The effects of glucose and mineral nitrogen amendment on the respiratory quotient (RQ), defined as the ratio of moles CO2 evolved per moles of O2 consumed, were tested on an agricultural and a forest soil (A horizon) of the Bornhöved Lake District, northern Germany. Both substrate-induced respiration rate and RQ value increased with increasing glucose concentration. Glucose plus nitrate addition gave the highest RQ value in the agricultural soil whereas addition of glucose alone gave the highest RQ value in the forest soil. Generally, glucose plus ammonium addition resulted in lower RQ values than glucose addition alone. When glucose was added to soil at field rates of C input, the RQ value showed fluctuations between 0.44 and 1.35. Low C supply caused rapid return of the initially enhanced RQ value towards those of basal respiration, that was typically <1. The experiment showed that both respiratory activity and respiratory quotient increased with increasing amount of available C. The RQ values >1 refer to both anabolic C uptake and ‘unbalanced’ C degradation. The microbial communities in the agricultural soil were less efficient in C use than those in the forest soil.
Respiration is a key process in the global carbon cycle and of crucial importance in the partitioning of energy in soil. Respiratory processes consume O2 and liberate CO2 and lead to loss of C from ecosystems to the atmosphere. In soil, organic matter, readily available litter and plant exudates are transformed and mineralised via respiratory processes mainly by the microbiota.
Under aerobic conditions, respiration encompasses the glycolysis and the subsequent oxidative decarboxylation of pyruvate, feeding into the citric acid cycle coupled to a respiratory chain (Fig. 1). When all these reactions are fully involved and alternative electron acceptors like NO3−, Fe3+, … are hardly used, the respiratory quotient (RQ) for substrates such as glucose is 1. Under these conditions, degradation of the substrate is considered as ‘balanced’ since a number of moles of CO2 are consumed as O2 are evolved. However, soil contains a wide spectrum of substrates  that may be transformed and immobilised, and their complete oxidation may also be retarded by environmental and nutritional factors. Thus, soil management may modify or ‘unbalance’ the RQ .
Soil respiration rate is commonly determined either on the basis of the rate of CO2 evolution or on the rate of O2 uptake. Then, the RQ is frequently assumed to have a value of 1 though in soil this may not be the case owing to environmental conditions . For example, the RQ value decreased with the decline of the initial phosphate level in a batch culture with Azotobacter vinelandii. How the RQ of soil varies when C and N availability is varied has not been extensively investigated. Investigations on factors controlling RQ values are needed since data on soil respiration are fundamental for evaluating soil quality and analysing organic matter transformation in ecosystems.
Glucose application generally leads to the induction and de-repression of metabolic processes and to the stimulation of microbial growth and enzyme activity [5,6]. Concurrently, RQ values were frequently approximately 1.3 between 4 and 24 h after substrate addition . In microbiological assays for the determination of microbial biomass and the active component either based on CO2 evolution or O2 uptake over 0–30 h, however, the glucose concentration is often varied and nutrients such as nitrogen and phosphorus may be added . In general, soil microbiological methods are usually carried out using high doses even though under natural conditions nutrient concentrations are generally limiting.
To estimate differences between CO2 and O2 data and to evaluate the physiological response of soil microbiota, this paper addressed soil microbial RQ values (i) at glucose concentrations typical for soil microbiological assays such as the substrate-induced respiration (SIR) method , (ii) with simultaneous glucose and nitrogen addition typical for studies on soil communities , and (iii) at relatively low glucose availability corresponding to C input under field conditions.
Materials and methods
The soils were typical for northern Germany and sampled in the ‘Bornhöved Lake District’ (54°06′N, 10°14′E). The landscape has a moderate oceanic climate with a long-term annual temperature of 8.3°C and 757 mm precipitation. The agricultural soil was under crop rotation (mixture of Lolium perenne L. and Festuca rubra L. in the sampling year), fertilised with organic manure previously and with 40 kg NH4NO3–N ha−1 in the sampling year. The forest soil was under beech (Fagus sylvatica L.). The forest topsoil was located below a litter horizon of approximately 5 cm thickness. The arable soil was an Eutri-cambic Arenosol  with a pH [H2O] of 6.4, an organic C of 14.4 mg C g−1 dry soil, a C/N ratio of 10 and organic C extracted with 0.5 M K2SO4 was 19 μg C g−1 soil. The beech forest soil unit was a Dystri-cambic Arenosol with a pH of 4.1, an organic C of 33.8 mg C g−1 dry soil, a C/N ratio of 14 and extractable organic C was 83 μg C g−1 soil. The microbial communities in these two soils showed different metabolic, anabolic and enzymatic responses to nutrient addition [5,11].
Three independent soil samples consisting of eight to 10 soil cores each were taken from each site. After removal of living plant residue, the fresh soil was sieved (<2 mm) and mixed. Samples were stored at 4°C for no longer than two months until analysis. The samples were generally pre-conditioned for about 3 days at laboratory temperature (approximately 22°C) before analyses. The water content was 17 and 23% H2O g−1 dry soil for the agricultural and the beech forest soil, respectively, which corresponds to 224±15 and 133±27 hPa (pF 2.4 and 2.1) and 31±2 and 34±7% water-filled pore space. The water content was constant during the experiment and considered to enable aerobic metabolism .
The oxygen uptake by the soil was continuously monitored in a Sapromat respirometer (Fa. IBUK, Königsbronn, Germany). Sodium hydroxide (1 M) was used to absorb CO2. The total CO2 produced during the course of the experiments was determined from the volume of 0.1 M HCl required to titrate the alkali absorbent until phenolphthalein became colourless.
After determining basal respiration for approximately 24 h, the alkali trap was exchanged and glucose and nitrogen were added. Thereafter, the alkali trap was again exchanged after approximately 4, 24, 48, 72 and 144 h.
The following three experiments were carried out:
Glucose added at rates of 0.2, 0.5, 2.0 and 5.0 mg C g−1 soil analogous to glucose concentrations used when estimating microbial biomass by SIR to induced maximal respiratory response.
Glucose addition inducing maximal initial respiratory response plus N supplement. The 2 mg glucose-C g−1 soil was considered as being optimal for maximal initial respiratory response and active biomass determinations . In addition, nitrogen was added either as (NH4)2SO4, KNO3 or NH4NO3 at a C:N ratio of 5:1 [w/w], which occur in microbial tissue .
Glucose was added in one pulse representing field rates and was much lower than used for SIR in (i). The annual input was estimated to be 3600–5067 kg C ha−1 in a maize monoculture and field crop rotation, respectively . The 5067 kg C ha−1 is equivalent to approximately 1.29 mg C g−1 soil for 30 cm soil depth and the bulk density of 1.3 g cm−3 and, thus, 0.025, 0.05, 0.1 and 0.2 mg glucose-C g−1 soil (here mixed with 1 mg talcum g−1 soil and added as one dose) corresponded to the approximate C input of 1, 2, 4 and 8 weeks. Similar annual C inputs ranging between 3620 and 5180 kg C ha−1 were determined in the beech forest .
Three independent samples of each soil were analysed. Statistical analyses were performed using Microsoft Excel 2000 and SigmaStat (Jandel Scientific, Erkrath, Germany).
Results and discussion
Correlation between O2 uptake and CO2 evolution rate
For the data from the three sets of experiments, high correlation coefficients were determined between soil O2 uptake and CO2 evolution rate. The lowest Spearman rank order correlation coefficient was calculated for basal respiration (r=0.22, P=0.069, n=72). The coefficient was r=0.83, 0.99, 0.99, 0.98 and 0.97 for 4, 24, 48, 72 and 144 h after glucose addition, respectively (P<0.001, n=72). Despite the fact that significant correlations between the two methods were detected, microbial RQ values – that is, the ratio between the two correlated activities – varied significantly in the soils dependent on nutritional conditions, as subsequently discussed.
The RQ value was frequently <1 when estimating basal respiration (Figs. 2–4). On average, the RQ value was here 0.77±0.06 and 0.80±0.05 for the agricultural and forest soil, respectively (confidence limit, P<0.05, n=36). Thus, CO2 evolution rate was 20–23% lower than O2 consumption rate. This shows that soil respiration rates on the basis of the rate of CO2 evolution or on the rate of O2 uptake may vary considerably and, furthermore, the soil microbial physiology in these soils has a relatively high oxygen requirement for basal metabolism. These results confirmed previous investigations . The RQ data were similar to those of Sabra et al.  studying Azotobacter vinelandii in laboratory experiments but generally higher than those reported by Aon et al.  from field studies with soils subjected to conventional or no-till management. Aon et al.  reported RQ values as low as 0.27.
The low RQ values suggest that aliphatic organic compounds, amino acids or refractory compounds containing relatively low O content were predominantly mineralised . The elemental composition of humic acids in soil may be approximated as C308H328O90N5. This corresponds to a C/O ratio [mol mol−1] of approximately 3.42. Dependent on the degradation pathway, their oxidation concurs with RQ values lower than 1, RQ being 0.29 and 0.906, respectively . In addition, nitrification may have lowered the RQ value by extra O2 uptake. Studies with Thiosphaera pantotropha showed that the amount of oxygen necessary for nitrification was 11% of the total oxygen uptake . It is, however, necessary to recognise that the respiratory quotients were here determined at 22°C. According to Chapman and Thurlow , a temperature lower than 22°C increases the RQ and may favour conditions for the mineralisation of cellulose rather than for lignin.
Glucose at high concentrations (typical for SIR)
Shortly after the addition of high glucose concentrations, the RQ value approached 1 with values of 0.98±0.08 and 0.95±0.10 (confidence limit, P<0.05, n=6) for the agricultural and the beech forest soil, respectively (Fig. 2). The results of this ‘maximal initial respiratory response’ indicate that the two gases can both be reliably used to estimate the maximal initial respiratory response that is a physiological estimate for the soil microbial biomass . This conclusion confirms previous investigations [7,22], although extreme RQ values of up to 0.84 were occasionally determined representing 16% differences between CO2 and O2 data. Therefore, caution is required when comparing biomass data on CO2 evolution with those on O2 uptake.
After the first 4 h, the RQ value increased over time and was highest for the highest glucose concentration, with 1.40 in the agricultural soil after 1 day and with up to 1.44 in the forest soil after 2 days. The microbial communities in the beech forest soil seem to have the higher C demand.
The respiratory quotient >1 after the addition of glucose can be partly explained by the occurrence of anabolic processes and the related carbon entering the biomass. For pure cultures, Herbert  gave the following elementary balance equation for complete glucose anabolism: 0.314 C6H12O6+0.75 O2+0.19 NH3→CH1.82O0.47N0.19+0.90 CO2+1.18 H2O. This equation corresponds to the RQ value of 1.2. Values higher than 1.2 may be attributed to ‘unbalanced’ glucose degradation (Fig. 1) and dominant specific metabolic pathways producing much more CO2 relative to O2. Soil microbial communities may produce more moles CO2 relative to the uptake of moles O2 by performing to a larger extent substrate decarboxylation to obtain precursor metabolites for growth [24,25]. Low rates of oxidative phosphorylation via the respiratory chain for ATP production or alternative electron acceptors such as NO3−, Fe3+, SO42- or organic acids result in a low O2 demand (Fig. 1).
For the lowest glucose concentration, the RQ value declined after 2 days, as did respiratory activity (Fig. 2), most likely attributed to substrate depletion. During the 6 days after substrate addition the RQ values varied between the treatments between −11 and 9% (Table 1).
|RQ [mol CO2 mol−1 O2]|
|0.2 C||0.8 C||2.0 C||5.0 C|
|2 C||2 C+0.4 NH4+||2 C+0.4 NO3−||2 C+0.4 NH4NO3|
|0.025 C||0.05 C||0.1 C||0.2 C|
|RQ [mol CO2 mol−1 O2]|
|0.2 C||0.8 C||2.0 C||5.0 C|
|2 C||2 C+0.4 NH4+||2 C+0.4 NO3−||2 C+0.4 NH4NO3|
|0.025 C||0.05 C||0.1 C||0.2 C|
Glucose plus nitrogen
When nitrogen was added together with glucose, the respiration rates were much higher. Simultaneously, the RQ value increased more steadily and persisted for a longer time than glucose alone (Fig. 3). This was particularly evident for the agricultural soil. Glucose and glucose plus nitrate induced the overall highest RQ values in the forest soil after 2 days. In contrast, ammonium and ammonium plus nitrate significantly reduced the respiratory quotient within the first day. Compared with nitrate addition, the ammonium amendment reduced the RQ value by 16% at day 1 for the agricultural soil and by 18% at day 2 for the forest soil. Since microbial growth was similar for all treatments with nitrogen in the agricultural soil (data not shown), nitrate may have been used more extensively as an alternative electron acceptor by microbial communities in the agricultural than in the forest soil, leading to higher RQ value by the reduced O2 uptake. However, N extractable with 0.5 mol potassium sulfate in the non-fumigated soil was lower in the ammonium treatment (data not shown).
Approximately 4 h after glucose addition, respiratory activity increased continuously and maximal respiration rates were often observed after around 24–48 h (Figs. 2 and 3). The N addition stimulated respiration and microbial growth, which agrees with earlier findings for these soils [11,26]. With a factor of 1.3, the active microbiota or the growing populations may be overestimated by 30% when determining CO2 rather than O2[27–29]. In addition, more than 15% lower values will be achieved when using ammonium sulfate in comparison to potassium nitrate (Table 1) as used by Stenstrøm et al. . During the 6 days following substrate addition, the RQ values differed between the treatments between −5 and 20% (Table 1).
Low glucose concentration (related to field input)
The glucose concentration corresponding to the 1-week field rate significantly increased microbial respiration rate over 4 h (Fig. 4). The maximal initial respiratory response increased with increasing C concentration and RQ values were highest for the highest glucose concentration and lowest for the lowest glucose concentration. The RQ values ranged between 0.44 and 1.35. The glucose concentrations related to field input produced significantly lower mean RQ values than the glucose concentration typical for SIR (Table 1).
The respiration rate over the first 4 h continued only for 24 h when the 8-week field rate was added. When the same substrate concentrations are compared, the respiratory activity and RQ value continued at a higher level in the beech forest than in the agricultural soil. Thus, the microbial communities in the forest soil were able to use the C source longer for metabolism.
Soil respiration rate returned to the initial level after 1, 2 and 3 days for the 2, 4 and 8-week dose. When respiration rate declined, the RQ value became concomitantly lower. The RQ values were >1 during the first 4 h for the 8-week dose. For all other treatments, the RQ value went up to approximately 1 and then declined with minimal values of approximately 0.5. This indicated that a higher C supply induced a higher RQ value due to anabolic processes. The RQ value differed from 1 between −51 and −6%. The highest differences occurred with lowest nutrient addition (Table 1).
The agricultural soil showed a smaller increase in microbial biomass after glucose amendment (Table 2). This indicates that the microbial communities had a lower anabolic efficiency . In addition, the agricultural soil depleted the added substrate earlier than the forest soil (Table 2). Since similar microbial biomass contents were estimated in un-amended soils (Table 2) the microbial communities in the forest soil seem to have a higher anabolic efficiency. This was related to a lower proportion of the glucose-responsive biomass and the prevalence of K-strategists . The higher the glucose concentration was, the greater was the increase in RQ value and microbial biomass content.
|Nil||25 C||50 C||100 C||200 C|
|Nil||25 C||50 C||100 C||200 C|
Soil respiration measurements frequently gave RQ values ≠1 during basal metabolism and microbial growth, although O2 uptake and CO2 evolution rate were generally highly correlated. The RQ value during the maximal initial respiratory response approached 1, indicating that both CO2 evolution and O2 uptake may be reliably used for the microbial biomass using the SIR method. However, RQ values may vary significantly from 1 and between soils.
During glucose-stimulated microbial growth, the RQ value was generally >1, which suggests unbalanced glucose degradation and carbon entering the biomass. The RQ values were significantly modified by the glucose concentration, being highest at highest glucose rates, the presence of available nitrate and ammonium, and soil supporting its emergent microbial communities.
The RQ values during basal respiration were typically <1, suggesting that O-poor compounds were predominantly mineralised. Here, RQ values averaged 0.79, showing that differences between CO2 and O2 data were more than 20%. Such differences should be considered when evaluating soil quality and analysing organic matter transformation in ecosystems and comparing CO2 and O2 measurements. Low glucose concentration induced a moderate increase in RQ value that returned rapidly to those typical for basal respiration. The microbial communities in the beech forest soil were able to use the C source longer for metabolism indicated by more sustainable respiratory activity and RQ value, and also showed a higher microbial growth in comparison to the agricultural soil.
The author is grateful for the excellent technical assistance from Sabine Splitzer and Sandra Wulff, and for financial support by the German Research Foundation (DFG; Project no. MU 831/12-1) and the state of Bavaria and Schleswig-Holstein.