.W.T., 2012. Frequency of prescribed burning in an upland oak forest determines soil and litter properties and alters the soil microbial community. For. Ecol. Manage. 265, 241–247. Wilson, C.A., Mitchell, R.J., Boring, L.R., Hendricks, J.J., 2002. Soil nitrogen dynamics in a fire-maintained forest ecosystem: results over a 3-year burn interval. Soil Biol. Biochem. 34, 679–689. Wüthrich, C., Schaub, D., Weber, M., Marxer, P., Conedera, M., 2002. Soil respiration and soil microbial biomass after fire in a sweet chestnut forest in southern Switzerland. Catena 48, 201–215. Xu, W.H., Wan, S.Q., 2008. Water- and plant-mediated responses of soil respiration to topography, fire, and nitrogen fertilization in a semiarid grassland in northern China. Soil Biol. Biochem. 40 (3), 679–687.