Main authors: Giulia Bongiorno, Else K. Bünemann, Chidinma Oguejiofor, Jennifer Meier, Gerrit Gort, Paul Mäder, Lijbert Brussaard, Ron de Goede
iSQAPERiS editor: Jane Brandt
Source document: Bongiorno G., Bünemann E., de Goede R., Brussaard L., Mäder P. (2018) Screening of novel soil quality indicators. iSQAPER Project Deliverable 3.4, 66 pp

 

Note: The following publication is based on the material contained in this section of iSQAPERiS

  • Giulia Bongiorno, Else K. Bünemann, Chidinma U. Oguejiofor, Jennifer Meier, Gerrit Gort, Rob Comans, Paul Mäder, Lijbert Brussaard, Ronde Goede. 2019. Sensitivity of labile carbon fractions to tillage and organic matter management and their potential as comprehensive soil quality indicators across pedoclimatic conditions in Europe. Ecological Indicators 99, 38-50 https://www.sciencedirect.com/science/article/pii/S1470160X18309415

In this article we assess the suitability of five different labile carbon fractions - dissolved organic carbon (DOC), hydrophilic DOC (Hy), permanganate oxidizable carbon (POXC), hot water extractable carbon (HWEC) and particulate organic matter carbon (POM-C) - as soil quality indicators across pedoclimatic zones.


Contents table
1. Introduction
2. Labile carbon measurements
3. Statistical analysis
4. Results
5. Discussion
6. Conclusions

1. Introduction

Soil organic carbon (SOC) is one of the most widely used soil quality indicators together with pH and available P and K (Bünemann et al., 2018). It affects various soil chemical, physical and biological properties and plays a primary role in multiple soil functions in agricultural soils, e.g. nutrient cycling, soil structural stability, and water retention (Reeves, 1997). Soil organic carbon also plays an important role in climate regulation with the potential of increasing carbon sequestration and counteracting the devastations created by extreme weather events (Lal, 2004). Despite its importance, SOC depletion is one of the main soil threats for agricultural soils. Agricultural measures which are aimed at increasing SOC stocks are therefore becoming a priority world-wide. For example, the »“4 per Mille” Initiative, established during the COP21 in Paris, aims at implementing soil managements practices, such as reduced tillage and the use of cover crops, which can effectively increase SOC stocks (Lal, 2016). Such practices have the potential to increase carbon stocks directly to the addition of organic materials but also indirectly through promoting soil structure by aggregate formation (Deb et al., 2015).

Soil organic carbon consists of a multiplicity of compounds, from simple to more complex molecules which can have different stability and functions in the soil, such as nutrient cycling, soil aggregation, and carbon storage (Deb et al., 2015). Since changes induced by soil practices are often difficult to detect in the recalcitrant part of SOC content, especially in the short term (Haynes, 2005), measuring rapidly changing SOC pools, such as labile carbon pools, might be more informative to assess soil quality (Gregorich et al., 1994; Wander, 2004).

Labile carbon is the SOC pool which is directly available for microbial metabolism and, hence, is considered to be the primary energy source for microorganisms (Chantigny, 2003; Haynes, 2005). Labile carbon is involved in different soil processes such as soil aggregate stabilization, and carbon and nitrogen cycling (Tisdall and Oades, 1982). Labile organic matter in soil mainly originates from the decomposition of organic materials, root exudates, and deceased microbial biomass (Bolan et al., 2011). Management practices with the addition of organic matter as fertilizer (Gattinger et al., 2012), and reduced tillage with a proven effect on soil organic matter in the top soil layer will likely increase labile organic carbon (Cooper et al., 2016). In addition, these practices have the potential to enhance soil aggregation which is one of the primary mechanisms through which labile C is sequestered in soil (Panettieri et al., 2015). These characteristics make labile organic carbon an interesting soil parameter bridging soil chemical, physical and biological characteristics. Moreover, labile organic carbon has been shown to be more sensitive to agricultural management than total organic carbon (Awale et al., 2017; Quanying et al., 2014).

Multiple labile carbon fractions have been characterized in the last thirty years. They are distinguished based on the nature of the fractionation methodology, which can be chemical, physical or biological (Haynes, 2005). Labile carbon fractions determined by chemical fractionation are extracted from the soil with different chemical compounds. Dissolved organic carbon (DOC) represents the organic carbon in the soil solution extracted with water and passing a mesh with a pore size of 0.45 µm (often called water-extractable organic carbon if extracted with water, WEOC, or water soluble carbon, WSC). Hy and DOC are small, soluble fractions of TOC, mainly comprised of root and microbial exudates and products of hydrolysis and leachates from organic matter. They can turn over very rapidly and be adsorbed to mineral surfaces (Leinemann et al., 2018; Lundquist et al., 1999). Labile carbon can also be extracted with hot water (hot water extractable carbon, HWEC), whose quantity is higher than DOC (Ghani et al., 2003). Permanganate oxidizable carbon (POXC), K2SO4 extractable C, and acid-hydrolysable C (H2SO4, HCl) are based on the use of other extractants then water. Although the quantities of POXC and HWEC were similar, they are most likely derived from different organic matter fractions. Both fractions probably comprise carbon derived from dissolved organic and microbial biomass carbon (in this study 1-2 % of TOC). POXC contains more recalcitrant compounds like lignin and complex polysaccharides, while HWEC largely (45-60%) comprises of carbohydrates and amides derived from soil microorganisms, enzymes, root exudates and lysates (Ghani et al., 2003; Haynes and Beare, 1997). HWEC is mainly present in the soil solution or loosely bound to soil minerals, and is prone to short-term seasonal variation (Leinweber et al., 1995). Physical (i.e. particle size or density) fractionation determines particulate organic matter carbon (POM-C) which consists mainly of partially decomposed organic residues (Haynes, 2005) and contains dissolved organic carbon, microbial biomass together with fresh plant residues and decomposing organic matter (Gregorich et al., 1994; Sequeira and Alley, 2011). Finally, microbial biomass carbon (MBC) and mineralizable C are also considered labile organic carbon fractions (called biological fractionation) (Haynes, 2005).

Many studies have used labile carbon to assess the impact of agricultural management and land use change on soil quality (Awale et al., 2017; Geraei et al., 2016; Ibrahim et al., 2013; Mirsky et al., 2008). However, it is unresolved which labile carbon fraction is the most sensitive to management and can be usefully related to soil functions, and as such be used as a sensitive soil quality indicator. Different fractions have been suggested as the most sensitive to soil management, and various methodologies and protocols have been applied, hampering comparisons between studies. Moreover, the linkage between labile carbon fractions and soil functions is often assumed and not established, and the generality of applying labile carbon fractions as soil quality indicators has never been assessed across different European pedoclimatic zones and agricultural management systems.

The general objective of this study was to facilitate the assessment of soil quality in agricultural systems by identifying a biochemical parameter which is sensitive to soil disturbance and linked with soil functions. The specific objective of our study was to assess the suitability of five different labile carbon fractions

  • dissolved organic carbon (DOC),
  • hydrophilic DOC (Hy),
  • permanganate oxidizable carbon (POXC),
  • hot water extractable carbon (HWEC) and
  • particulate organic matter carbon (POM-C)

as soil quality indicators across pedoclimatic zones. To do so, we tested the sensitivity of the labile carbon fractions to tillage and organic matter input in 10 European long-term field experiments. Moreover, we assessed the correlation of the different labile carbon fractions with soil physical, chemical and biological soil parameters linked to soil functions. We hypothesised that labile carbon concentrations would increase with the application of reduced tillage and high organic matter input. Moreover, we expected that labile carbon fractions would be positively correlated to chemical, physical and biological soil parameters related to nutrient cycling, soil organic carbon sequestration, and soil aggregation.

2. Labile carbon measurements

Dissolved organic carbon and Hydrophylics (DOC and Hy)

Twenty grams of field moist soil were used to extract dissolved organic carbon (DOC) as described in Van Agtmaal et al. (2017), and adapted as follows. Briefly, the samples were mixed with ultrapure water at a soil-to-solution ratio of 1:2 (dry wt/vol) in DOC-free polypropylene tubes, shaken for 1 hour, centrifuged for 20 minutes at 3750 rpm and subsequently for 10 minutes at 10000 rpm. The samples were then filtered at 0.45 µm with cellulose acetate Whatman® Puradisc membrane filters to obtain total DOC. Filters were pre-rinsed with ultrapure water and flushed with air to avoid any release of DOC during filtration. A fraction of the DOC obtained was subsequently acidified to pH 1 with 6 M HCl to extract the hydrophilic part of the DOC (Hy) using a simplified DOC fractionation scheme adapted from Van Zomeren and Comans (2007). During the fractionation the hydrophobic components of DOC present in solution (humic and fulvic acid, and hydrophobic neutrals) bind to an added insoluble polymeric absorbent (SupeliteTM DAX-8, Sigma-Aldrich). Only the hydrophilic part of the DOC remains in solution not binding to the resin, and can be quantified. Briefly, the DAX-8 resin was added to the acidified solutions to reach a ratio of 1:5 (wt/vol). The solution was then shaken horizontally for one hour at 180 rpm, centrifuged for 5 minutes at 3750 rpm, and the supernatant containing the hydrophilic part of DOC was collected. The total carbon (C) content of the DOC solution and the supernatant was determined on a TOC-5050A analyser (Shimadzu Corporation, Kyoto, Japan). DOC and Hy fractions were further analysed for specific ultraviolet absorbance (SUVA) to assess their aromaticity (Weishaar et al., 2003). To this end, 1.5 ml extracted DOC and Hy from each samples were analysed with a spectrophotometer (Genesys 10S UV-VIS, Thermo Fisher Scientific Inc., Waltham MA, USA) and ultrapure water was used as a blank. The aromaticity of the two fractions expressed by the SUVA (l g-1 cm-1) at 254 nm was calculated as described in Weishaar et al. (2003) and adapted by Amery et al. (2008):

D34 eq01

Where A254 is absorbance at 254 nm (dimensionless), b is the path length (cm) and DOC (or Hy) is the dissolved organic carbon concentration (mg l-1) of the solution.

Hot water extractable carbon (HWEC)

Hot water extractable carbon (HWEC) was determined according to the methodology of Ghani et al. (2003). Briefly, 4 g of soil were mixed with 30 ml of deionized water in a 50 ml polypropylene centrifuge tube. The tube was shaken horizontally for 30 minutes at 150 rpm and centrifuged for 20 minutes at 3500 rpm. The supernatant obtained at this stage (water-soluble carbon) was discarded. An additional 30 ml of deionized water was added to the sediments remaining in the tube and the tube was shaken for 10 seconds to suspend the soil in the water. Subsequently, the tubes were placed in an oven at 80oC for 16 hours. After this step, the tubes were shaken for 10 seconds in a vortex shaker and centrifuged for 20 minutes at 3500 rpm, and additionally for 10 minutes at 10000 rpm if necessary (to bring down the solid). The supernatants were filtered using 0.45 µm cellulose nitrate filter membrane and total carbon was determined on a TOC analyser.

Permanganate oxidizable carbon (POXC)

The permanganate oxidizable carbon (POXC) was extracted and analysed following the procedure of Weil et al. (2003) modified as follows. Briefly, 2.5 g of air dried soil aswas weighted into a polypropylene tube and 18 ml of demineralized water and 2 ml of 0.2 M K2MnO4 was added. The tube was shaken for 2 minutes at 120 rpm and thereafter left undisturbed on a lab bench for 8 minutes to continue the oxidation reaction. Subsequently, 0.5 ml of solution was taken fromthe tube and placed in another tube with 49.5 ml of demineralized water, allowing the reaction to stop. The absorbance of each sample at 550 nm (Abs) was determined using a GENESYS 10S UV-VIS Spectrophotometer. Permanganate oxidizable carbon was calculated according to Weil et al. (2003):

D34 eq02

where 0.02 mol L-1 is the concentration of the K2MnO4 solution, a is the intercept and b is the slope of the standard curve, 9 kg is the amount of carbon oxidized by 1 mol of MnO4 changing from Mn+7 to Mn+4, 0.02 L is the volume of the K2MnO4 reacting with the samples, and the Wt is the mass of soil in kg used for the reaction.

Carbon from particulate organic matter (POM-C)

The particulate organic matter was characterized as reported by Wyngaard et al. (2016) modified from Salas et al. (2003). Briefly, 10 g of dry soil samples aswas shaken for 15 hours with 30 ml of 1 M NaCl on a horizontal shaker. Subsequently the suspension was wet-sieved through a 53 µm sieve. The material on the top of the sieve was transferred to a crucible and dried overnight at 105˚C. The samples were weighted (M1) and placed in a furnace at 550˚C for 4 hours before weighing them again (M2). The POM was calculated by loss of ignition, i.e. as the weight loss during combustion at 550˚C in the muffle furnace. The POM-C was calculated dividing POM values for 1.724, assuming that the percentage of organic carbon in the POM was 58%. This conversion factor has been criticized, but for the purpose of this study we will not enter in the discussion (Pribyl, 2010).

Labile carbon stocks

Labile carbon and TOC stocks were calculated in the different layers taken into account in the study as:

D34 eq03

Where BD is the bulk density expressed in g cm3, the soil depth is the soil layer sampled, and the Labile C concentration is the concentration of labile carbon measured in g kg -1. For the LTEs where the two layers were sampled, C stock were calculated in the two layers separately and then added to obtain the value of the stocks in the 0-20 cm layer.

3. Statistical analysis

The effects of soil management on the labile carbon fractions (presented either in mg kg-1, percentage of TOC or as C stocks) per site across the 10 European long-term field experiments were assessed using linear mixed effects models. Mixed models were used to take into account the possible correlations introduced by the multi-site field experiments and to generalize the effect of the management practices across the different LTEs (Bradford et al., 2013; Lucas and Weil, 2012). The tillage and/or the soil organic matter addition and, if distinguished, the layer, their two-way and, if applicable, three-way interactions were used as fixed factors. Random effects of trials, blocks, main plots and subplots were introduced in the models to represent the experimental designs of the different trials. The effect of the soil pedoclimatic zone was not included in the fixed part of the model because we were interested in the management effects across the pedoclimatic zone. Three separate linear mixed effect models were applied to three subsets of the LTEs:

  1. Tillage model. The primary factor of interest in this analysis was tillage, followed by OM. To assess the effect of tillage and layer, the LTEs CH1, CH2, NL1, NL2, SL1 and HU4 were used and the analysis was performed on each of the soil layers (0-10 cm and 10-20 cm). For these trials, the stratification ratio for the labile carbon fractions in RT and CT was calculated and analysed in the linear mixed effect model according to Franzluebbers (2002) :
    D34 eq04
  2. OM model. The primary factor of interest in this analysis was the OM input, followed by tillage. For this analysis the LTEs analysed were NL1, NL2, SL1, CH3, HU1, PT1, ES4 and the 0-20 cm layer was used. In the LTEs in which the two layers were sampled separately, the value of the 0-20 cm layer was taken as the average of the 0-10 and the 10-20 layers.
  3. Stocks model. The factors of interest in this analysis were tillage and OM input. For the analysis of the labile carbon stocks (Mg ha-1) in the 0-20 layer, all ten trials were used. Additionally, the effect of agricultural management and the layer, if applicable, on the labile carbon concentrations was assessed in each long term field experiment with linear mixed effect models.

The effects of tillage and fertilization and their interaction on the labile carbon fractions were addressed by performing analysis of variance (function anova) on the fitted linear mixed effect model. For all the studied variables, the model assumptions of normality and homogeneity of variances of the residuals were checked both visually (plotting sample quantiles versus theoretical quantiles and residuals versus fitted values) and with the Shapiro-Wilk and Levene’s tests (Zuur, 2009). Variables whose residuals did not meet these assumptions were log-transformed or squared root-transformed and then used for analysis. If the transformation did not meet the criteria, the function weights were used in the linear mixed model effect formula to take into account the non-homogeneous variance structure introduced by the factors studied (Zuur, 2009). The function emmeans was used to estimate the marginal means and Tukey HSD post-hoc tests were used to assess significant differences between treatments when the ANOVA indicated a statistically significant effect. All test results were considered statistically significant at p≤0.05.

Spearman’s rank correlation was used to examine the relationships between labile carbon fractions and biological, physical and chemical soil quality parameters across the LTEs. Correlation analysis was done on log-transformed or squared root-transformed variables. The relationship between labile carbon fractions and soil parameters was validated using partial correlations, correcting for variation caused by the intrinsic differences of the LTEs (pedoclimatic zones). Partial correlations can, in fact, remove the effect of a random variable (in this case the LTE) which might control the observed relationship between two variables. When partial correlations are applied, the relationship between two variables is independent from the controlling variable. All statistical calculations were carried out using R version 3.3.2 (R Development Core Team, 2013). For the linear mixed effects model, the packages nlme (Pinheiro et al., 2018) and emmeans (Lenth et al., 2018) were used, while for the correlation analysis the packages car (Fox and Weisberg, 2011) and stats were used

4. Results

The concentrations of Hy, DOC, POXC, HWEC and POM-C differed widely among the LTEs, but, consistently across the 10 LTEs, Hy was the least abundant fraction per unit of soil or per unit of total organic C (0.004-0.05 % of TOC), followed by DOC (0.06-0.4% of TOC), POXC (1.45-4.32% of TOC), HWEC (0.98-6% of TOC) and finally POM-C (8-52% of TOC, and POM 14-89% of TOC) (Figure 2, Table S2). In comparison, microbial biomass carbon was intermediate between DOC and POXC and HWEC (0.12-2.84% of TOC). POXC and HWEC were similar in their concentration and total share in the TOC. Among the labile carbon fractions, the one with the lowest coefficient of variation was POXC (32%), followed by DOC (42%), Hy (43%), HWEC (51%) and POM-C (52%). The LTEs HU1 and PT1 had the lowest concentrations of labile carbon across the different fractions. Most labile carbon fractions had lower concentrations in the lower than in the upper layer, with the exception of DOC which was often higher in the lower layer. The labile carbon fractions expressed in % of TOC were more variable across the LTEs, and we did not find specific LTEs which had consistently higher or lower labile carbon fractions expressed as % of TOC.

D34 fig02
Figure 2

Effect of tillage on the labile carbon fractions

The labile carbon fractions differed in their sensitivity to tillage (Table 2). Concentrations of POXC, HWEC, and POM-C were higher under RT than CT, but only in the upper layer and in plots which received low OM input. Looking at the F statistics, POXC and POM-C (mg kg-1 soil) were the fractions most sensitive to tillage. In the lower layer, we observed a trend towards higher labile carbon concentrations in CT than RT. Moreover, RT had higher concentrations of Hy, POXC, HWEC and POM-C in the upper layer compared to the lower layer, while in CT the two layers had similar labile carbon concentrations. We found higher values of stratification ratio in RT than CT with low and high organic matter input for POXC, HWEC and POM-C (Table S4). For Hy and DOC the stratification ratio was higher in RT only with low organic matter input. Fields with high OM input tended to have higher concentrations of Hy, POXC, HWEC and POM than sites with low OM input, but the difference was not significant (Table 2). In SL1 and NL2, DOC tended to be higher in CT compared to RT, but the other labile carbon fractions had the opposite trend (Table S4). When expressed as % of TOC, only POXC and POM-C were higher in RT in the upper layer and only for plots which received low OM input (Table S5). We found that labile carbon expressed as % of TOC tended to be higher in the lower compared to the upper layer, especially in plots which received high organic matter input. The only fraction which was still higher in the upper layer if expressed as % of TOC was the HWEC.

Aromaticity of the DOC and Hy as measured by SUVA254 were not affected by tillage treatment across the sites (Table S6).

Table 2. Effects of tillage (CT vs. RT) and organic matter input (LOW vs. HIGH) on the labile carbon fractions for the tillage trials as analysed with mixed linear effect model (number of observations= 120). In the upper part of the table the estimated means and 95% confidence intervals (in parentheses) of Hy, DOC (mg kg-1 soil), POXC, HWEC, POM-C and TOC (g kg-1 soil) under tillage and OM management are reported. Different letters following mean and se have to be read per columns and per layer; they show treatments which are significantly different (p< 0.05) according to Tukey post-hoc test. In the lower part of the table F statistics and p-values for the main factors and their interactions are reported.

 D34b tab02

Effect of OM addition on the labile carbon fractions

All labile carbon fractions were significantly higher in high OM compared to low OM input trials (Table 3). In the analyses, the type of tillage applied to the plots was also taken into account. POXC, HWEC and POM-C (mg kg-1) were significantly increased in the RT compared to the CT plots. POXC, Hy POM-C (mg kg-1) were the more sensitive labile carbon fractions (taking into account the F statistics). When labile carbon fractions were expressed as % of total organic carbon, only Hy, POXC and POM-C were significantly higher in the high OM compared to low OM input trials (Table S7). In addition, the positive effect exerted by the high organic matter input on POXC was stronger in trials with CT. Aromaticity of DOC and Hy as measured by SUVA254 were not affected by the organic matter input type across all the sites (Table S8).

Table 3. Effects of organic matter input (LOW vs. HIGH) and tillage (CT vs. RT) on the labile carbon fractions for the OM input trials as analysed with mixed linear effect model (number of observations= 119). In the upper part of the table the estimated means and 95% confidence intervals (in parentheses) of Hy, DOC (mg kg-1 soil), POXC, HWEC, POM-C and TOC (g kg-1 soil) under OM and tillage management are reported. Different letters following means and se have to be read per column; they show treatments which are significantly different (p< 0.05) according to the Tukey post-hoc test. In the lower part of the table F statistics and p-values for the main factors and their interactions are reported.

D34b tab03

Effect of tillage and OM input on the labile carbon stocks across the 10 LTEs

Reduced tillage and high OM input both significantly increased labile carbon stocks expressed in Mg ha-1, i.e. stocks of all the fractions (Table 4). POXC and POM-C were affected most by the two management factors (higher F statistics). The TOC stock was less sensitive than the stocks of labile C fractions, being affected neither by organic matter addition nor by tillage.  

Table 4. Effect of organic matter input (LOW vs. HIGH) and tillage (CT vs. RT) on the labile carbon stocks expressed in Mg C ha-1 for the OM input trials as analysed with mixed linear effect model (number of observations= 101). In the upper part of the table the estimated means and 95% confidence intervals (in parenthesis) of the labile carbon fractions and TOC in organic matter and tillage management are reported. Different letters following mean and se have to be read per column; they show treatments which are significantly different (p< 0.05) according to the Tukey post-hoc test. In the lower part of the table F statistics and p-values for the main factors and their interactions are reported.

D34b tab04

Correlation between labile organic carbon fractions and soil parameters

We tested the bivariate relationships between the labile carbon fractions and soil chemical, physical and biological indicators across both soil layers where applicable. In addition to bivariate correlations, we validated the obtained relationships by carrying out partial correlations where we corrected for the variation caused by the LTE (Table 5). POXC was the labile C fraction that was most significantly (p-values), and strongly (Spearman’s correlation coefficients, ρ), correlated with the soil chemical, physical and biological indicators related to nutrient cycling, soil structure and biodiversity both in the bivariate (Table S9) and the partial (Table 5) correlations. Moreover, POXC proved to be highly positively correlated (p< 0.0001) with Hy-DOC (ρ=0.59), DOC (ρ= 0.41), HWEC (ρ= 0.60) and POMC (ρ= 0.70) (Table 5 residuals and S10 original data). The other carbon fractions were correlated with each other but not so strongly, with the only exception of strong positive correlations between Hy-DOC and (in addition to POXC) DOC, HWEC, and POMC (p< 0.0001, ρ= 0.41; p< 0.001, ρ= 0.41; p< 0.0001, ρ= 0.50, respectively) (Table 6).  

Table 5. Partial correlation coefficients (ρ) between the labile organic carbon fractions expressed in mg kg-1 soil (Hy-DOC, DOC, POXC, HWEC and POMC) and % (TOC) and various soil chemical, physical and biological indicators used as dependent variable, corrected for the long term field experiments (LTEs). The number of samples used in the analyses was 167, but 101 for earthworm number, and earthworms biomass.

D34b tab05

Table 6. Partial correlation coefficients (ρ) between the labile organic carbon fractions expressed in mg kg-1 (Hy-DOC, DOC, POXC, HWEC and POMC) and L g C-1 m-1 (Hy SUVA and DOC SUVA) used as dependent variable, corrected for the long term field experiments (LTEs). In addition, for comparison with the labile carbon fractions, also the correlation with TOC expressed in mg kg-1 has been reported. The number of samples used in the analyses was 167.

D34b tab06

5. Discussion

The ranges of labile organic C fractions measured in this study were in accordance with those reported previously (Benbi et al., 2015; Lucas and Weil, 2012; Margenot et al., 2017). Hy accounted for the smallest part of TOC, followed by DOC, POXC, HWEC and POM-C (Figure 2).

Effect of tillage on the labile carbon fractions

In the analysis across the 6 tillage trials, POXC, HWEC and POM-C (mg kg-1 soil) were higher in RT compared to CT in the upper layer, but only significantly so in plots which received low organic matter input (Table 1). The addition of organic material might favour the formation of macro- and micro-aggregates which can increase labile carbon retention in soil and counteract the negative effect of CT. Moreover, CT could promote the release of labile organic carbon from the organic matter added to the soil. Mando et al. (2005) observed that deep tillage decreased POM-C, but less so when manure was applied to the fields. The authors attributed this to the higher capacity of the soil to preserve soil organic carbon in association with clay. If soils are rich in carbon (or saturated) their potential to protect and accumulate organic matter may be limited under agricultural practices which are supposed to increase soil C levels (Schmidt et al., 2015; Six et al., 2002).

Several studies reported that RT increases the concentration of soil labile carbon compared to CT (Aziz et al., 2013; Liu et al., 2014; Neogi et al., 2014). Tillage disrupts macro- and micro-aggregates, increases soil temperature, aeration and releases soil organic matter which is protected in these physical structures (Six et al., 1999). Soil organic matter can subsequently be more available to soil organisms, increasing CO2 emission. This phenomenon is fostered by the incorporation of residues, mineral fertilizers and organic amendments in deeper soil layers. In reduced tillage, on the other hand, the labile carbon protected in aggregates can accumulate in the soil (Jastrow et al., 2006). Under these conditions, microbial community abundance and activity can be favoured, increasing the production of enzymes which can increase soil labile C fractions (Melero et al., 2011).

Our study corroborates previous findings which also detected HWEC, POXC and POM-C as the most sensitive labile C fractions to tillage (Chen et al., 2009; Ćirić et al., 2016), in particular POXC and POM-C (Culman et al., 2012; Plaza-Bonilla et al., 2014; Prasad et al., 2016).

Dissolved organic carbon and Hy were less sensitive to tillage. Dissolved organic carbon is very much dependent on environmental conditions (i.e. temperature and precipitation), and short-term management (Federici et al., 2017; Mouloubou et al., 2016; Soon et al., 2007). Moreover, in spring, which coincided with our sampling time, the level of DOC is the lowest throughout the year (Haynes, 2005; Schiedung et al., 2017).

Effect of organic matter input on the labile carbon fractions

High OM addition increased the concentration of all labile carbon fractions compared to low OM addition. This agricultural practice had greater impact on the labile carbon fractions than tillage indicating the important role of organic matter addition in increasing C in soil (Table 3). Permanganate oxidizable C, Hy and POM-C were the fractions more sensitive to OM additions (Ibrahim et al., 2013; Mirsky et al., 2008; Tatzber et al., 2015). Previous studies found that organic input increases the concentration of soil labile carbon (Benbi et al., 2015; Li et al., 2018; Pezzolla et al., 2015; Tatzber et al., 2015). Organic matter input can increase labile carbon, favouring soil biota biomass and activity directly through the addition of organic substrates of different complexity and nutrients, and indirectly through the creation of suitable environments for living organisms. Moreover, organic matter addition can introduce external microbial populations which also can contribute to an increase of the labile organic carbon pools (Bastida et al., 2008).

Some studies did not find effects of tillage and fertilization on the labile C fractions (Ladoni et al., 2015; Margenot et al., 2017; Sequeira and Alley, 2011). This can be due to the soil properties, the non-homogeneous distribution of plant and microbial residues, organic matter input type and quantity, environmental conditions and time of sampling. The soil type, for example, can influence the extent to which agricultural management can affect soil organic carbon. In soils with light texture, organic matter additions can have a higher beneficial effect on TOC and labile carbon than conservation tillage (Chivenge et al., 2007). Our approach, based on the selection of LTEs from different pedoclimatic zones and contrasting soil types, permitted us to identify overall trends correcting for these differences in pedoclimatic zones. Even after such corrections, we found tillage and organic matter additions to have an effect on the labile carbon fractions, and in particular on POXC and POMC.

Labile carbon as soil quality indicator

All the labile carbon fractions had a positive relationship between them (Table 6). However, the different strengths of the correlation between labile carbon fractions suggest that they don’t represent the same pools. All the labile carbon fractions had a positive relationship with TOC (Table 5) suggesting that TOC is their main determinant in soil (Geraei et al., 2016; Yu et al., 2017). Therefore, dynamics in labile C fractions could be used as an indication of TOC’s dynamic in soils due to agricultural management. Labile carbon is, in fact, an essential starting point for the formation of more stable soil organic matter (Cotrufo et al., 2013). However, the different strengths of the relationship between labile carbon fractions and TOC suggest that these fractions quantify distinct parts of the TOC (Table 6). POXC and POM-C were the two labile carbon fractions that showed the strongest relationship of all labile carbon fractions with TOC. In addition, POXC was a sensitive labile carbon fraction measured with less variation compared to the other fractions (Table S2) POXC was the labile carbon fraction most strongly related to soil chemical, physical and biological quality indicators (Table 5). The correlations between POXC, TOC and MBC have been attributed to specific characteristics of the extraction methods used to determine the three fractions (Geraei et al., 2016). The oxidation of POXC mimics microbial decomposition of organic matter, which is confirmed by its often positive correlation with basal respiration, substrate-induced respiration, microbial biomass and soluble carbohydrates (Weil et al., 2003). POXC was also correlated very strongly with both the labile carbon fractions that are more labile than POXC (i.e. Hy and DOC) and the labile fractions that are equally to (HWEC) or less labile (POM-C) than POXC. This suggests that POXC is the most representative labile organic carbon indicator.

The positive correlation between POXC and HWC, WSA and CEC, which are parameters known to be influenced by more complex organic matter (Wander, 2004), can be explained by the fact that the oxidation during the POXC reaction targets labile but also affects more recalcitrant forms of SOM. Specialized microorganisms can make use of more complex compounds (Lehmann and Kleber, 2015), which could explain the relationship between POXC and microbial biomass and activity even if permanganate reacts with more complex compounds also, as recently confirmed by Romero et al. (2018). Hence, POXC strongly relates to TOC, but also a variety of other soil quality parameters underlining its role as a multifunctional soil quality indicator. Moreover, POXC can be measured relatively cheaply and fast (Table S11). The different strengths of the correlations between labile carbon fractions and other soil quality indicators, including TOC, suggest that these fractions quantify distinct parts of the TOC with different functional characteristics.

Currently, little is known about the chemical composition and the seasonal dynamics of POXC. However, there is evidence for the sensitivity of POXC to other types of soil management beside tillage and organic matter input such as the use of cover crops, but this should be validated with further studies (Culman et al., 2012). POXC was found to be linked with various soil quality indicators related to multiple soil functions, which is a very important characteristic for effective and informative soil quality indicators. As the quantitative relationships between currently used indicators and soil functions are generally under-investigated, establishing those relationships is of high priority and future studies should particularly address these quantitative linkages.

6. Conclusions

The labile organic carbon fractions investigated in 10 LTE fields covering a range of pedoclimatic zones appeared sensitive to soil management, showing in general increased values for reduced tillage and high organic matter input systems. Our results suggest that the different labile carbon fractions represent different soil organic carbon pools. POXC and POM-C represent organic carbon pools which appeared to be more sensitive to agricultural management than the other labile carbon fractions. This makes them more suitable as soil quality indicators than the highly labile carbon fractions such as DOC, Hy and the slowly changing TOC. Also, concentrations of POXC and POM-C are an order of magnitude higher than Hy and DOC, which strongly facilitates their measurement.

Our results suggest that POXC is to be preferred over POM-C as soil quality indicator. POXC appears be the carbon fraction that is the most informative about total soil organic matter, nutrients, soil structure, and microbial pools and activity. The response of POXC is not too fast to show only short-term management effects, but fast enough to show long-term management effects at the short term. Moreover, POXC is easily measured at low cost, which makes its use feasible in practice.

 


Note: For full references to papers quoted in this article see

» References

 

Go To Top