Geophysical Research Letters

High alternative oxidase activity in cold soils and its implication to the Dole Effect


  • Alon Angert,

    Corresponding author
    1. Institute of Earth Sciences, Hebrew University of Jerusalem, Jerusalem, Israel
      Corresponding author: A. Angert, Institute of Earth Sciences, Hebrew University of Jerusalem, Givat Ram, Jerusalem 91904, Israel. (
    Search for more papers by this author
  • Mirco Rodeghiero,

    1. Sustainable Agro-ecosystems and Bioresources Department, Research and Innovation Centre, IASMA, Fondazione Edmund Mach, San Michele all'Adige, Italy
    Search for more papers by this author
  • Kevin Griffin

    1. Department of Earth and Environmental Sciences and Department of Ecology, Evolution and Environmental Biology, Lamont-Doherty Earth Observatory, Earth Institute at Columbia University, Palisades, New York, USA
    Search for more papers by this author

Corresponding author: A. Angert, Institute of Earth Sciences, Hebrew University of Jerusalem, Givat Ram, Jerusalem 91904, Israel. (


[1] Variations in the Dole Effect, which have been used to infer past changes in biospheric productivity, are strongly affected by isotopic discrimination in soil respiration. Respiration through the alternative oxidase (AOX) pathway is associated with a higher discrimination than the one associated with the “normal” dark respiration pathway (the cytochrome pathway, COX). However, observations of O2 discrimination and AOX activity in undisturbed natural environments are scarce. In the current study we measured the O2 concentration and stable isotopes in the root zone of tundra, boreal forest and alpine forest soils. To estimate the discrimination from this data, we have performed O2diffusion experiments in gamma-sterilized soil columns, with varying soil clay content. The discrimination found in the diffusion experiments was independent of clay content, and the value found, 14 ± 2‰, is the same as the one for binary diffusion of O2 in N2, indicating no interaction between the O2 and clay particles. Based on the field and laboratory results, the respiratory discrimination in the soils studied is 15–31‰, with the higher values associated with colder soils. The high discrimination found for cold (<6°C) soils indicates that AOX is a major respiratory pathway in these soils. This relationship between soil temperature and discrimination can be used in future interpretations of Dole Effect variations.

1. Introduction

[2] Soil respiration, the sum of respiration by heterotrophic organisms (bacteria, fungi) and living roots, is the dominant terrestrial source of CO2 to the atmosphere. This fact, and the sensitivity of global climate to the atmospheric concentrations of CO2, has directed much attention to this process in recent decades. Respiration in the soils of boreal forests and tundras makes a significant contribution to global soil respiration [Bond-Lamberty and Thomson, 2010; Raich and Potter, 1995]. Thus, it is highly important to understand the processes that govern respiration in these cold soils. Angert et al. [2003] reported high contributions of the alternative oxidase (AOX, see below) respiratory pathway in the soils of two adjacent boreal forest sites, based on estimates of high O2 discrimination (22.5 ± 3.6‰) in these soils. In agreement, Lee et al. [2003] reported high discrimination at northern Canada in a waste pile (of mostly excavated sand) covering dewatered organic lacustrine sediments. By contrast, low discrimination values have been reported for tropical soils [Angert et al., 2003]. This contrast between tropical and northern sites has been used by Severinghaus et al. [2009] to support his suggestion that past variations in the Dole Effect are mainly driven by changes in monsoon activity, which strongly controls tropical forests productivity and soil respiration. However, this important assumption is currently based on limited data, with only two natural boreal forest sites, which are located close to each other, and therefore requires further verification from a larger data set.

[3] The low O2 discrimination in tropical soils results from high respiration rates relative to the rates of O2 diffusion into roots and soils aggregates which induce a low O2 concentration at the site of consumption [Angert and Luz, 2001; Angert et al., 2001]. This effect of diffusion on the discrimination in O2 uptake is similar to the well known isotopic effect which takes place in CO2 diffusion and uptake in leaves [Farquhar et al., 1982]. The high discrimination values in the boreal forest soils were explained by high engagement of the AOX, which consumes O2 with a discrimination of 25–30‰ that is considerably higher than the ∼20‰ discrimination in “normal” dark respiration through the cytochrome oxidase pathway (COX) [Ribas-Carbo et al., 1997].

[4] The COX and AOX represent two branches of the electron transport chain of mitochondrial respiration in plants, some fungi and some bacteria. Both the COX and AOX are responsible for the terminal step in the chain, which transfers electrons to O2, reducing it to water. The AOX pathway (also known as the cyanide-resistant pathway) bypasses two sites of trans-membrane proton pumping and, as a result, produces at most only one third of the ATP produced by the COX pathway. While the precise ecological function of AOX activity is unknown, it is thought to be generally related to stress [Moore et al., 2002; Rachmilevitch et al., 2007] and has been shown to suppress the production of reactive oxygen species [Vanlerberghe and McIntosh, 1997]. The AOX is also believed to be engaged when plants produce excess of organic acids [Lambers et al., 2008]. Ambient temperatures also influence AOX activity, with in vivo electron partitioning through the pathway typically increasing under cooler temperatures [Armstrong et al., 2008; Gonzalez-Meler et al., 1999; Ribas-Carbo et al., 2000; Searle and Turnbull, 2011; Searle et al., 2011].

[5] Aggarwal et al. [2004] suggest that discrimination in soils is strongly controlled by the interaction between the O2and clay particles in the soil column. According to this idea the soil column acts similar to a gas-chromatograph (GC) column, and fractionates the O2 isotopes as they diffuse through the soil. It should be noted that if this claim is correct, then the estimates of soil discrimination of Angert et al. [2003] are biased, since these estimates assume that the discrimination of O2diffusing through the soil air-filled pores is identical to that of O2 diffusion through air. This assumption is included in the equation used to estimate the soil O2 discrimination against 18O in soil respiration with respect to soil air (Dsoil) [Angert et al., 2003]:

display math

where δ18O values are relative to the atmosphere, and subscripts “s” and “a” stand for soil air and atmosphere respectively. Ddiff is the discrimination factor in diffusion of O2in air (14‰ - derived from the theory of binary diffusion of gases [Mason and Marrero, 1970]), and the term δ18Otd stands for the δ18O of soil air that would be expected due to thermal diffusion only, and is calculated as in Angert et al. [2003]. If the soil column does behave as a GC column, then the strength of interaction with the O2, and hence the discrimination resulting from diffusion through the soil, are expected to be temperature dependent. Thus, it is possible that the different discrimination values estimated for cold and warm soils might be an artifact of this effect. For example, if this effect caused a shift of 10‰ in Ddiff toward lower values in cold soils, the estimated discrimination will also be 10‰ lower (equation (1)), which will make the values of boreal soils similar to those of tropical soils.

[6] The goal of the current study is to address the two problems outlined above. First, we have conducted O2 diffusion experiments with natural soils, at different temperatures, to determine if the discrimination in diffusion through the soil column deviates from the discrimination for diffusion in air. And second, we have extended the range of cold soils studied by conducting research at tundra and boreal forest sites in northern Alaska, and alpine forest sites along altitudinal temperature gradient in northern Italy.

2. Materials and Methods

2.1. Field Study

[7] Soil air was collected at 3 sites in Alaska on June 27–30th 2010, and 3 sites in Italy during June 7–10th 2011. The first two sites in Alaska (AK-CT, and AK-GH) were tussock tundra sites near Toolik Lake (68°38′N, 149°34′W, elevation 780 m), and are part of a long-term field-manipulation that began in 1989 at the Arctic Long-Term Ecological Research site. In this site the mean annual air temperature is −7.2°C, and the average annual rainfall is 250 mm. Two experimental plots were used; one served as an ambient temperature control (AK-CT), and the second a warmed plot where the growing season temperature has been passively increased by an average of 4°C using a small greenhouse (AK-GH). The greenhouses were constructed of a rectangular wood frame (∼2.5 × 5 m) with a 1.5 m tall gabled roof and 65 cm tall sides. The wooden frame is covered each growing season with transparent 0.15 mm (6 mil) plastic sheeting (Cloud 9 commercial greenhouse plastic; Monsanto, Incorporated, St. Louis, Missouri, USA) from late May, following snowmelt, to late August. The effects of these structures on air and soil temperatures, relative humidity, and canopy PAR have been described elsewhere [Hobbie and Chapin, 1998; Chapin et al., 1995]. Although the main effects include a 20% reduction in PAR, increased humidity and decreased turbulence, the canopy structure and allometry of individual species, appear to be little affected [Shaver et al., 2001; Bret-Harte et al., 2002] and model analyses of the effects of the light reduction indicate that it has little effect on C cycling [McKane et al., 1997]. The ambient environmental conditions of both plots have been logged on an hourly and daily basis, year-round since the start of environmental treatments in 1988 with a Campbell 21x datalogger (Campbell Instruments, Logan, UT, USA). The third site in Alaska was a boreal forest site located 16 km north of the town of Coldfoot Alaska (67°23′N, 150°05′W, elevation 481 m). The area is a typical boreal forest with small individual spruce (Picea sp.) and alder (Alnus sp.) trees on a gentle SW facing slope to the east side of the Dalton Highway.

[8] The three sites in Italy are described in detail by Rodeghiero and Cescatti [2005] where they are referred to as S6, S8 and S11. Site S6 is European beech forest (Fagus sylvatica L.) while sites S8 and S11 are Norway spruce (Picea abies (L.) Karsten) forests. The mean annual air temperature is 8.6°C, 5.9°C, and 4.2°C, and the average annual rainfall is 976, 1015, and 1008 mm for S6, S8 and S11, respectively.

[9] Soil air was sampled by a stainless steel tube (with 10 mm inside-diameter and 12.5 mm outer-diameter) that was hammered into the soil, as inAngert et al. [2001]. The tube end was pointed to ensure easy insertion, and 2-mm-diameter holes were drilled above the pointed end for soil air collection. An 8-mm-diameter plastic rod inserted inside the tube reduced its dead volume. Samples of the soil air were collected in pre-evacuated ∼3.6 mL glass flasks with a Louwers-Hapert™ O-ring valve. Before sampling, the dead volume in the tubing and flasks' necks were purged with soil air.

[10] Sample preparation and mass spectrometry were performed according to Barkan and Luz [2003]. The preparation of the sample included cryogenic removal of water vapor and CO2 and chromatographic separation of N2 by a fully automated system. Elimination of N2 prevented the need for correction for the effect of N2 interference in the ion source of the mass spectrometer. The oxygen concentrations were calculated from the ratio of O2 to Ar (expressed as δO2/Ar), under the assumption of constant Ar concentration. All measurements were performed on a Finnigan-MAT Delta-Plus (Thermo Scientific, Waltham, MA, USA) dual-inlet mass-spectrometer. The precision inδO2/Ar was 1‰, and the precision in δ18O determination was 0.03‰. All the results are presented with respect to atmospheric air standard [Barkan and Luz, 2003].

2.2. Diffusion Experiments

[11] Soil columns were prepared by filling a glass tube, (8 cm long, 0.6 cm outer diameter, 0.4 cm internal diameter) with 2.0–2.4 g loose soil (or sand). The soils samples were: 1) Terra-rossa soil (correlates with chromatic luvisols in the FAO classification system) with clay content of 49%, sampled at a site with natural vegetation and Mediterranean climate in Judean mountains (31°42′N, 35°3′E); 2) Sample from site S6 - clay content 42%; 3) Sample from site S8 - clay content 31%; 4) Sample from site S11 - clay content 10%; and 5) Acid-washed sand (Merck) - clay content 0%. The soils were sterilized by gamma radiation from a Cesium-137 source for at least 5 hours. Overnight incubation of the gamma-treated soils showed no CO2 emission and no O2consumption even after re-wetting the soils, which indicates that the sterilization was successful.

[12] The experiments were conducted with both dry and wet soils. In the wet soils experiment deionized water was added in a proportion of ∼1/8 of the weight of the soil. Plugs made of alumina wool were inserted in both ends of the glass tube to keep the soil in place, while allowing air movement. The soil column was connected to a 3.6 mL glass flask equipped with a Louwers-Hapert™ O-ring high-vacuum-valve. The other end of column was connected through a Y connector (Figure 1) to N2 flow. This system was placed inside a thermostatic cooling bath, and the N2 flow to the system passed through a heat exchanger. The bath was set to either 3°C, to simulate the conditions in cold soils, or to room temperature (20°C). Before an experiment started, the tubes and the soil column were flushed with N2. Then the N2 flow was temporarily stopped, the soil column was connected to the flask, the flask was opened, and the N2 flow was resumed. During the experiment O2 diffused out of the flask, through the soil column, and was then removed by the N2 flow. After 30–45 minutes from the beginning of the experiment the flask was closed and taken for analysis. The experiments were done in identical pairs, and in each pair one flask was taken for O2 concentration analysis and one for O2 δ18O measurement.

Figure 1.

Schematic representation of the soil-column diffusion experiments setup.

[13] The method for determination of O2 δ18O for the diffusion experiments flasks was identical to that described above for the field study flasks. However, since Ar was not conservative during the diffusion in N2 experiments, the O2/Ar ratio could not be used as a measure of O2 concentrations. In these experiments the O2concentration was estimated on one flask from each pair by a fuel-cell based O2analyzer (Sable Systems FC-10) connected to an air circulating system similar to that described inAngert and Sherer [2011]. The analyzer [O2] reading was corrected for the system's internal pressure and for dilution by water vapor. Water vapor concentration was determined by a Li-840A (LI-COR, Lincoln, NE, USA) infra-red-gas-analyzer, through which the air flow in the circulating system passed before entering the oxygen analyzer. The accuracy and precision in [O2] determination by this method was estimated by measuring duplicates of environmental samples by both this system and by the O2/Ar method, and was found to be ±0.04% O2 (data not shown). The discrimination (D) was calculated from the Rayleigh distillation equation. The flasks were vented before each experiment, so the starting point could be taken as atmospheric air, and conducting the experiments under water prevented possible leaks.

3. Results

[14] The O2 concentrations and O2 δ18O values measured in the soil profiles at the field sites are given in Table 1. The O2 concentrations ranged from 9.24% to 20.35%, while the δ18O values ranged from 0.33‰ to 5.92‰ (vs. atmospheric air). The results of the soil columns diffusion experiments are summarized in Table 2. The O2 concentrations ranged from 12.89% to 18.16‰ and the δ18O values ranged from 1.83‰ to 6.89‰. The discrimination calculated for these diffusion experiments ranged from 12‰ to 17‰, with an average of 14‰ and a standard deviation of 2‰.

Table 1. Summary of the Field Experiments Including Ecosystem Type, the Temperature at the Soil Surface (Tsurface), and at the Sol-Air Sampling Depth (Tsoil), and the Isotopic Discrimination Calculated According to Equation (1)
S6Alpine F.19.160.659.49.322
S6Alpine F.19.410.519.49.321
S6Alpine F.19.060.389.28.818
S6Alpine F.19.140.339.28.818
S8Alpine F.20.350.434.94.529
S8Alpine F.19.990.705.75.429
S8Alpine F.19.940.595.75.426
S8Alpine F.20.210.515.55.128
S8Alpine F.20.110.695.55.131
S11Alpine F.19.270.618.25.6521
S11Alpine F.18.730.878.25.6522
S11Alpine F.19.900.956.73.531
S11Alpine F.19.630.856.44.327
AK (GH)Tundra9.245.9216.94.524
AK (GH)Tundra9.593.7516.94.520
AK (CT)Tundra18.871.1710.23.024
AK (Bo)Boreal F.18.680.5723.210.816
AK (Bo)Boreal F.17.280.5023.210.815
Table 2. Summary of the Soil Column Diffusion Experimentsa
SoilMoistureT (°C)[O2] (%)δ18OD (‰)
  • a

    The average discrimination was found to be 14‰ as expected for diffusion of O2 in air.

Judean M.dry2016.713.4615
Judean M.dry316.812.9713
Judean M.dry318.161.8313
Judean M.dry314.815.0014
Judean M.moist316.193.6914
Judean M.moist315.054.3213
Judean M.moist315.295.1416
Average    14 ± 2

4. Discussion

[15] The O2 discrimination values of the soil respiration at the field sites were estimated by equation (1), from the measured O2 δ18O values and O2 concentrations (Table 1), based on the assumption of discrimination in diffusion through the soil of 14‰ (the value for binary diffusion of O2 in N2). These calculated discrimination values are presented in Figure 2, against soil temperature and compared to a previous estimate of soil discrimination values [Angert et al., 2003]. The values found in the current study show the same tendency for increasing discrimination with lower temperature as in Angert et al. [2003], but extended it to lower temperature and higher discrimination values. In that previous study the lowest discrimination ranged from an average of 10.1‰ for tropical forest to an average of 22.5‰ for boreal forest (with Tsoil = 8.4 to 19.4°C), while in the current study the discrimination ranged from 15.1‰ (at Tsoil = 10.8°C) to 31‰ (at Tsoil = 3.5°C). The value of R2 for a linear regression line of D versus Tsoil in the combined dataset is 0.59.

Figure 2.

The discrimination (D) estimated for the field sites in Angert et al. [2003] and in the current study (Italy Alps and Alaska).

[16] The relationship between colder soils and higher estimates of discrimination in O2 uptake can be explained in two possible ways. First, if the discrimination in diffusion of air through soil is indeed 14‰, then this estimate of high discrimination in cold soils is correct, and needs to be explained. Second, it is possible that as Aggarwal et al. [2004] suggested, interactions of O2 with the soil clay particles cause isotopic discrimination in the transport rate, resulting in effective value for discrimination of O2 in diffusion through soils which is lower than the value for diffusion through air or through porous inert material. If this interaction is stronger in cold temperature and can cause a shift in the order of 10‰ in Ddiff, it may explain our field observations with no need to invoke a mechanism with high discrimination in O2 uptake.

[17] Our soil column diffusion experiments were designed to determine if this is the case, or if the discrimination for diffusion of O2 through soil columns is ∼14‰ even in low temperatures and in soils that contain clays. The discriminations calculated for the diffusion experiments are summarized in Table 2. The average discrimination was 14‰ ± 2‰, with no significant difference between sand and clay containing soils. There were also no relationship between the discrimination and clay content of the soil, the experiment temperature, nor any clear difference between wet and dry soils. Overall the results indicate that diffusion of O2 through soils cause a discrimination of 14‰, as expected for diffusion through inert porous medium, and that no additional effects are associated with possible interaction with the clays. Given these results, we have decided that there is no point in conducting a full factorial experiment in which all soils will be tested over a range of temperatures and water contents. The interaction effects described in Aggarwal et al. [2004]may be related to that study experimental setup, soil pre-treatment (drying and sterilizing at 110°C for 24 hours), or the assumptions or structure of the model used to estimate the discrimination. The lack of any clay-interaction effect on the magnitude of 10‰ needed to explain our observations in cold-soils sites with typical discrimination of COX (∼20‰) indicates that the discrimination in uptake in sites with Tsoil < 6°C was extremely high (20.4–31‰). These results indicated the high discrimination reported in Angert et al. [2003] for two adjacent boreal forest in Alaska is probably a general phenomena in cold soils, since it was also found for the fairly different tundra and alpine forest sites.

[18] Such high discrimination values have not been observed for the COX pathway over a wide range of organisms ranging from plants [Ribas-Carbo et al., 1997] to marine eukaryotes and marine bacteria [Kiddon et al., 1993]. In the tundra soils, there can be additional O2 consumption due to oxidation of reduced species produced deeper in the profile, in the waterlogged soil above the permafrost. However, such oxidation cannot explain the high discrimination we found, since: 1) The high D values were also found in the alpine soils which are well drained; and 2) the discrimination values associated with this processes are actually lower than that of the COX: 4.5–12‰ for abiological reduction by aqueous sulfide or ferrous iron [Oba and Poulson, 2009a, 2009b], 16.1–17.5‰ for biological methane oxidation [Mandernack et al., 2009], and 16.1‰ for biological hydrogen sulfide oxidation [Luz et al., 2002]. The only known process with high discrimination matching the high values estimated for the cold sites is the AOX pathway.

[19] A simple mass balance can be made if we neglect the effect of diffusion limitation and assume discriminations of 20‰ for the COX and 30‰ for AOX (although a value of 25‰ was reported for roots [Ribas-Carbo et al., 1997], the value for the soil microbiota can range from 25–30‰). Based on such balance, a discrimination of 25‰ indicates that about 50% of the soil respiration is through the AOX, and discrimination of ∼30‰ indicates that approximately 100% of the respiration is through the AOX. Although such high rates of AOX engagement may seem surprising, there is a scarcity of studies of AOX activity in undisturbed natural environments and thus little data to compare this with. Such studies are only possible by estimating the O2discrimination with high-accuracy field measurements combined with simple modeling. Other field studies also report evidence of strong discrimination. For example,Luz and Barkan [2011] reported discrimination values of ∼27‰ for the ocean mixed layer in various sites during or just before spring blooms, and Angert et al. [2012]report discrimination values ranging between 12.6‰ and 21.5‰ for tree-stem respiration. The lower range indicates the effect of diffusion, which limits the effective discrimination, while the higher end probably indicates that the large discrimination associated with the AOX can overcome the diffusion effect. High discrimination (24–27‰) was also reported for an artificial cold site (Tsoil ∼ 1°C) of a mine waste pile [Lee et al., 2003]. By contrast, “inversed discrimination” (D = −30‰) was reported below the roots zone (depth of 1–15 m) of boreal forest [Wassenaar and Hendry, 2007], and was suggested to be related to processes other than respiration.

[20] Currently we cannot determine the biological source for the high engagement of AOX in the cold soil sites we studied, as it may be a response of either the microbiota and/or of the plants-roots for cold stress [Armstrong et al., 2008; Gonzalez-Meler et al., 1999; Ribas-Carbo et al., 2000; Searle and Turnbull, 2011; Searle et al., 2011]. Since the AOX pathway produce the same reductive power (NADPH) as COX pathway, but just third of the ATP, it was also suggested that the AOX will be important under conditions in which the demand for carbon skeletons is high and the demand for ATP is low - for example when roots excrete large amounts of organic acids [Lambers et al., 2008]. Future studies are needed to determine if the high AOX engagement is in roots, the microbiota, or both. In addition, studying the temporal variations in the AOX activity in relationship to other environmental parameters may provide important clues regarding the function of this ubiquitous but, thus far poorly understood respiratory pathway.

[21] Regardless to the questions of biotic control on the AOX, the high discrimination we found has implications for the Dole Effect. Our results support the suggestion of negative correlation between soil temperature and the soil respiratory O2 discrimination [Angert et al., 2003], and the use of this correlation to infer paleo-changes in the land biosphere from past variations in the Dole Effect [Severinghaus et al., 2009].


[22] We thank Leonardo Montagnani, Stefano Minerbi, Mauro Cavagna, Cristina Martinez and Matthew Turnbull for their help with field sampling, Shunit Mazeh for help with diffusion experiments, and Eugeni Barkan for assisting with the O2 stable isotopes analysis. AA was supported by ISF grant 870/08 and by Ring Foundation grant, and KLG was supported by project 0732664 (IPY) from the National Science Foundation.

[23] The Editor thanks three anonymous reviewers for assisting in the evaluation of this paper.