Impact of temperature and moisture on heterotrophic soil respiration along a moist tropical forest gradient in Australia.
Tropical forests maintain the largest terrestrial biodiversity and are vulnerable to changes in climatic conditions (Foster 2001; Williams et al. 2003). Furthermore, they store the largest amount of carbon (C) of any biome, with >40% of world's terrestrial C stored in tropical ecosystems (Malhi et al. 1999; Raich et al. 2006; Schwendenmann and Pendall 2008). Disturbances to these vulnerable ecosystems will therefore likely lead to considerable changes in C02 exchanges with the atmosphere (Meir et al. 2008; Willis and Bhagwat 2009). The impact of warming on C fluxes in tropical rainforests has been explored in a range of studies (Kiese and Butterbach-Bahl 2002; Sotta et al. 2004; Saatchi et al. 2007; Cox et al. 2013). To estimate the impact of changing temperature and rainfall patterns on the C cycle of tropical rainforests in situ, several studies along altitudinal gradients have also been conducted (Girardin et al. 2010; Malhi et al. 2010; Moser et al. 2011). Most studies report an increase in soil C stocks with increased elevation (Raich et al. 2006; Wilcke et al. 2008; Dieleman et al. 2013), with soil C stocks at high elevations demonstrated to be higher than aboveground biomass C stocks in lowland rainforests (Malhi et al. 1999; Zimmermann et al. 2010). However, the fate of soil organic C (SOC) under projected climate change scenarios is still not well understood. Zimmermann et al. (2009) and Girardin et al. (2010) measured soil respiration along a tropical forest altitudinal gradient in Peru and found that total soil respiration (Rs) did not change along a temperature gradient of 14[degrees]C, but that relative biomass C allocation changed from above to below the ground, leading to different contributions of autotrophic and heterotrophic sources to [R.sub.s].
For analysing the effect of changed climatic conditions on soil C cycling, translocation of soil cores along altitudinal gradients was shown to be a successful approach in different ecosystems. Conant et al. (2008) translocated soil mesocosms across three semi-arid ecosystems in the San Francisco Mountains, USA, and found that heterotrophic soil respiration ([R.sub.sh]) rates were highest at the wettest site, whereas temperature was negatively correlated with [R.sub.sh]. In this semi-arid system, Rsh was controlled mainly by the size of soil C pools and by moisture. Hart (2006) replaced soil cores reciprocally between two forest sites along an altitudinal gradient in Arizona, USA, where there was a difference in mean annual temperature of 2.7[degrees]C and mean annual rainfall of 228 mm between the sites, and measured an increase in [R.sub.sh] rates of 190% when soil cores were translocated downslope. The [R.sub.sh] rates correlated as well with temperature as they did with moisture. Zimmermann et al. (2009) translocated soil cores along a tropical forest gradient in Peru and conducted a temperature-sensitivity analysis, which revealed that soil C-stabilisation processes mainly drove changes in temperature sensitivities. However, the effect of soil moisture on temperature sensitivity remained unclear.
The impact of changing climatic conditions on soil organic matter (SOM) decomposition of tropical soils has also been investigated in various laboratory incubation experiments (Bekku et al. 2003; Schwendenmann and Pendall 2008), but these incubation studies did not consider diurnal or seasonal climate patterns. Although the relationship between [R.sub.sh] and temperature has been explored in many studies across various ecosystems (Conant et al. 2011), the combined effect of temperature and soil moisture on SOM is not well explored. Soil moisture is known to limit soil respiration through limited access of microbial decomposers at low moisture levels and lack of oxygen at saturation levels of water (Chambers et al. 2004; Suseela et al. 2012). However, the predictive capacity of current model approaches for the respiration-moisture relation remains quite limited (Moyano et al. 2013).
In order to analyse the temperature and moisture effect on heterotrophic soil C effluxes, we translocated soil cores from three different moist tropical forest sites with different climatic conditions. The [R.sub.sh] rates were then measured regularly for 1 year, and model functions considering temperature and soil moisture were fitted to the measured [R.sub.sh] rates. The hypotheses we tested were: (i) soil moisture is the more dominant driver of [R.sub.sh] than temperature in moist tropical forests; (ii) translocation of soil cores from moist sites to sites with much lower rainfall will significantly alter the average [R.sub.sh] rates; and (iii) translocation of soil cores among different sites will not affect the relation between [R.sub.sh] and soil temperature.
Materials and methods
Soil translocation was initiated in February 2010 in the Wet Tropics Bioregion of Far North Tropical Queensland, Australia. Soil cores from three tropical forest sites at different elevations with different climatic conditions were excavated and translocated to the other sites, with some cores re-emplaced at their original location to act as controls. The low- and high-elevation sites were at the base and top of the Mt Bellendcn Ker Range and the mid-elevation site was at Robson Creek, further to the east.
The highest site was on top of Mt Bellenden Ker (site BT, 17[degrees]15'55"S, 145[degrees]51'14"E) with elevation 1540m a.s.l., mean annual temperature 14.2[degrees]C and mean annual rainfall 8100 mm. The forest type was a simple microphyll vine-fern thicket, on a Leptosol (according to WRB 2006) formed on granite. The mid-elevation site was Robson Creek (site RC, 17[degrees]07'00"S, 145[degrees]37'50"E) at 700 m a.s.l. with mean annual temperature 20.4[degrees]C and mean annual rainfall 1770 mm. The soil type was a Gleysol developed on metamorphic sediments in a simple notophyll vine forest. The lowest site was at the base of Mt Bellenden Ker (site BB, 17[degrees]16'11"S, 145[degrees]54'01"E) at 100m a.s.l. Mean annual temperature is 23.4[degrees]C and mean annual rainfall 4630 mm. The forest growing on this Ferralsol overlying the same granite bedrock as at BT was a complex mesophyll vine forest. A further two sites in open wooded savannah were chosen for installation of excavated soil cores from the three sites described above to expose the soil cores to a larger range of climatic conditions. These sites were Davies Creek at 670m a.s.l. (site DC, 17[degrees]01'17"S, 145[degrees]35'05"E) with mean annual temperature 22.2[degrees]C and mean annual rainfall 1260 mm, and Smithfield at 40 m a.s.l. (site JC, 16[degrees]48'59"S, 145[degrees]40'56"E) with mean annual temperature 25.1[degrees]C and mean annual rainfall 1990 mm. Climatic parameters were taken from the Bureau of Meteorology (www.bom.gov.au/climate), and further details regarding some of the sites can be found in Butterbach-Bahl et al. (2004), Graham (2006) and McJannet et al. (2007).
Soil cores were taken according to the methodology of Zimmermann et al. (2009, 2010). At each of the three sampling sites, leaf surface litter was removed and 15 plastic tubes of 10 cm diameter and 30 cm length were hammered into the ground. The core length was selected because 30 cm represented the maximum soil depth at the BT site. If the soil cores were compacted >5% (1.5 cm), the cores were discharged and newly taken. The tubes were then dug out and transported back to the laboratory, where a soil moisture sensor of 10 cm length (Vegetronix, Riverton, UT, USA) was pushed from the surface into the soil core, with the sensor cable led out of the core through a small hole, which was sealed with silicone after installation. To hinder root growth into the cores and to stop the loss of soil material from the base, a 63-pm nylon mesh was installed across the bottom of each core. Three cores from each sampling site were then installed in the ground at each of the five field sites by drilling holes with a 12-cm-wide auger. A layer of glass marbles was put on the soil core surfaces to protect the soils from splash erosion, a partial plastic cap was put on the top of the tubes to reduce the rainfall input, and a removable chicken wire to stop leaf litter input. The cap diameter was chosen to reduce the rainfall input by 20%, which is the average transpiratory loss of forests in the study region (McJannet et al. 2007). Because roots were excluded from the cores, there was no transpiratory loss from the cores. Although this 20% might be too small for soil moisture values at the highest elevation where cloud interception could contribute up to 66% of monthly water input (McJannet et al. 2007), and too low for drier sites, the reduction in rainfall input was necessary to avoid waterlogging within the cores and was based on best data available.
Dataloggers (XR5-SE; Pace Scientific, Mooresville, NC, USA) were installed at all five sites for recording rainfall (rainfall sensor RS-100; Pace Scientific) under the canopy, and soil temperature (temperature probe PT956; Pace Scientific) and soil moisture (soil moisture sensor Echo EC10; Degagon, Pullman, WA, USA) at 10 and 30 cm soil depths at half-hourly intervals. Additional temperature sensors (ibutton; Maxim, Sunnyvale, CA, USA) were installed in the shade at 10 cm above the soil surface.
To quantify the soil C stocks within the top 30 cm at the three sampling sites, three soil profiles at each site were sampled in 5-cm intervals with metal tubes of 5 cm length and diameter. All soil samples were then dried at 60[degrees]C, crushed and sieved to 2 mm to remove stones and roots >2 mm. The soil samples were then ground and the C concentrations measured with an elemental analyser (EA 4010; Costech, Valencia, CA, USA). Fine soil densities, corrected for stone and root volumes, were determined, and the C stocks calculated. All values are presented with standard errors. The pH was measured in composite soil samples from the layers 0-10, 10-20 and 20-30 cm, by using an electrode (PHM210; Radiometer Analytical, Villeurbanne, France) in a mixture of dry soil and 0.01 m Ca[Cl.sub.2] at a ratio of 1:10. Soil texture values for composite samples were determined using a laser diffraction analyser (Mastersizer 2000; Malvern, Worcestershire, UK).
Soil respiration measurements
Effluxes of [R.sub.sh] from the soil cores were measured every 3 weeks from April 2010 to May 2011 in the field with an LI-8100 (LICOR, Lincoln, NE, USA) portable infrared gas analyser. With this equipment, the C[O.sub.2] increases in a closed airflow system between the soil cores and a 10-cm survey chamber. This increase was measured over 150s and the C[O.sub.2] flux rate calculated based on exponential best fit equations (Kutzbach et al. 2007). On each measurement occasion, soil temperature at 10 cm depth ([T.sub.s]) next to the cores and volumetric soil water content (VWC) from the surface to a depth of 10 cm within the cores were also recorded.
To check for decreases in [R.sub.sh] rates over time due to lower C availability within the decomposing soil cores, [R.sub.sh] rates of control cores at the start and at the end of the translocation experiment were compared. For this, periods with very similar temperature and moisture conditions were selected and [R.sub.sh] rates of the corresponding periods evaluated.
To evaluate the impact of temperature and moisture on the measured [R.sub.sh] rates, we used the following equations, to calculate the temperature effect alone:
[R.sub.sh] = [R.sub.10] x [Q.sup.(Ts-10)/10.sub.10] (1)
to calculate the soil moisture effect alone:
[R.sub.sh] = a + b x VWC - c x [VWC.sup.2] (2)
and to calculate the combined effect of temperature and soil moisture:
[R.sub.sh] = ([R.sub.10] x [Q.sup.(Ts-10)/10.sub.10]) x (a + b x VWC - c x [VWC.sup.2] - c x [VWC.sup.2]) (3)
where [R.sub.sh] is measured heterotrophic respiration rate ([micro]mol C[O.sub.2] [m.sup.-2] [s.sup.-1]); [T.sub.s] is soil temperature ([degrees]C) at 10 cm depth; VWC is volumetric water content (%) at 10 cm depth; and [R.sub.10], [Q.sub.10], a, b and c are fitted parameters. Furthermore, [Q.sub.10] gives the temperature sensitivity of the respiration as a factor by which the respiration accelerates with an increase in temperature of 10[degrees]C, and [R.sub.10] the fitted [R.sub.sh] rate at 10[degrees]C. This temperature sensitivity function was successfully used for annual soil respiration data of different ecosystems (Janssens and Pilegaard 2003; Reichstein et al. 2005; Zimmermann and Bird 2012), and the soil moisture term describes the site-specific limitation of [R.sub.sh] at very high (saturation) or low (drought) water contents (Chambers et al. 2004; Zimmermann et al. 2009). Davidson et al. (2000) assumed that a quadratic function relating soil moisture to microbial soil respiration might be the best option, and Moyano et al. (2013) confirmed the usefulness of this approach in a recent review.
Data were checked for normal distribution with a Shapiro-Wilk test, and differences among measurements tested for significance with an analysis of variance (ANOVA) for normally distributed data, and a Kruskal-Wallis one-way ANOVA for non-normally distributed data. Measured respiration rates were fitted to the model Eqns 1-3 by means of least-square regression using SIGMAPLOT 12.0 (SYSTAT Software Inc., San Jose, CA, USA). For this, data were first tested for variance homogeneity. Regression fits were then calculated individually for soil cores from the same origin installed at each site, and then for all pooled respiration measurements of soil cores from the same origin.
The model performances were then compared by using Akaike's information criterion (AIC); the model fit with the lowest AIC value was considered to be the best one (Burnham and Anderson 2004). AIC values were calculated only for model fits with significant correlations between measured and modelled data (P < 0.05). These model predictions were tested for normal distribution of errors, in which case the criterion was calculated as:
AIC = n x log([[sigma].sup.2]) + 2 x K (4)
where [[sigma].sup.2] = (residual sum of squares)/n; n is the sample size; and K the number of estimated parameters where the variance was also counted as an estimated parameter (Burnham and Anderson 2004).
Climatic parameters for the period March 2010-April 2011 coincided well with long-term averages provided by the Bureau of Meteorology (Table 1). Annual rainfall was highest at site BT, with 8019 mm (under canopy), followed by BB, JC, RC and DC with 1260 mm (Fig. 1). Measured rainfall was highest during February 2011, when category 5 Tropical Cyclone Yasi crossed the Queensland coast 75 km south of Bellenden Kcr. Monthly rainfall data were also available from the Bureau of Meteorology for BT and BB, and revealed a very high correlation between values measured at open vegetation sites and under the canopy (correlation coefficients [R.sup.2] > 0.95) with 60.6% and 36.2% of the total rainfall reaching the soil surface at BT and BB, respectively. However, VWC did not follow the same trend as rainfall, mostly because of the different soil textures and densities at each site. At BT, the topsoil was very moist throughout the year, whereas the soil layers at 20-30 cm depth drained much more quickly due to coarse fragments in the B/C layer of the Leptosol. VWCs at RC and BB were more similar for 10 and 30 cm depths. Average annual soil temperature at 10 cm depth was highest at site JC at 40 m a.s.l. (25.3[degrees]C) and decreased by ~0.56[degrees]C per 100 m in altitude to 15.9[degrees]C at BT at 1540 m a.s.l.
Soil C stocks in the top 30 cm were highest at BT (10.66 [+ or -] 0.15 kg C [m.sup.-2]) and decreased linearly with decreasing altitude to 6.13 [+ or -]0.06 kg C [m.sup.-2] at BB (C stock=0.0031 x altitude + 6.002, [R.sup.2] = 0.94, P < 0.01) (Table 1), which equates to a difference in soil C stock of 471 g C per [degrees]C. Average pH was lowest at BT (3.83 [+ or -] 0.13) and slightly higher at RC (4.33 [+ or -] 0.09) and BB (4.20 [+ or -] 0.02).
Soil respiration rates
The [R.sub.sh] rates for all translocated soil cores measured during the experiment are shown in Fig. 2. At each site, the three cores from the same origin were grouped together per measurement event, and the average values are presented with standard errors. [R.sub.sh] rates were not measured for the first 2 months after installation of the cores, because severed dead roots in the cores might have enhanced [R.sub.sh] rates (Kuzyakov 2006). After this initial stabilisation phase, [R.sub.sh] rates did not correlate with time since installation (Pearson product moment correlations not significant between [R.sub.sh] rates and days since installation of the cores; P > 0.05 in all cases), indicating no bias in the results from dead roots or depletion of SOM in the cores. [R.sub.sh] rates did not reveal any declining trend over the experiment, although C availability decreased over time due to the experimental setup hindering any new C input from reaching the cores. [R.sub.sh] rates of control cores from BT were not significantly different in April 2010 and January 2011 ([+ or -] = 0.73, 2-tailed f-test), when Ts values were similar (15.6 [degrees] C in April 2010 and 15.8 [degrees] C in January 2011). The same was true for control cores from RC when comparing [R.sub.sh] measurements in April 2010 and January 2011 ([+ or -] = 0.41), with Ts of 20.6[degrees]C and 20.8[degrees]C, respectively, and [R.sub.sh] rates of control cores from BB ([+ or -] = 0.76), with Ts of 23.5[degrees]C in April 2010 and 23.4[degrees]C in January 2011.
Cores installed at DC were only partly measured in 2011, because several cores became inhabited by termites, which led to respiration rates being an order of magnitude higher than for non-inhabited cores. [R.sub.sh] rates of all soil cores installed at BB were not considered after February 2011, this site being disturbed by a tropical cyclone. The translocation of the soils to new sites had a significant impact on average [R.sub.sh] rates measured during the study period. Figure 3 presents boxplots of the measured [R.sub.sh] rates. Translocation of the soil cores among sites allowed measurement of [R.sub.sh] rates across a temperature range of ~15[degrees]C, whereas the in situ temperature ranges during the [R.sub.sh] measurements over the year were 7.2[degrees]C for BT, 5.8[degrees]C for RC, and 8.5[degrees]C for BB. Soil cores from BT were only warmed up by the translocation, whereas soil cores from RC and BB were warmed and cooled compared with their native in situ [T.sub.s]. Furthermore, soil cores from BT received less rain by up to 6840 mm when installed at DC, whereas soil cores from RC received more rain at three of the five installation sites. Cores from BB received 3407 mm more rain at BT and up to 3370 mm less rain at the other sites.
Rates of [R.sub.sh] of soil cores from the coolest and wettest site, BT, were increased by ~150% when installed at RC and BB compared with their site of origin, and by ~300% when installed at sites DC and JC. [R.sub.sh] rates of soil cores from RC revealed a different pattern with decreased rates when installed at BT (-56%), at BB (-42%) and at DC (-4%), but increased rates when installed at JC (+18%). The lower [R.sub.sh] rates at BT and BB were significantly different from the [R.sub.sh] rates of the control cores installed at RC, whereas cores installed at DC and JC were not significantly different from the control cores. Soil cores taken at BB delivered significantly lower [R.sub.sh] rates when installed at BT (-50%), but slightly higher rates at RC (+18%), DC (+17%) and JC (+45%).
Modelled temperature and moisture dependence of [R.sub.sh]
Data analysis revealed variance homogeneity of [R.sub.sh] rates across all temperature ranges considered and allowed the calculation of regression models according to Eqns 1-3. The calculated parameters of these different model functions are given in Tables 2 and 3.
Soil cores from BT revealed a significant correlation of Rsh rate with [T.sub.s] at DC only. The [R.sub.sh] rates of cores from RC showed significant correlations with [T.sub.s] at RC, BB and DC, whereas the explained variance by temperature was rather low in RC and DC ([R.sup.2] of 0.47 and 0.46, respectively). However, comparing model fits of measured [R.sub.sh] rates of cores from RC revealed best results for soil cores reinstalled at their origin site RC, because AIC was lowest for this site (Table 2). As at BT, the soil cores from BB correlated significantly with Ts at DC only.
Use of Eqn 2 to relate [R.sub.sh] rates to VWC at the different host sites did not result in improved prediction models compared with Eqn 1. Soil cores from the wettest site, BT, showed a significant relation between [R.sub.sh] and VWC when installed at RC ([R.sup.2] of 0.70). However, [R.sub.sh] rates from all other soil cores installed at the different sites did not show significant relationships with VWC (Table 3).
Use of the combined Ts and VWC equation gave highly significant model fits for [R.sub.sh] rates of soil cores from BT installed at RC and BB, and for soil cores from RC in DC, but not for [R.sub.sh] rates from any other soil cores (Table 4).
Comparison of model fits considering AIC values was possible in only two cases. [R.sub.sh] rates of soil cores from BT installed at RC were better projected with the moisture alone (AIC = -9.65) than with the combined Eqn 3 (AIC = -5.67), and rates from cores originating from RC installed in DC were more significantly predicted with the combined Eqn 3 (AIC 3.84) than with the temperature alone Eqn 1 (AIC = 7.41).
Consideration of [R.sub.sh] rates from all host sites in one equation led to significant model fits in most cases (Table 5). [R.sub.sh] rates could be correlated significantly to Ts for all three soils and delivered [Q.sub.10] values of 2.63 for the soil cores from BT, 2.06 for soils from RC, and 2.00 for the soil cores from BB. The only significant fit between [R.sub.sh] rates and VWC was found for the soil cores originating from BT, where the explained variation was rather poor ([R.sup.2] = 0.16) and the shape of the regression fit did not show an optimal maximal moisture content for [R.sub.sh] but rather a decline in [R.sub.sh] rates at medium soil-moisture contents (Fig. 4). The combined [T.sub.s] and VWC approach of Eqn 3 resulted in highly significant regression fits for all three soils, with correlation coefficients 0.60-0.39, comparable to the temperature-alone fit of Eqn 1. Furthermore, temperature sensitivity values, Qw, were also very close to those obtained by Eqn 1. However, AIC values for Eqn 1 were always lower for the temperature-alone Eqn 1 than for the combined temperature and VWC Eqn 3, even though standard errors of estimates were almost identical (Table 5).
As a further analysis, we related average [R.sub.sh] rates as measured over the entire study year to the site-specific factors mean annual temperature, mean annual VWC and total rainfall (Table 6). The regression equations used were the same as for soil temperature and soil moisture, and a linear approach was chosen for total rainfall. Correlation coefficients between mean annual [R.sub.sh] rates and mean annual temperature were 0.45-0.98, between mean annual [R.sub.sh] rates and mean annual VWC 0.24-0.72, between mean annual [R.sub.sh] rates and rainfall 0.62-0.75, and between mean annual [R.sub.sh] rates and a combination of mean annual temperature and rainfall 0.83-1. Reduction of data to five points weakened the confidence in the modelled parameters, resulting in P values slightly greater than 0.05 in most cases (Table 6). By this approach, mean annual [R.sub.sh] rates for soil cores from RC could best be explained by total rainfall ([R.sup.2] = 0.81, P = 0.04), and mean annual [R.sub.sh] rates for soil cores from BB by the combination of mean annual temperature and rainfall ([R.sup.2] = 1.0, P < 0.01). However, the fit of measured data to mean annual temperature and rainfall in BB resulted in only one function parameter being significant. The only case where the AIC was better than in any approach before was the projection of [R.sub.sh] rates from RC by considering mean annual rainfall alone (AIC = 3.07), which also reduced the standard error of estimation from 1.12 to 0.66.
Average soil respiration rates
Total soil respiration consists of autotrophic respiration ([R.sub.sa]) from plant sources and [R.sub.sh] from microbes decomposing accumulated SOM (Hanson et al. 2000; Bond-Lamberty and Thomson 2010). Most studies of soil C[O.sub.2] effluxes are based on Rs, and studies measuring [R.sub.sh] are mainly conducted as laboratory incubation experiments (von Liitzow and Kogel-Knabner 2009). The field method we used allows direct measurement of [R.sub.sh] in the field, because no living roots remained in the cores, and no enhanced C[O.sub.2] effluxes (potentially resulting from severed dead roots priming initial [R.sub.sh] rates) were observed after 2 months. This has the advantage that the soils experience 'real' environmental conditions, including episodic rainfall and diurnal temperature variation. The [R.sub.sh] rates of the soil cores taken and installed at the same sites represent the [R.sub.sh] components of the corresponding site, and they ranged from 0.91 to 4.22 [micro]mol C[O.sub.2] [m.sup.-2][s.sup.-1]. These values are in the upper range of reported [R.sub.sh] rates for tropical forests. Li et al. (2004) reported soil respiration rates that excluded root and litter in a tropical forest in Puerto Rico, with climatic conditions similar to our study gradient, of 0.46 [micro]mol C[O.sub.2] [m.sup.-2][s.sup.-1], and Zimmermann et al. (2010) reported measured [R.sub.sh] rates of 1.1-2.7 [micro]mol C[O.sub.2] [m.sup.-2][s.sup.-1] along a 3000-m altitudinal tropical forest gradient in Peru.
Climate dependence of heterotrophic respiration rates
According to theory (Lloyd and Taylor 1994; Davidson and Janssens 2006), [R.sub.sh] rates should increase at higher temperatures, as long as moisture or oxygen are not limiting. Soil cores originating from BT increased at all sites with higher temperatures and drier conditions, indicating no significant moisture limitation at any site. This is especially surprising, because total annual rainfall dropped to as low as 15% (in DC) relative to its native site. [R.sub.sh] rates of soil cores from RC, however, did not increase significantly at the warmer host sites, and even decreased at site BB, which was 3[degrees]C warmer but experienced 2.6 times the rainfall of RC. Soil cores from BB showed increased [R.sub.sh] rates at the warmest site, JC, but the lower rain input at RC and DC seemed not to have any major impact on [R.sub.sh] rates, if considered separately. These patterns were also confirmed by the regression models considering rain and mean annual temperature, because RC had the lowest AIC value for the model with total annual rainfall, and BT and BB for the model based on [T.sub.s] alone.
Although soil moisture is known to affect respiration rates (Davidson et al. 2000; Sotta et al. 2004; Meir et al. 2008), the soil moisture term in the combined functions analysed here did not show any significant improvement in fitting measured [R.sub.sh] rates, as indicted by the AIC values. The slightly better correlation coefficients between measured and modelled data accounting for soil temperature and moisture were not large enough to justify the two additionally estimated parameters for the moisture term in Eqn 3. Soil moisture alone was in no case a significant predictor for the measured [R.sub.sh] rates. A reason for this might be that the range of VWC along the translocation gradient was still not large enough to observe any effect of drought or saturation, although the soil cores were placed along sites with a difference in mean annual rainfall of ~6650 mm [year.sup.-1]. Davidson et al. (2000) found for an Amazonian forest soil (~1800 mm rain [year.sup.-1]) that [R.sub.s] rates were suppressed by high soil moisture only very close to saturation (matrix potential-0.005 MPa), and by low water contents at a matrix potential of ~-10MPa, which corresponds to a VWC of <5%. However, Sotta et al. (2004) found a cubic relation between VWC and [R.sub.s] ([R.sup.2] = 0.40) at values of 42- 45% VWC for a tropical forest soil with a mean annual rainfall of 2200 mm, and Hashimoto et al. (2004) reported an increase in [R.sub.s] at 10-40% VWC in 10 cm soil depth of a tropical monsoon forest in Thailand. Although our translocated soil cores covered a range of 19-37% VWC, soil moisture was not a significant predictor for [R.sub.sh]. There are two possible conclusions for this observation: (i) VWC was not a driving factor of [R.sub.sh] in our moist study forests, or (ii) the suggested cubic VWC function was not an adequate model to relate [R.sub.sh] rates to VWC measurements.
However, the decomposition of SOM depends not only on temperature and moisture, but also on substrate quality and microbial access to SOM (Davidson and Janssens 2006; Conant et al. 2011). Therefore, comparison of temperature sensitivity [Q.sub.10] values among soils with different C inputs and different textural protection mechanisms is somewhat arbitrary. Furthermore, calculated [Q.sub.10] values are highly dependent not only on the applied temperature function (Fang and Moncrieff 2001; Tuomi et al. 2008) but also on the considered temperature range, as our results showed. Consideration of the full translocation range of temperatures resulted in a [Q.sub.10] of 2.07 for the RC soil cores, but [Q.sub.10] was 2.75 when calculated for the native Ts only. Rates of [R.sub.sh] for soil cores from BT and BB did not result in significant model fits considering the on-site measurements only. Anyway, the [Q.sub.10] values calculated in this study for the SOM substrate respired in the first year were well within the range of other reported [Q.sub.10] values of 1.0-5.6 for tropical soils (Bekku et al. 2003; Bahn et al. 2010).
Impact of future climatic conditions on respiration rates
Suppiah et al. (2007) reported climate predictions for the study region based on 23 global climate models and emission scenarios A2 and A1B of the 1PCC Report Emission Scenarios (1PCC SRES 2000). Their analysis revealed that temperature will increase by ~2.6[degrees]C by 2080 (compared with 2010) in the Wet Tropics Bioregion of Far North Tropical Queensland. Annual rainfall was predicted to decrease by ~23% in the dry season and 1% in the wet season by 2080. The greatest reduction in annual rainfall as achieved by translocation was by bringing soil cores from BT to DC, which reduced the rainfall amount by 84%, and the smallest reduction was by bringing soil cores from RC to DC, which reduced the incoming water amount by 29%. Therefore, the translocation covered the entire range in rainfall and temperature as predicted by Suppiah et al. (2007) for the year 2080.
Assuming that Ts will increase at the same rate as air temperature, [R.sub.sh] would increase over the next 70 years by 29% in BT compared with the respiration rates in 2010-11, 21% in RC, and 19% in BB. The greater increases in [R.sub.sh] rates for the sites with the larger soil C stocks could lead to a total net C loss at the three study sites, as was also observed for soils along altitudinal gradients in South America (Girardin et al. 2010; Zimmermann et al. 2010). However, it must considered that these predictions ignore future potential depletion of C stocks with ongoing decomposition (Conant et al. 2008), changes in C input rates and substrate qualities as caused by potential shifts in tree species (Feeley and Silman 2010), or transformations in the microbial communities of the SOM decomposers (Bradford et al. 2008). Estimated future [R.sub.sh] rates as calculated here were based solely on temperature changes.
Advantage of translocation soil cores over laboratorial incubation studies
The translocation of relative large monoliths along the altitudinal gradient was successful for determining the impact of changes in temperatures and rainfall. Incubations of soil samples in the laboratory under controlled climatic conditions are normally conducted on much smaller soil cores (Bekku et al. 2003; Chen et al. 2010), on which the impact of cutting roots, disturbed soil structure and missing C input is larger, and [R.sub.sh] rates decline over the experiment duration. Zimmermann and Bird (2012) conducted an incubation study at the same sites with smaller tubes (5 cm diameter, 20 cm length) and reported repeated incubation rates that declined by 50% after 6 months. This effect is well known and is a major problem in incubation studies conducted in the laboratory taking several weeks to months (Schaufler et al. 2010). The larger soil cores translocated here did not show declining [R.sub.sh] rates due to lower C availability, indicating that C substrate supply to microbes was not limiting over the experiment period. Therefore, translocated soil monoliths with a large volume are well suited to study changes in [R.sub.sh] rates under natural conditions, even if the studied temperature range cannot be adjusted as well as in the laboratory, and temperature and moisture effects must be considered simultaneously.
Tropical rain forests are substantial global C stores and sensitive to changes in climatic conditions. Our results showed that warming would have the largest impact on tropical forest soils with high C stocks, because the temperature sensitivity of [R.sub.sh] rates increased with increasing elevations. From a global perspective, the amount of C stored in soils is more moisture-dependent than it is temperature-driven (Scharlemann et al. 2014). However, this is based on steady-state conditions. Rapid changes in climatic conditions as simulated here can be better estimated with temperature functions alone, because moisture was hardly a limiting factor in the moist tropical forest sites studied. An exception to this, however, might be extreme events with long drought periods that could change [R.sub.sh] rates substantially.
Received 25 July 2014, accepted 12 December 2014, published online 2 April 2015
This study was financed by an Australian Research Council Federation Fellowship to MIB. We thank the Department of Environment and Resource Management, Queensland Government Australia, for access to the national park sites, and Broadcast Australia for access and transport to Mt Bellenden Ker. A special thanks to Spiro Buhagiar and Ian McConnell for their support on transportation.
Bahn M, Reichstein M, Davidson EA, Griinzweig J, Jung M, Carbone MS, Epron D, Mission L, Novellon Y, Roupsard O, Savage K, Trumbore SE, Gimeno C, Curiel Yuste J, Tang J, Vargas R, Janssens IA (2010) Soil respiration at mean annual temperature predicts annual total across vegetation types and biomes. Biogeosciences 7, 2147-2157. doi: 10.5194/bg-7-2147-2010
Bekku YS, Nakatsubo T, Kume A, Adachi M, Koizumi H (2003) Effect of warming on the temperature dependence of soil, respiration rate in arctic, temperate and tropical soils. Applied Soil Ecology 22, 205-210. doi: 10.1016/S0929-1393(02)00158-0
Bond-Lamberty B, Thomson A (2010) A global database of soil respiration data. Biogeosciences 7, 1915-1926. doi: 10.5194/bg-7-1915-2010
Bradford MA, Davies CA, Frey SD, Maddox TR, Melillo JM, Mohan JE, Reynolds JF, Treseder KK, Wallenstein MD (2008) Thermal adaptation of soil microbial respiration to elevated temperature. Ecology Letters 11, 1316-1327. doi: 10.1111/j.1461-0248.2008.01251.x
Burnham KP, Anderson DR (2004) Multimodel inference: Understanding AIC and BIC in model selection. Sociological Methods & Research 33, 261-304. doi: 10.1177/0049124104268644
Butterbach-Bahl K, Kock M, Willibald G, Hewett B, Buhagiar S, Papen H, Kiese R (2004) Temporal variations of fluxes of NO, N[O.sub.2], [N.sub.2O], C[O.sub.2], and C[H.sub.4] in a tropical rain forest ecosystem. Global Biogeochemical Cycles 18, GB3012. doi: 10.1029/2004GB002243
Chambers JQ, Tribuzy ES, Toledo LC, Crispim BF, Hihuchi N, dos Santos J, Araujo AC, Kruijt B, Nobre AD, Trumbore SE (2004) Respiration from a tropical forest ecosystem: partitioning of sources and low carbon use efficiency. Ecological Applications 14, 72-88. doi: 10.1890/01-6012
Chen X, Tang J, Jiang L, Li B, Chen J, Fang C (2010) Evaluating the impacts of incubation procedures on estimated Q10 values of soil respiration. Soil Biology & Biochemistry 42, 2282-2288. doi: 10.1016/j.soilbio.2010.08.030
Conant RT, Drijber RA, Haddix ML, Parton WJ, Paul EA, Plante AF, Six J, Steinweg JM (2008) Sensitivity of organic matter decomposition to warming varies with its quality. Global Change Biology 14, 868-877. doi: 10.1111/j. 1365-2486.2008.01541.x
Conant RT, Ryan MG, Agren GI, Birge HE, Davidson EA, Eliasson PE, Evans SE, Frey SD, Giardina CP, Hopkins FM, Hyvonen R, Kirschbaum MUF, Lavallee JM, Leifeld J, Parton WJ, Steinweg JM, Wallenstein MD, Martin Wetterstedt JA, Bradford MA (2011) Temperature and soil organic matter decomposition rates--synthesis of current knowledge and a way forward. Global Change Biology 17, 3392-3404. doi: 10.1111/j.1365-2486.2011.02496.x
Cox PM, Pearson D, Booth BB, Friedlingstein P, Huntingford C, Jones CD, Luke CM (2013) Sensitivity of tropical carbon to climate change constrained by carbon dioxide variability. Nature 494, 341-344. doi:10.1038/naturel 1882
Davidson EA, Janssens 1A (2006) Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature 440, 165-173. doi: 10.1038/nature04514
Davidson EA, Verchot LV, Cattanio JH, Ackerman IL, Carvalho JEM (2000) Effects of soil water content on soil respiration in forests and cattle pastures of eastern Amazonia. Biogeochemistry 48, 53-69. doi: 10.1023/A:1006204113917
Dieleman WIJ, Venter M, Ramachandra A, Krockenberger AK, Bird MI (2013) Soil carbon stocks vary predictably with altitude in tropical forests: Implications for soil carbon storage. Geoderma 204-205, 59-67. doi: 10.1016/j.geoderma.2013.04.005
Fang C, Moncrieff JB (2001) The dependence of soil C02 efflux on temperature. Soil Biology & Biochemistry 33, 155-165.
Feeley KJ, Silman MR (2010) Modelling the responses of Andean and Amazonian plant species to climate change: the effects of georeferencing errors and the importance of data filtering. Journal of Biogeography 37, 733-740. doi: 10.1111/j.1365-2699.2009.02240.x
Foster P (2001) The potential negative impacts of global climate change on tropical montane cloud forests. Earth-Science Reviews 55, 73-106. doi: 10.1016/S0012-8252(01)00056-3
Girardin CAJ, Malhi Y, Aragao LEOC, Mamani M, Huaraca HW, Durand L, Feeley KJ, Rapp J, Silva-Espejo JE, Silman M, Salinas N, Whittaker RJ (2010) Net primary productivity allocation and cycling of carbon along a tropical forest elevational transect in the Peruvian Andes. Global Change Biology 16, 3176-3192.
Graham AW (2006) 'The CSIRO rainforest permanent plots of North Queensland--site, structural, floristic and edaphic descriptions.' (CSIRO and the Cooperative Research Centre for Tropical Rainforest Ecology and Management: Cairns, Qld)
Hanson PJ, Edwards NT, Garten CT, Andrews JA (2000) Separating root and soil microbial contributions to soil respiration: A review of methods and observations. Biogeochemistry 48, 115-146. doi:10.1023/A:1006244819642
Hart SC (2006) Potential impacts of climate change on nitrogen transformations and greenhouse gas fluxes in forests: a soil transfer study. Global Change Biology 12, 1032-1046. doi:10.1111/j.13652486.2006.01159.x
Hashimoto S, Tanaka N, Suzuki M, Inouc A, Takizawa H, Kosaka I, Tanaka K, Tantasirin C, Tangtham N (2004) Soil respiration and soil C[O.sub.2] concentration in a tropical forest, Thailand. Journal of Forest Research 9, 75-79.
IPCC SRES (2000) 'Special Report on Emissions Scenarios: A special report of Working Group III of the Intergovernmental Panel on Climate Change.' (Eds N Nakicenovic, R Swart) (Cambridge University Press: Cambridge, UK)
Janssens 1A, Pilegaard K (2003) Large seasonal changes in Q(10) of soil respiration in a beech forest. Global Change Biology 9, 911-918. doi: 10.1046/j.1365-2486.2003.00636.X
Kiese R, Butterbach-Bahl K (2002) N20 and C02 emissions from three different tropical forest sites in the wet tropics of Queensland, Australia. Soil Biology & Biochemistry 34, 975-987. doi: 10.1016/S0038-0717(02) 00031-7
Kutzbach L, Schneider J, Sachs T, Giebels M, Nykanen H, Shurpali NJ, Martikainen PJ, Aim J, Wilmking M (2007) C[O.sub.2] flux determination by closed-chamber methods can be seriously biased by inappropriate application of linear regression. Biogeosciences 4, 1005-1025. doi: 10.5194/bg-4-1005-2007
Kuzyakov Y (2006) Sources of C02 efflux from soil and review of partitioning methods. Soil Biology & Biochemistry 38, 425-448. doi: 10.1016/j.soilbio.2005.08.020
Li Y, Xu M, Sun OJ, Cui W (2004) Effects of root and litter exclusion on soil CO2 efflux and microbial biomass in wet tropical forests. Soil Biology and Biochemistry 36, 2111-2114.
Lloyd J, Taylor JA (1994) On the temperature-dependence of soil respiration. Functional Ecology 8, 315-323.
Malhi Y, Baldocchi DD, Jarvis PG (1999) The carbon balance of tropical, temperate and boreal forests. Plant, Cell & Environment 22, 715-740. doi: 10.1046/j.1365-3040.1999.00453.x
Malhi Y, Silman M, Salinas N, Bush M, Meir P, Saatchi S (2010) Introduction: Elevation gradients in the tropics: laboratories for ecosystem ecology and global change research. Global Change Biology 16, 3171-3175. doi:10.1111/j.1365-2486.2010.02323.x
McJannet D, Fitch P, Disher M, Wallace J (2007) Measurements of transpiration in four tropical rainforest types of north Queensland, Australia. Hydrological Processes 21, 3549-3564. doi: 10.1002/hyp.6576
Meir P, Metcalfe DB, Costa ACL, Fisher RA (2008) The fate of assimilated carbon during drought: impacts on respiration in Amazon rainforests. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 363, 1849-1855. doi:10.1098/rstb.2007.0021
Moser G, LeuschnerC, Hertel D, Graefe S, Soethc N, lost S (2011) Elevation effects on the carbon budget of tropical mountain forests (S Ecuador): the role of the belowground compartment. Global Change Biology 17, 2211-2226. doi: 10.1111/j.1365-2486.2010.02367.x
Moyano FE, Manzoni S, Chenu C (2013) Responses of soil heterotrophic respiration to moisture availability: An exploration of processes and models. Soil Biology & Biochemistry 59, 72-85. doi: 10.1016/j.soilbio.2013.01.002
Raich JW, Russell AE, Kitayama K, Parton WJ, Vitousek PM (2006) Temperature influences carbon accumulation in moist tropical forests. Ecology 87, 76-87. doi: 10.1890/05-0023
Reichstein M, Subke JA, Angeli AC, Tenhunen JD (2005) Does the temperature sensitivity of decomposition of soil organic matter depend upon water content, soil horizon, or incubation time? Global Change Biology 11, 1754-1767. doi: 10.1111/j.1365-2486.2005.001010.x
Saatchi SS, Houghton RA, Alvala R, Soares JV, Yu Y (2007) Distribution of aboveground live biomass in the Amazon basin. Global Change Biology 13, 816-837. doi:10.1111/j.1365-2486.2007.01323.x
Scharlemann JPW, Tanner EVJ, Hiederer R, Kapos V (2014) Global soil carbon: understanding and managing the largest terrestrial carbon pool. Carbon Management 5, 81-91. doi:10.4155/cmt,13.77
Schaufler G, Kitzler B, Schindlbacher A, Skiba U, Sutton MA, Zechmeister-Boltenstern S (2010) Greenhouse gas emissions from European soils under different land use: effects of soil moisture and temperature. European Journal of Soil Science 61, 683-696. doi:10.1111/j.13652389.2010.01277.x
Schwendenmann L, Pendall E (2008) Response of soil organic matter dynamics to conversion from tropical forest to grassland as determined by long-term incubation. Biology and Fertility of Soils 44, 1053-1062. doi: 10.1007/s00374-008-0294-2
Sotta ED, Meir P, Malhi Y, Nobre AD, Hodnett M, Grace J (2004) Soil C[O.sub.2] efflux in a tropical forest in the central Amazon. Global Change Biology 10, 601-617. doi: 10.1111/j. 1529-8817.2003.00761 .x
Suppiah R, Macadam I, Whetton PH (2007) 'Climate change projections for the tropical rainforest region of North Queensland.' (Marine and Tropical Sciences Research Facility: Caims, Qld)
Suseela V, Conant RT, Wallenstein MD, Dukes JS (2012) Effects of soil moisture on the temperature sensitivity of heterotrophic respiration vary seasonally in an old-field climate change experiment. Global Change Biology 18, 336-348. doi:10.1111/j. 1365-2486.2011.02516.x
Tuomi M, Vanhala P, Karhu K, Fritze H, Liski J (2008) Heterotrophic soil respiration--comparison of different models describing its temperature dependence. Ecological Modelling 211, 182-190. doi: 10.1016/j.ecolmodel.2007.09.003
von Liitzow M, Kogel-Knabner I (2009) Temperature sensitivity of soil organic matter decomposition--what do we know? Biology and Fertility of Soils 46, 1-15. doi: 10.1007/s00374-009-0413-8
Wilcke W, Oelmann Y, Schmitt A, Valarezo C, Zech W, Homeier J (2008) Soil properties and tree growth along an altitudinal transect in Ecuadorian tropical montane forest. Journal of Plant Nutrition and Soil Science 171, 220-230. doi:10.1002/jpln.200625210
Williams SE, Bolitho EE, Fox S (2003) Climate change in Australian tropical rainforests: an impending environmental catastrophe. Proceedings of the Royal Society of London. Series B, Biological Sciences 270, 1887-1892. doi: 10.1098/rspb.2003.2464
Willis KJ, Bhagwat SA (2009) Biodiversity and climate change. Science 326, 806-807. doi:10.1126/science.1178838
WRB (2006) 'World reference base for soil resources.' World Soil Resources Reports No. 103. (FAO: Rome)
Zimmermann M, Bird MI (2012) Temperature sensitivity of tropical forest soil respiration increase along an altitudinal gradient with ongoing decomposition. Geoderma 187-188, 8-15. doi: 10.1016/j.geoderma. 2012.04.015
Zimmermann M, Meir P, Bird Ml, Malhi Y, Ccahuana AJQ (2009) Climate dependence of heterotrophic soil respiration from a soil-translocation experiment along a 3000 m tropical forest altitudinal gradient. European Journal of Soil Science 60, 895-906. doi: 10.1111/j.1365-2389.2009.01175.x
Zimmermann M, Meir P, Bird MI, Malhi Y, Ccahuana AJQ (2010) Temporal variation and climate dependence of soil respiration and its components along a 3000 m altitudinal tropical forest gradient. Global Biogeochemical Cycles 24, GB4012. doi: 10.1029/2010GB003787
M. Zimmermann (A,B,C), K. Davies (A), V. T. V. Pena de Zimmermann (A), and M. I. Bird (A)
(A) Centre for Tropical Environmental and Sustainability Science and School of Earth and Environmental Sciences, James Cook University, Cairns, Qld 4870, Australia.
(B) Institute of Soil Research, University of Natural Resources and Life Sciences, Peter Jordan St. 82, 1190 Vienna, Austria.
(C) Corresponding author. Email: Michael.email@example.com
Table 1. Site properties, soil carbon stocks (kg [m.sup.-2]), soil texture (%) and pH values for the three soil sampling sites Mean annual temperature (MAT) and rainfall were taken from long-term records from the Bureau of Meteorology, Australia Depth (cm) C stock Clay Silt Sand pH Bellenden Ker, top: 1540m a.s.l., MAT 14.2[degrees]C, total rain 8100mm 0-10 5.02 [+ or -] 0.37 5.4 11.3 83.3 3.66 10-20 3.67 [+ or -] 0.29 6.1 13.9 80.1 3.60 20-30 1.97 [+ or -] 0.23 38.3 46.4 15.3 3.99 Robson Creek: 700 m a.s., l, MAT 20.4[degrees]C. total rain 1770mm 0-10 3.97 [+ or -] 0.33 17.9 26.1 55.9 4.51 10-20 2.63 [+ or -] 0.32 24.8 39.7 35.5 4.27 20-30 1.89 [+ or -] 0.14 16.6 30.7 52.7 4.21 Bellenden Ker, bottom: 100m a.s.l., MAT 23.4[degrees]C, total rain 4630mm 0-10 2.81 [+ or -] 0.21 12.8 26.6 60.6 4.19 10-20 1.90 [+ or -] 0.07 19.1 32.7 48.1 4.18 20-30 1.42 [+ or -] 0.05 24.0 33.1 42.9 4.23 Table 2. Parameter fits, correlation coefficients and error probabilities for regressions between heterotrophic soil respiration and soil temperature according to the equation [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup.(Ts - 10)/10] for soil cores from the sites Bellenden Ker, top (BT), Robson Creek (RC) and Bellenden Ker, bottom (BB) installed at the sites BT, RC, BB, DC (Davies Creek) and JC (Smithfield) SEE, Standard error of estimation; AIC, Akaike information criterion for significant model fits. * P<0.05; ** P<0.01 Core Installation n [R.sub.10] [Q.sub.10] [R.sup.2] origin site BT BT 10 1.26 ** 0.56 0.16 RC 14 3.08 ** 0.77 ** 0.06 BB 7 5.85 0.49 * 0.47 DC 7 1.61 * 2.26 * 0.66 JC 11 1.94 1.74 * 0.17 RC BT 7 2.16 * 0.74 0.06 RC 13 1.62 * 2.75 * 0.47 BB 8 0.66 * 3.06 * 0.79 DC 10 1.41 2.62 * 0.46 JC 9 3.04 * 1.43 ** 0.28 BB BT 9 2.11 * 0.35 0.26 RC 13 1.69 1.82 0.13 BB 10 0.93 2.21 0.27 DC 11 1.34 * 2.00 ** 0.40 JC 11 2.79 * 1.22 ** 0.04 Core Installation P SEE AIC origin site BT BT 0.25 0.34 RC 0.39 0.44 BB 0.09 0.39 DC 0.03 0.62 2.05 JC 0.21 0.99 RC BT 0.60 0.59 RC <0.01 0.76 1.96 BB <0.01 0.17 6.07 DC 0.03 0.96 4.69 JC 0.15 0.66 BB BT 0.17 0.47 RC 0.22 0.78 BB 0.13 0.66 DC 0.04 0.58 -0.13 JC 0.54 0.70 Table 3. Parameter fits, correlation coefficients and error probabilities for regressions between heterotrophic soil respiration and volumetric soil water content (VWC) according to the equation [J.sub.sh] = a + b x VWC - c x [VWC.sup.2] for soil cores from the sites Bellenden Ker, top (BT), Robson Creek (RC) and Bellenden Ker, bottom (BB) installed at the sites BT, RC, BB, DC (Davies Creek) and JC (Smithfield) SEE, Standard error of estimation; AIC, Akaike information criterion for significant model fits. * P<0.05 Core Installation n a b c origin site BT BT 10 -9.38 0.66 -0.01 RC 14 -29.97 * 2.10 * -0.03 * BB 7 153.62 -9.42 0.15 DC 7 72.79 -4.95 0.09 JC 11 133.11 -7.35 0.10 RC BT 7 -47.28 3.19 -0.05 RC 13 -53.92 3.71 -0.06 BB 8 -244.67 13.64 -0.19 DC 10 211.50 * -13.08 * 0.21 * JC 9 -574.68 33.44 -0.48 BB BT 9 12.97 0.72 -0.01 RC 13 -1756.18 98.35 -1.37 BB 10 485.77 -26.77 0.37 DC 11 -9.58 0.87 -0.01 JC 11 -3.70 0.60 -0.01 Core Installation [R.sup.2] P SEE origin site BT BT 0.05 0.84 0.39 RC 0.70 0.01 0.26 BB 0.48 0.27 0.43 DC 0.70 0.09 0.65 JC 0.01 0.97 1.15 RC BT 0.47 0.28 0.50 RC 0.17 0.39 0.99 BB 0.30 0.42 0.34 DC 0.53 0.07 0.96 JC 0.56 0.09 0.56 BB BT 0.13 0.72 0.60 RC 0.26 0.23 0.76 BB 0.10 0.68 0.79 DC 0.13 0.57 0.75 JC 0.23 0.36 0.67 Table 4. Parameter fits, correlation coefficients and error probabilities for regressions between heterotrophic soil respiration and soil temperature together with volumetric soil water content (VWC) according to the equation [R.sub.sh] = ([R.sub.10] x [Q.sub.10.sup.(Ts - 10)/10)]) x (a + b x VWC - c x [VWC.sup.2]) for soil cores from the sites Bellenden Ker, top (BT), Robson Creek (RC) and Bellenden Ker, bottom (BB) installed at the sites BT, RC, BB, DC (Davies Creek) and JC (Smithfield) SEE, Standard error of estimation; AIC, Akaike information criterion for significant model fits. * P<0.05; ** P<0.01; *** P<0.001 Core Installation n [R.sub.10] [Q.sub.10] a origin site BT BT 10 9.28 0.20 -5.76 RC 14 11.55 0.94 * -2.78 BB 7 5.08 ** 1.56 * 6.56 *** DC 7 5.66 1.71 3.67 JC 11 47.08 *** 2.06 4.13 RC BT 7 7.13 1.03 -6.71 RC 13 2.34 2.57 * -3.24 BB 8 25.03 *** 3.68 *** 3.67 * DC 10 18.10 *** 3.08 * 2.33 JC 9 65.55 *** 1.14 -6.55 BB BT 9 8.81 *** 0.29 -6.26 RC 13 151.46 *** 1.57 -4.85 BB 10 13.10 *** 2.35 ** 4.87 DC 11 1.73 2.23 * -3.84 JC 11 2.58 1.49 4.83 ** Core Installation b c [R.sup.2] P SEE AIC origin site BT BT 0.38 -0.006 0.58 0.28 0.30 RC 0.19 -0.003 0.70 0.02 0.29 -5.67 BB -0.25 0.004 1.00 0.01 0.10 -5.81 DC -0.23 0.004 0.83 0.30 0.68 JC -0.23 0.003 0.24 0.75 1.16 RC BT 0.45 -0.007 0.47 0.78 0.70 RC 0.25 -0.004 0.48 0.21 0.88 BB -0.21 0.003 0.85 0.13 0.20 DC -0.15 0.002 0.89 0.01 0.55 3.84 JC 0.38 -0.005 0.58 0.38 0.66 BB BT 0.36 -0.005 0.45 0.68 0.61 RC 0.27 -0.004 0.28 0.56 0.83 BB -0.26 0.003 0.37 0.60 0.78 DC 0.29 -0.005 0.53 0.21 0.63 JC -0.19 0.002 0.40 0.48 0.68 Table 5. Parameter fits, correlation coefficients and error probabilities for regressions between pooled heterotrophic soil respiration rates ([R.sub.sh]) and soil temperatures together with volumetric soil water contents (VWC) according to the equations for soil cores from the sites Bellcnden Ker, top (BT), Robson Creek (RC) and Bellcnden Ker, bottom (BB) installed at all other sites SEE, Standard error of estimation; AIC, Akaike information criterion for significant model fits. * P<0.05; ** P<0.01; *** P<0.001 Core [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup.((Ts - 10)/10) origin n [R.sub.10] [Q.sub.10] [R.sup.2] P BT 49 0.99 *** 2.63 *** 0.50 <0.01 RC 47 1.71 *** 2.07 *** 0.37 <0.01 BB 54 1.29 *** 2.00 *** 0.40 <0.01 Core [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup.((Ts - 10)/10) origin SEE AIC BT 0.96 3.54 RC 1.12 9.65 BB 0.80 -5.43 Core [R.sub.sh] = a + 6 x VWC + c x [VWC.sup.2] origin n a b c [R.sup.2] BT 49 19.67 ** -1.14 * 0.019 * 0.16 RC 47 41.57 2.88 -0.045 0.05 BB 54 -0.53 0.29 -0.006 0.04 Core [R.sub.sh] = a + 6 x VWC + c x [VWC.sup.2] origin P SEE AIC BT 0.02 1.26 16.46 RC 0.29 1.38 BB 0.33 0.97 Core [R.sub.sh] - [R.sub.10] x [Q.sub.10.sup.((Ts - 10)/10)] origin x (a + b x VWC + c x [VWC.sup.2]) n [R.sub.10] 010 a b BT 49 0.91 2.62 *** 4.35 -0.20 RC 47 1.86 2.08 -5.28 ** 0.41 BB 54 0.78 1.87 *** -4.4Q 0.41 Core [R.sub.sh] - [R.sub.10] x [Q.sub.10.sup.((Ts - 10)/10)] origin x (a + b x VWC + c x [VWC.sup.2]) c [R.sup.2] P SEE AIC BT 0.003 0.60 <0.01 0.89 4.77 RC -0.007 0.41 <0.01 1.11 14.27 BB -9.007 0.39 <0.01 0.80 -1.97 Table 6. Parameter fits, correlation coefficients and error probabilities for regressions between heterotrophic soil respiration ([R.sub.sh]) and mean annual temperature (MAT), mean annual soil moisture (VWC), mean annual rainfall (rain) and a combination of MAT and rain as given by the equations for soil cores originating from Bellenden Ker, top (BT), Robson Creek (RC) and Bellenden Ker, bottom (BB) installed at the different sites SEE, Standard errors of estimation; AIC, Akaike information criterion for significant model fits. * P<0.05; ** P<0.01 [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup.((MAT - 10/10)] Core n [R.sub.10] [Q.sub.10] [R.sup.2] P origin BT 5 0.80 2.93 ** 0.70 0.08 RC 5 1.67 1.91 0.45 0.22 BB 5 0.93 ** 2.47 ** 0.98 <0.01 [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup.((MAT - 10/10)] Core SEE AIC origin BT 0.82 RC 1.13 BB 0.16 -3.03 Core [R.sub.sh] = a + b x VWC + c x [VWC.sup.2] origin n a h c [R.sup.2] BT 5 67.16 -4.59 0.079 0.72 RC 5 1176.25 -69.19 1.019 0.72 BB 5 202.98 -11.68 0.170 0.24 Core [R.sub.sh] = a + b x VWC + c x [VWC.sup.2] origin P SEE AIC BT 0.28 RC 0.28 BB 0.76 Core [R.sub.sh] = a + b x rain origin n a h [R.sup.2] P BT 5 4.20 ** -0.0004 0.75 0.06 RC 5 4.99 ** -0.0004 * 0.81 0.04 BB 5 3 54 -0.0003 0.62 0.11 Core [R.sub.sh] = a + b x rain origin SEE AIC BT 0.746 RC 0.658 3.07 BB 0.630 Core [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup. origin ((MAT - 10/10)] x (a + b x rain) n [R.sub.10] [Q.sub.10] a b BT 5 0.0049 2.23 300.75 -0.0250 RC 5 0.0089 1.21 438.62 -0.0347 BB 5 0.0040 2.25 ** 283.22 -0.0074 Core [R.sub.sh] = [R.sub.10] x [Q.sub.10.sup. origin ((MAT - 10/10)] x (a + b x rain) [R.sup.2] P SEE AIC BT 0.92 0.36 0.74 RC 0.83 0.52 1.10 BB 1.00 <0.01 0.01 -12.76
|Printer friendly Cite/link Email Feedback|
|Author:||Zimmermann, M.; Davies, K.; de Zimmermann, V.T.V. Pena; Bird, M.I.|
|Date:||May 1, 2015|
|Previous Article:||Assessment of tillage effects on soil quality of pastures in South Africa with indexing methods.|
|Next Article:||Effect of straw returning in winter fallow in Chinese rice fields on greenhouse gas emissions: evidence from an incubation study.|