Performance of base hydrolysis methods in extracting ...

Accepted Manuscript Performance of base hydrolysis methods in extracting bound lipids from plant material, soils, and sediments Gerrit Angst, Tomá š Cajthaml, Š árka Angst, Kevin E. Mueller, Ingrid KögelKnabner, Sebastian Beggel, Stefanie Kriegs, Carsten W. Mueller PII: DOI: Reference:

S0146-6380(17)30266-8 http://dx.doi.org/10.1016/j.orggeochem.2017.08.004 OG 3602

To appear in:

Organic Geochemistry

Received Date: Revised Date: Accepted Date:

27 April 2017 8 August 2017 10 August 2017

Please cite this article as: Angst, G., Cajthaml, T., Angst, S., Mueller, K.E., Kögel-Knabner, I., Beggel, S., Kriegs, S., Mueller, C.W., Performance of base hydrolysis methods in extracting bound lipids from plant material, soils, and sediments, Organic Geochemistry (2017), doi: http://dx.doi.org/10.1016/j.orggeochem.2017.08.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Performance of base hydrolysis methods in extracting bound lipids from plant material, soils, and sediments

Gerrit Angst a,b,*, Tomáš Cajthaml c, Šárka Angst b, Kevin E. Mueller d, Ingrid Kögel-Knabner a,e , Sebastian Beggel f, Stefanie Kriegs a, Carsten W. Mueller a

a

Chair of Soil Science, Technical University of Munich, Emil-Ramann-Straße 2, D-85354

Freising, Germany b

Institute of Soil Biology & SoWa RI, Biology Centre, Czech Academy of Sciences, Na Sádkách

7, CZ 37005 České Budějovice, Czech Republic c

Institute for Environmental Studies, Faculty of Science, Charles University,

Benátská 2, CZ 12800, Praha, Czech Republic d

Biological, Geological and Environmental Sciences, Cleveland State University, Cleveland,

Ohio 44115, USA e Institute

for Advanced Study, Technical University of Munich, Lichtenbergstraße 2a, D-85748

Garching, Germany f

Aquatic Systems Biology Unit, Department of Ecology and Ecosystem Management, Technical

University of Munich, Muehlenweg 22, D-85354 Freising, Germany

*corresponding author: phone: +420 387 775 785; e-mail: [email protected]

1

ABSTRACT Base hydrolysis methods are widely used to extract bound lipid biomarkers such as cutin and suberin from plant material, soils, and sediments. However, non-uniform treatments among studies hamper their comparability and information on differences between commonly used saponification methods does not exist. The aim of the present study was to compare frequently used base hydrolysis methods and evaluate their performance in extracting bound lipids from common European forest and agricultural plant species (Fagus sylvatica L. and Zea mays L.), and corresponding soils and sediments. We compared extraction in Teflon-lined bombs, ultrasound assisted extraction, and extraction with reflux in quintuplicate for each leaf, soil, and sediment material. The reflux method gave consistently higher yields of bound lipids as compared to the other two methods, while requiring the same amount of time and solvent and featuring lower standard errors. Some lipids, including common building blocks of cutin and, to a lesser extent, suberin, were more efficiently extracted by the reflux method than others, indicating that source assignment of organic matter in soils and sediments via biomarkers might be sensitive to the employed method. Although the other extraction methods may perform similarly under different conditions, such as other sample matrices, extraction conditions of the reflux method may potentially be more easily controlled across replicates and laboratories as compared to the bomb or ultrasound methods. Thus, the use of reflux methods ensures meaningful yields of hydrolysable lipids while providing high reproducibility.

Keywords: Cutin, Suberin, Soil organic matter, Biomarkers, European beech, Fagus sylvatica L., Maize, Zea mays L.

1.

Introduction

2

The chemical nature of soil organic matter (SOM) and the pathways of SOM formation are current key research issues (e.g., Cotrufo et al., 2013, 2015; Castellano et al., 2015) aiming at reliably parametrizing carbon (C) models and managing C sequestration efficiently (Jandl et al., 2007; Campbell and Paustian, 2015). To evaluate the fate of organic matter entering the soil, it is important to trace individual compounds with the aid of suitable tools, such as isotopic or biomarker techniques. In this regard, the hydrolysable lipid biopolymers cutin, present in the cuticle of angiosperms and gymnosperms (Kolattukudy, 1980), and suberin, mainly occurring in the periderm of roots (Kolattukudy, 1980), have increasingly been used to disentangle the amount and turnover of C from above- and below-ground sources in SOM (Crow et al., 2009; Feng et al., 2010; Carrington et al., 2012; Spielvogel et al., 2014; Angst et al., 2016b), or to evaluate the extent of pedogenic processes, such as bioturbation (Nierop and Verstraten, 2004). Although hydrolysable lipids have widely been applied in soil science and biogeochemistry, there is no uniform method for their extraction from plant material, soils or sediments. Chemical methods span from acid/base hydrolysis (Holloway et al., 1981; Ray et al., 1995; Graça et al., 2002; Feng and Simpson, 2008) to transesterification with BF3/MeOH (Kolattukudy et al., 1975; Matzke and Riederer, 1991; Kögel-Knabner et al., 1992; Riederer et al., 1993; Fiorentino et al., 2006), or cupric oxide oxidation (Filley et al., 2008; Crow et al., 2009; Carrington et al., 2012; Angst et al., 2017). More recently, Mendez-Millan et al. (2010a) identified base hydrolysis as being the most efficient and precise method for the extraction of cutin from maize leaves and soil. Within the field of base hydrolysis, however, treatments differ in their underlying extraction principles, such as ultrasonication vs stirring or pressurized extraction vs extraction at atmospheric pressure. As a consequence, different methods yield results that may deviate from each other thus hampering the comparability of studies and leading to varying conclusions about the contribution of plant-derived compounds to soil or turnover of root- and shoot-derived SOM

3

for example. Three types of saponification methods are commonly used: reflux methods at moderate temperatures and atmospheric pressure (Nierop et al., 2003; Otto and Simpson, 2006; Mendez-Millan et al., 2010b), ultrasonication assisted extractions (Rumpel et al., 2005; Mueller et al., 2012), and the extraction within Teflon-lined ‘bombs’ at higher temperatures and pressures (Feng et al., 2010; Spielvogel et al., 2014; Angst et al., 2016a). The aim of the present study was to compare the performance of these saponification methods in extracting lipids from leaves, soils, and sediments in terms of yield and reproducibility. We particularly focused on the acid lipid fraction because it has the highest diagnostic value for the identification of bound lipid biomarkers, such as cutin and suberin (Spielvogel et al., 2014). For our analyses, we chose a forest site (European beech; Fagus sylvatica L.) and an agricultural site (maize; Zea mays L.), vegetated with species highly abundant in central Europe (Geßler et al., 2007).

2.

Materials and methods

2.1

Study sites and sampling Soils were sampled in a pure, mature European beech (Fagus sylvatica L.) stand (48° 55'

17" N 14° 56' 19" E) near the village of Zajíc, Czech Republic, and a maize (Zea mays L.) field (48° 23' 42" N, 11° 48' 03" E) near the city of Freising, Germany (Table 1) in February 2016. Sites without understory vegetation were chosen to ensure direct comparisons between the vegetation and the soils/sediments. Samples were collected from the Ah horizons of a Luvic Stagnosol (beech site) and a Calcic Gleysol (maize site; IUSS Working Group WRB, 2014), respectively, using 100 cm³ steel rings. Six samples were taken in a distance of 2 m to each other to cover possible spatial heterogeneities. Because measured pH values and C contents of the samples at each site were very similar (Table 1), we chose to combine them into a composite sample. Sediments were

4

collected from the top layers covering the bed of creeks flowing through (maize site) or having its source in the respective study area (beech site) using a box sampler (Pander et al., 2015). Previously collected and freeze-dried green beech and maize leaves from the study sites were used for analysis of the plant source material.

2.2

Laboratory analyzes Soils and sediments were air-dried and sieved to ≤ 2 mm, the leaves were finely ground

to a powder using an ultracentrifuge mill. Carbon contents were measured in duplicate on an elemental analyzer (EuroVector, Milan, Italy) via dry combustion. Aliquots of soils and sediments were finely ground prior to analysis using a ball mill. Inorganic C contents of soils and sediments were determined by the quantification of evolved CO2 after fumigation with HCl (‘Scheibler’ method; DIN EN ISO 10693). The total C was corrected for inorganic C so that all C data presented in this study correspond to organic C contents. All samples were subjected to a sequential lipid extraction procedure to release hydrolysable lipids. In a first step, solvent-extractable lipids were removed from the samples by accelerated solvent extraction (ASE200, Dionex GmbH, Idstein, Germany). Aliquots (30 g of soil and sediment; 5 g of leaves) were extracted with a mixture of dichloromethane DCM/methanol (9:1, v:v) at a pressure of 17 × 106 Pa and a temperature of 75 °C (Wiesenberg et al., 2004; Jansen et al., 2006; Angst et al., 2016a). The heating phase and the static extraction time were set to 5 min. All samples were extracted twice under the same conditions to assure a thorough removal of solvent-extractable lipids (Wiesenberg et al., 2004; Angst et al., 2016a). In a second step, the pre-extracted samples (5 g of soils and sediments and 300 mg of plant material; n = 5 for each method and material) were hydrolyzed with methanolic KOH (1 5

M) using three different methods to release hydrolysable lipids. We compared extraction in Teflon-lined bombs (M-bomb), ultrasound assisted extraction (M-ultrasound), and extraction with reflux (M-reflux). The M-bomb method was performed according to Spielvogel et al. (2014), Pisani et al. (2015), and Angst et al. (2016a). Briefly, the samples were weighed into 50 ml teflon-lined bombs and extracted with 20 ml of 1M methanolic KOH for 3 h at 100 °C. The M-ultrasound method was performed by modifying a procedure by Rumpel et al. (2005) and Mueller et al. (2012). The samples were weighed into 100 ml centrifuge glass tubes that were closed with Teflon-lined lids and saturated with 20 ml of 1M methanolic KOH solution. The centrifuge tubes were submerged in an ultrasonication bath to the depth of the KOH solution and extracted at 80 °C for 3 h. The samples were additionally ultrasonicated for the first 15 min of each hour to avoid excessive heating. The M-reflux method was similar to the procedure used by Nierop et al. (2003), Otto and Simpson (2006), or Mendez-Millan et al. (2011). The samples were weighed into 100 ml round bottom flasks and saturated with 20 ml of 1M methanolic KOH. The samples were then refluxed for three hours using condenser tubes that were cooled with water flowing through them. The temperature was adjusted to 80 °C using patio heaters and gentle boiling of the solution was monitored using boiling chips. After cooling, the samples were centrifuged and the supernatant KOH solution transferred to glass vials. The solid soil residues were washed twice with a DCM/methanol (1:1, v:v) solution, which was subsequently combined with the KOH extracts. The combined extracts were evaporated to total dryness and re-dissolved in deionized water and DCM. The acid fractions were separated from the re-dissolved extracts by liquid-liquid extraction according to Angst et al. (2016a). The glass vials containing deionized water and DCM were vortexed and the DCM phase was removed and transferred to separate vials. Fresh DCM was added to the glass 6

vials containing the deionized water phase and the previous step was repeated twice. The deionized water phase was then adjusted to pH 1 using concentrated hydrochloric acid and the acid lipid fraction was extracted 3× with DCM by repeated addition of DCM, vortexing, and transferring the DCM phase to separate vials. Note that the adjustment to pH 1 may have led to an underestimation of n-carboxylic acids in the extracts (Naafs and Van Bergen, 2002); but because the liquid-liquid extraction was identical for all employed methods, our approach did not affect the differences among the method of base hydrolysis. To remove any residual water, the DCM extracts were filtered over fritted columns containing Na2SO4. The filtered extracts were dried under a gentle stream of nitrogen and stored in the freezer until further processing. For GC–MS analysis, the acid fractions were sequentially derivatized using two procedures. For methylation, the dry samples were dissolved in 3 ml of DCM and aliquots of 300 μl were transferred to 2 ml glass vials. The samples were evaporated under a stream of nitrogen and re-dissolved in 300 μl of ethyl acetate containing anthracene (10 μg/ml) as internal standard. Subsequently, 10 μl of methanol and 500 μl of diazomethane in diethyl ether were added. The vials were tightly closed and incubated at 5 °C for 1 h. The samples were then dried under nitrogen, re-dissolved in 600 μl of ethyl acetate and analyzed for methyl esters of carboxylic acids using GC–MS. The etheric solution of diazomethane was prepared according to a protocol published previously (Křesinová et al., 2014). The samples were subsequently dried again and silylated using BSTFA containing 1% of TMCS (Sigma). The samples were incubated at 70 °C for 30 min, dried and dissolved in ethyl acetate (600 μl) prior to the second GC–MS analysis that was focused on hydroxylated carboxylic acids. We chose a two-step GC approach because a direct double derivatization resulted in a higher baseline and some overlapping peaks, which complicated the accurate identification of analytes.

7

The GC–MS analyses were performed on a Varian 450-GC (USA) instrument equipped with a Combi-Pal autosampler (CTC Analytics, Sweden) and a Varian 240 MS ion trap detector employing an electron-impact ion source. The samples (1 µl) were injected onto the GC column (DB-1MS: 30 m length × 0.25 mm I.D., 0.25 µm film thickness; Agilent, USA) using a split/splitless mode with an injector temperature of 290 °C. The GC oven temperature program started at 60 °C (held for 1 min in splitless mode), then the splitter was opened and the temperature was raised to 160 °C (held for 2 min) at 40 °C/min, and was then raised to 290 °C at 2.5 °C/min held isothermally for 15 min. Helium (99.9%) was used as carrier gas with a constant flow rate of 1.0 ml/min. The ion trap and transfer line temperatures were 240 °C and 290 °C, respectively. Data (TIC mode) were collected from 6 to 72 min in the mass range of 50 to 600 Da. Monomers were identified according to their fragmentation pattern, previously published spectra, and retention times of external standard compounds, aided by comparison with a mass spectral library (NIST 08). The analytes were quantified using calibration curves derived from different concentrations of external standards provided by Sigma-Aldrich (Czech Republic), Larodan (Sweden) and TCI (China), consisting of homologues of n-carboxylic, α,ω-dicarboxylic, and ω-hydroxycarboxylic acids. Lipid yields were corrected for small shifts in the response of the internal standard in the sample relative to that in the external standard solution during measurement, and are reported per g C (Figs. 1 and 2; Supplementary Tables S3 and S4) and per g dry weight (DW; Supplementary Tables S1 and S2).

2.3

Statistics and calculations Statistical means of the five replicates for each extracted material, standard errors (SE),

and coefficients of variation (CV) were calculated using Microsoft Excel 2013 for Windows

8

(Microsoft, Redmond, WA, USA). All other statistics were conducted using the R 3.0.3 Software for Windows (R Core Team, 2013). The data were screened for normality and homoscedasticity using the Shapiro-Wilk and Bartlett tests, respectively. Depending on the tests’ outcomes, significant differences between the methods (M-bomb, M-ultrasound, and M-reflux) for each extracted lipid were either assessed by one-way analysis of variance (ANOVA) followed by Tukey’s honestly significant difference (HSD) test, or by the Kruskal-Wallis test followed by Dunn’s post-hoc test. Differences were regarded as significant at a level of p < 0.05. Differences at a level of 0.05 > p < 0.1 are reported as marginally significant.

3. 3.1

Results and discussion Total lipid yields and extracted monomers The total lipid yields for beech varied from 4.1 to 8.5 mg/g C for leaves, 2.9 to 6.4 mg/g

C for soil and 2.7 to 8.5 mg/g C for sediment, mostly increasing in the order M-bomb < Multrasound < M-reflux (Fig. 1). Total lipid yields extracted from maize leaves, soil, and sediment ranged from 0.7 to 2.0 mg/g C, 1.5 to 3.0 mg/g C, and 2.6 to 6.5 mg/g C, respectively. Despite these rather large variations, the total lipid concentrations were generally within the range of values reported in the literature for similar methods and sample types (Mendez-Millan et al., 2011, 2010a; Mueller et al., 2012; Angst et al., 2016a, 2016b). Deviations in lipid yield among studies likely occur because lipid composition depends on variations in environmental conditions, genetics, and traits of the same plant species at different study sites, such as the lifespan of leaves and roots (Mueller et al., 2012; Freimuth et al., 2017). Also hydrolyzing time may explain variable lipid yields across studies. For example, the lipid yields for maize leaves extracted by Mendez-Millan et al. (2010a) were several magnitudes higher (~7 mg/g DW) than

9

in the current study (~1 mg/g DW; Supplementary Table S2). The authors used a reflux time of 18 h as compared to a reflux time of 3 h in the present study. For soil, however, the lipid yields in the present study were higher than those reported by Mendez-Millan et al. (2010a) (~167 cf. 60 µg/g DW). Apart from site-dependent variations in the lipid composition of maize leaves, these results may reflect different stabilization of the monomers in soil and/or a matrix dependent effect of hydrolysis time. However, in accordance with the majority of studies using base hydrolysis, employing a hydrolysis time of 3 h or less (Nierop et al., 2003; Rumpel et al., 2005; Otto and Simpson, 2006; Feng and Simpson, 2008; Mueller et al., 2012; Spielvogel et al., 2014; Angst et al., 2016a), we adapted an extraction time of 3 h. All extraction methods released a series of n-carboxylic, α,ω-dicarboxylic, ωhydroxycarboxylic, and mid-chain substituted hydroxycarboxylic acids with different chain lengths (Figs. 1 and 2; an overview of concentrations normalized to DW and C are shown in Supplementary Tables S1 to S4). The n-carboxylic acids are present in microbes and various plant tissues, with different chain lengths characteristic of plant waxes (≥ C20) and microbes (mainly mono- and di-unsaturated C16 or iso-carboxylic C16/C18 acids) (Eglinton and Hamilton, 1967; Lichtfouse et al., 1995; Otto et al., 2005; Wiesenberg et al., 2010). The ω-hydroxy, midchain substituted hydroxy, and α,ω-dicarboxylic acids ≥ C16 are monomers of the plant biopolymers cutin or suberin. Although typical cutin and suberin monomers have previously been described (x,ω-dihydroxyhexadecanoic acids (x = 8, 9, or 10; x,ω-diC16) and 9,10,ωtrihydroxyoctadecanoic acid (9,10, ω-C18) for cutin and ω-hydroxycarboxylic acids ≥ C20 and dicarboxylic acids C16 and C18 for suberin; Kögel-Knabner, 2002; Kolattukudy, 1980), many studies perform a site-dependent assignment of monomers that sometimes substantially deviate from the ‘standard’ set of cutin/suberin monomers (Dignac and Rumpel, 2006; Hamer et al.,

10

2012; Spielvogel et al., 2014); this is because the composition of plant waxes, cutin, and suberin varies across plant species and environmental conditions (Diefendorf et al., 2011; Mueller et al., 2012). Because we focused on the performance of different methods in extracting lipids from various sample matrices and because leaves were the only plant organ sampled, we cannot unambiguously assign monomers in our soil/sediment samples to cutin or suberin, but suggest the potential origin of a monomer when appropriate.

3.2

Differences among methods Despite lower temperatures than M-bomb and lower pressures than both M-bomb and M-

ultrasound, M-reflux released consistently higher yields of the total sum of lipids regardless of extracted material (significant for maize leaves, soil, and sediment, and marginally significant for beech leaves and soil; Figs. 1a, 1b, and 2). The only exception to this pattern was the sediment from the beech site, where M-bomb provided similar yields albeit featuring a substantially higher standard error as compared to M-reflux (38% cf. 15%). The comparably high standard errors for M-bomb may result from the extraction conditions within the bombs that can likely not as easily be controlled across replicates as compared to those of M-reflux (see below). This may particularly be true for complex sample matrices, such as the sediment matrix in the present study. Several types of lipids showed significantly higher yields when treated with M-reflux as compared to the other methods (0.5–4.5 times higher); this was evident for the ubiquitously occurring n-C16 and n-C18 acids, the n-carboxylic acids ≥ C20 (dominant in plant waxes) and several hydroxycarboxylic acids ascribed to cutin or suberin of beech or maize (Dignac and Rumpel, 2006; Mendez-Millan et al., 2010a; Angst et al., 2016a), including x,ω-diC16

11

hydroxycarboxylic acids, 9,10,ω-C18 hydroxycarboxylic acid, ω-hydroxycarboxylic acids ≥ C20, and ω-dicarboxylic acids ≥ C16. The difference between methods were largest for mid-chain substituted hydroxycarboxylic acids, followed by n-carboxylic acids, ω-hydroxycarboxylic acids, and α,ω-dicarboxylic acids, although the differences were not consistent across the sample types. For example, concentrations of the x,ω-diC16 acids, which have exclusively been ascribed to cutin of beech, maize, and other species (Hauff et al., 2010; Mendez-Millan et al., 2011; Pisani et al., 2015; Angst et al., 2016a, 2016b), were ~2 to 79 times higher using M-reflux for beech soils and leaves, soils and sediments of maize (Figs. 1a, 1b, 2b and 2c). The yields of C16DA and C18DA, often assigned to maize and sometimes to beech suberin (Mendez-Millan et al., 2011; Angst et al., 2016a), were ~2 to 2.5 times higher using M-reflux for the maize samples. These results indicate that the hydrolysis methods have different extraction efficiencies for some individual lipid monomers, so the proportion of certain lipids relative to each other, including those derived from cutin and suberin, may vary depending on the method of hydrolysis. The explanation of why M-reflux preferentially extracted bound lipids as compared to the other methods is not straightforward. Higher temperatures decrease solvent viscosity and can disrupt strong solute-matrix interactions (e.g., hydrogen bonding or dipole attractions; Richter et al., 1996), while pressure maintains the solvents in liquid state and favors its penetration into the sample matrix, usually making chemical extractions of soils or sediments more effective (Camel, 2001). The effect of pressure may be particularly important in soils or sediments with a high amount of clay, where high pressures force the solvent into small, water-blocked pores (Jansen et al., 2006), while it seemed to be negligible in the present study (clay contents from 16.7 to 18.7%; Table 1). The methods using higher pressure (M-bomb and M-ultrasound) as well as temperature (M-bomb) extracted fewer lipids than M-reflux. Moreover, cavitation phenomena

12

caused by the ultrasonic treatment (M-ultrasound) and thus disruption of soil particles (Capelo et al., 2005) did not seem to increase the extraction yields as compared to the other methods. Several factors may be responsible for these patterns: (1) Because the polarity of a solvent is reduced by increasing pressure (Jansen et al., 2006), lower yields of the more polar lipids, carrying one or more OH groups (such as the x,ω-diC16 acids), may have resulted for M-bomb and M-ultrasound as compared to M-reflux; (2) Likewise, harsher extraction conditions, such as increased temperatures and pressures, may have decreased lipid yields of M-bomb or Multrasound, especially those of ω- and mid-chain substituted hydroxycarboxylic acids, by converting these acids into their corresponding carboxylic counterparts (Goñi and Hedges, 1982, 1990; Mendez-Millan et al., 2010a). However, the yields of n-carboxylic acids ≥ C16 were mostly highest for M-reflux as well (Figs. 1 and 2); (3) An important factor responsible for overall higher lipid yields of M-reflux as compared to the other methods was likely a constant boiling of the solvent (cf. Section 2.2). This resulted in a continuous and more thorough mixing as compared to M-bomb and M-ultrasound, probably assuring that the entire solvent was brought into contact with the sample matrix. A more thorough mixing of solvent and sample may have enabled better cleavage of crosslinks within a biopolymer (such as cutin) in which ωhydroxycarboxylic and especially mid-chain substituted hydroxycarboxylic acids may be involved with one or two hydroxy groups, respectively (Kolattukudy, 1980), and lead to higher yields of these acids when employing M-reflux. Based on these results, rather conservative temperatures and no or little pressure are likely sufficient (or even preferable) for comparatively high extraction yields of bound lipids from plants, soils, and sediments. Similar results were obtained by Quénéa et al. (2012) for accelerated solvent extraction, where high temperature resulted in a decrease of lipid yields and

13

by Jansen et al. (2006) who observed a negative or no effect of increased pressure on solventextractable lipid yields in mineral soil horizons. The M-reflux method compares favorably not only with respect to the yield of plantderived lipids, but also with respect to reproducibility. M-reflux performed similarly as compared to M-ultrasound regarding the average CV over all extracted lipids from the five replicates per method and material (16.8% compared to 18.7%, respectively), but showed substantially smaller CV-values as compared to M-bomb (27.8%). These findings are especially crucial when the amount of biomarkers, such as the above-mentioned biopolymers cutin or suberin, are to be used for inferences about the source of organic carbon in soils and sediments, where often only one replicate is investigated (Feng et al., 2010; Spielvogel et al., 2014). However, a possible disadvantage of M-reflux as compared to M-ultrasound may be a limited sample throughput due to the space consuming setup of the reflux apparatus, while several sonication baths can be used at once for M-ultrasound. By adapting the ultrasound method in a way that samples are continuously sonicated, temperature is held constant during this process and headspace flushed with N2 to reduce oxidizing conditions, this method may be more effective and approximate the lipid yields of reflux methods (unpublished results; K.E. Mueller). However, reflux conditions appear to be less sensitive to the applied laboratory protocol (pressure is constant and solvent is always boiling and condensing) as compared to ultrasonication (and bomb) methods.

4.

Conclusions In summary, the reflux method used in the present study required the same amount of

solvents and time as compared to the used ultrasonication and ‘bomb’ methods, but gave

14

significantly higher yields of extracted lipids and featured the smallest standard errors. Notably, the extraction yields of particularly mid-chain substituted hydroxycarboxylic acids, including those often ascribed to cutin, were several times higher for the reflux as compared to the other methods, indicating that estimation of individual lipids may be sensitive to the method of hydrolysis. Our results are of particular relevance for the investigation of SOM fractions, where material for extraction is commonly scarce and high lipid yields and representativity of extractions are essential.

Acknowledgments This work was realized within the grant MU3021-4/1 by the Deutsche Forschungsgemeinschaft (DFG) and with the support of the SoWa Research Infrastructure funded by MEYS CZ grant LM2015075, programme "Projects of Large Infrastructure for Research, Development, and Innovations". Part of the equipment used in this study was purchased from the Operational Programme Prague-Competitiveness (Project CZ.2.16/3.1.00/21516). The authors would like to thank Prof. Dr. Thorsten E.E. Grams and Franz Buegger for supply of plant material and Franziska Fella, Tabea Bartels, and Maria Greiner for help in the laboratory. We thank the reviewers for their helpful comments.

Associate Editor–Klaas Nierop

References Angst, G., Heinrich, L., Kögel-Knabner, I., Mueller, C.W., 2016a. The fate of cutin and suberin of decaying leaves, needles and roots – Inferences from the initial decomposition of bound fatty acids. Organic Geochemistry 95, 81–92. 15

Angst, G., John, S., Mueller, C.W., Kögel-Knabner, I., Rethemeyer, J., 2016b. Tracing the sources and spatial distribution of organic carbon in subsoils using a multi-biomarker approach. Scientific Reports 6. doi:10.1038/srep29478 Angst, G., Mueller, K.E., Kögel-Knabner, I., Freeman, K.H., Mueller, C.W., 2017. Aggregation controls the stability of lignin and lipids in clay-sized particulate and mineral associated organic matter. Biogeochemistry 132, 307–324. Camel, V., 2001. Recent extraction techniques for solid matrices – supercritical fluid extraction, pressurized fluid extraction and microwave-assisted extraction: their potential and pitfalls. The Analyst 126, 1182–1193. Campbell, E.E., Paustian, K., 2015. Current developments in soil organic matter modeling and the expansion of model applications: a review. Environmental Research Letters 10, 123004. Capelo, J.L., Maduro, C., Vilhena, C., 2005. Discussion of parameters associated with the ultrasonic solid-liquid extraction for elemental analysis (total content) by electrothermal atomic absorption spectrometry. An overview. Ultrasonics Sonochemistry 12, 225–232. Carrington, E.M., Hernes, P.J., Dyda, R.Y., Plante, A.F., Six, J., 2012. Biochemical changes across a carbon saturation gradient: Lignin, cutin, and suberin decomposition and stabilization in fractionated carbon pools. Soil Biology & Biochemistry 47, 179–190. Castellano, M.J., Mueller, K.E., Olk, D.C., Sawyer, J.E., Six, J., 2015. Integrating plant litter quality, soil organic matter stabilization, and the carbon saturation concept. Global Change Biology 21, 3200–3209. Cotrufo, M.F., Soong, J.L., Horton, A.J., Campbell, E.E., Haddix, M.L., Wall, D.H., Parton, W.J., 2015. Formation of soil organic matter via biochemical and physical pathways of litter mass loss. Nature Geoscience 8, 776–779. Cotrufo, M.F., Wallenstein, M.D., Boot, C.M., Denef, K., Paul, E., 2013. The Microbial

16

Efficiency-Matrix Stabilization (MEMS) framework integrates plant litter decomposition with soil organic matter stabilization: Do labile plant inputs form stable soil organic matter? Global Change Biology 19, 988–995. Crow, S.E., Lajtha, K., Filley, T.R., Swanston, C.W., Bowden, R.D., Caldwell, B.A., 2009. Sources of plant-derived carbon and stability of organic matter in soil: implications for global change. Global Change Biology 15, 2003–2019. Diefendorf, A.F., Freeman, K.H., Wing, S.L., Graham, H. V., 2011. Production of n-alkyl lipids in living plants and implications for the geologic past. Geochimica et Cosmochimica Acta 75, 7472–7485. Dignac, M.F., Rumpel, C., 2006. Relative distributions of phenol dimers and hydroxy acids in a cultivated soil and above ground maize tissue. Organic Geochemistry 37, 1634–1638. Eglinton, G., Hamilton, R.J., 1967. Leaf epicuticular waxes. Science 156, 1322–1335. Feng, X., Simpson, M.J., 2008. Temperature responses of individual soil organic matter components. Journal of Geophysical Research: Biogeosciences 113. doi:10.1029/2008JG000743 Feng, X., Xu, Y., Jaffé, R., Schlesinger, W.H., Simpson, M.J., 2010. Turnover rates of hydrolysable aliphatic lipids in Duke Forest soils determined by compound specific 13C isotopic analysis. Organic Geochemistry 41, 573–579. Filley, T.R., McCormick, M.K., Crow, S.E., Szlavecz, K., Whigham, D.F., Johnston, C.T., van den Heuvel, R.N., 2008. Comparison of the chemical alteration trajectory of Liriodendron tulipifera L. leaf litter among forests with different earthworm abundance. Journal of Geophysical Research: Biogeosciences 113. doi:10.1029/2007JG000542 Fiorentino, G., Spaccini, R., Piccolo, A., 2006. Separation of molecular constituents from a humic acid by solid-phase extraction following a transesterification reaction. Talanta 68,

17

1135–1142. Freimuth, E.J., Diefendorf, A.F., Lowell, T. V., 2017. Hydrogen isotopes of n-alkanes and nalkanoic acids as tracers of precipitation in a temperate forest and implications for paleorecords. Geochimica et Cosmochimica Acta 206, 166–183. Geßler, A., Keitel, C., Kreuzwieser, J., Matyssek, R., Seiler, W., Rennenberg, H., 2007. Potential risks for European beech (Fagus sylvatica L.) in a changing climate. Trees 21, 1–11. Goñi, M.A., Hedges, J.I., 1982. Characterization of lignin by gas capillary chromatography of cupric oxide oxidation products. Analytical Chemistry 54, 174–178. Goñi, M.A., Hedges, J.I., 1990. Cutin-derived cupric oxide reaction products from purified cuticles and tree leaves. Geochimica et Cosmochimica Acta 54, 3065–3072. Graça, J., Schreiber, L., Rodrigues, J., Pereira, H., 2002. Glycerol and glyceryl esters of ωhydroxyacids in cutins. Phytochemistry 61, 205–215. Hamer, U., Rumpel, C., Dignac, M.F., 2012. Cutin and suberin biomarkers as tracers for the turnover of shoot and root derived organic matter along a chronosequence of Ecuadorian pasture soils. European Journal of Soil Science 63, 808–819. Hauff, S., Chefetz, B., Shechter, M., Vetter, W., 2010. Determination of hydroxylated fatty acids from the biopolymer of tomato cutin and their fate during incubation in soil. Phytochemical Analysis 21, 582–589. Holloway, P.J., Brown, G.A., Wattendorff, J., 1981. Ultrahistochemical detection of epoxides in plant cuticular membranes. Journal of Experimental Botany 32, 1051–1066. IUSS Working Group WRB, 2014. World Reference Base for Soil Resources 2014. International soil classification system for naming soils and creating legends for soil maps, World Soil Resources Reports No. 106. FAO, Rome. doi:10.1017/S0014479706394902 Jandl, R., Lindner, M., Vesterdal, L., Bauwens, B., Baritz, R., Hagedorn, F., Johnson, D.W.,

18

Minkkinen, K., Byrne, K.A., 2007. How strongly can forest management influence soil carbon sequestration? Geoderma 137, 253–268. Jansen, B., Nierop, K.G.J., Kotte, M.C., de Voogt, P., Verstraten, J.M., 2006. The applicability of accelerated solvent extraction (ASE) to extract lipid biomarkers from soils. Applied Geochemistry 21, 1006–1015. Kögel-Knabner, I., 2002. The macromolecular organic composition of plant and microbial residues as inputs to soil organic matter. Soil Biology and Biochemistry 34, 139–162. Kögel-Knabner, I., Hatcher, P.G., Tegelaar, E.W., de Leeuw, J.W., 1992. Aliphatic components 13

of forest soil organic matter as determined by solid-state C NMR and analytical pyrolysis. Science of The Total Environment 113, 89–106. Kolattukudy, P.E., 1980. Bio-polyester membranes of plants - cutin and suberin. Science 208, 990–1000. Kolattukudy, P.E., Kronman, K., Poulose, A.J., 1975. Determination of structure and composition of suberin from roots of carrot, parsnip, rutabaga, turnip, red beet and sweetpotato by combined gas-liquid-chromatography and mass-spectrometry. Plant Physiology 55, 567–573. Lichtfouse, É., Berthier, G., Houot, S., Barriuso, E., Bergheaud, V., Vallaeys, T., 1995. Stable carbon isotope evidence for the microbial origin of C14–C18 n-alkanoic acids in soils. Organic Geochemistry 23, 849–852. Matzke, K., Riederer, M., 1991. A comparative study into the chemical constitution of cutins and suberins from Picea abies L. Karst, Quercus robur L., and Fagus sylvatica L. Planta 185, 233–245. Mendez-Millan, M., Dignac, M.-F.F., Rumpel, C., Derenne, S., 2010. Quantitative and qualitative analysis of cutin in maize and a maize-cropped soil: Comparison of CuO

19

oxidation, transmethylation and saponification methods. Organic Geochemistry 41, 187– 191. Mendez-Millan, M., Dignac, M.F., Rumpel, C., Derenne, S., 2011. Can cutin and suberin biomarkers be used to trace shoot and root-derived organic matter? A molecular and isotopic approach. Biogeochemistry 106, 23–38. Mendez-Millan, M., Dignac, M.F., Rumpel, C., Rasse, D.P., Derenne, S., 2010. Molecular dynamics of shoot vs. root biomarkers in an agricultural soil estimated by natural abundance 13

C labelling. Soil Biology & Biochemistry 42, 169–177.

Mueller, K.E., Polissar, P.J., Oleksyn, J., Freeman, K.H., 2012. Differentiating temperate tree species and their organs using lipid biomarkers in leaves, roots and soil. Organic Geochemistry 52, 130–141. Naafs, D.F.W., Van Bergen, P.F., 2002. Effects of pH adjustments after base hydrolysis: Implications for understanding organic matter in soils. Geoderma 106, 191–217. Nierop, K.G.J., Naafs, D.F.W., Verstraten, J.M., 2003. Occurrence and distribution of esterbound lipids in Dutch coastal dune soils along a pH gradient. Organic Geochemistry 34, 719–729. Nierop, K.G.J., Verstraten, J.M., 2004. Rapid molecular assessment of the bioturbation extent in sandy soil horizons under, pine using ester-bound lipids by on-line thermally assisted hydrolysis and methylation-gas chromatography/mass spectrometry. Rapid Communications in Mass Spectrometry 18, 1081–1088. Otto, A., Shunthirasingham, C., Simpson, M.J., 2005. A comparison of plant and microbial biomarkers in grassland soils from the Prairie Ecozone of Canada. Organic Geochemistry 36, 425–448. Otto, A., Simpson, M.J., 2006. Sources and composition of hydrolysable aliphatic lipids and

20

phenols in soils from western Canada. Organic Geochemistry 37, 385–407. Pander, J., Mueller, M., Geist, J., 2015. A comparison of four stream substratum restoration techniques concerning interstitial conditions and downstream effects. River research and applications 31, 239–255. Pisani, O., Frey, S.D., Simpson, A.J., Simpson, M.J., 2015. Soil warming and nitrogen deposition alter soil organic matter composition at the molecular-level. Biogeochemistry 123, 391–409. Quénéa, K., Mathieu, J., Derenne, S., 2012. Soil lipids from accelerated solvent extraction: Influence of temperature and solvent on extract composition. Organic Geochemistry 44, 45– 52. Ray, A.K., Lin, Y.Y., Gerard, H., Chen, Z., Osman, S.F., Fett, W.F., Moreau, R.A., Stark, R.E., 1995. Separation and identification of lime cutin monomers by high performance liquid chromatography and mass spectrometry. Phytochemistry 38, 1361–1369. Richter, B.E., Jones, B.A., Ezzell, J.L., Porter, N.L., Avdalovic, N., Pohl, C., 1996. Accelerated solvent extraction: A technique for sample preparation. Analytical Chemistry 68, 1033– 1039. Riederer, M., Matzke, K., Ziegler, F., Kogelknabner, I., 1993. Occurence, distribution and fate of the lipid plant biopolymers cutin and suberin in temperate forest soils. Organic Geochemistry 20, 1063–1076. Rumpel, C., Seraphin, A., Dignac, M., Michaelis, W., Eusterhues, K., Ko, I., 2005. Effect of base hydrolysis on the chemical composition of organic matter of an acid forest soil. Organic Geochemistry 36, 239–249. Spielvogel, S., Prietzel, J., Leide, J., Riedel, M., Zemke, J., Kögel-Knabner, I., 2014. Distribution of cutin and suberin biomarkers under forest trees with different root systems. Plant and

21

Soil 381, 95–110. Wiesenberg, G.L.B., Dorodnikov, M., Kuzyakov, Y., 2010. Source determination of lipids in bulk soil and soil density fractions after four years of wheat cropping. Geoderma 156, 267– 277. Wiesenberg, G.L.B., Schwark, L., Schmidt, M.W.I., 2004. Improved automated extraction and separation procedure for soil lipid analyses. European Journal of Soil Science 55, 349–356.

22

Figure captions Fig. 1. Extraction yields for beech (a) leaves, (b) soil, and (c) sediment from M-bomb, Multrasound, and M-reflux ± SE. DA: α,ω-dicarboxylic acid; n-: n-carboxylic acid; ω-: ωhydroxycarboxylic acid; Monomers labelled differently are mid-chain substituted hydroxy or epoxy acids. Significant differences are indicated by different letters. Monomers not significantly differing between the methods are not labelled with a letter. Marginally significant differences (0.05 ≥ p ≤ 0.01) are marked by an asterisk. n = 5 for each bar. Fig. 2. Extraction yields for maize (a) leaves, (b) soil, and (c) sediment from M-bomb, Multrasound, and M-reflux ± SE. DA: α,ω-dicarboxylic acid; n-: n-carboxylic acid; ω-: ωhydroxycarboxylic acid; Monomers labelled differently are mid-chain substituted hydroxy or epoxy acids. Significant differences are indicated by different letters. Monomers not significantly differing between the methods are not labelled with a letter. Marginal significant differences (0.05 ≥ p ≤ 0.01) are marked by an asterisk. n = 5 for each bar.

23

24

25

Table 1 General characteristics of the study areas. Carbon contents correspond to organic C contents. Zajíc, Czech Republic

Freising, Germany

Soil type

Luvic Stagnosol

Calcic Gleysol

Vegetation

European beech

Maize

Soil (Ah)

3.4 ± 0.01

7.3 ± 0.1

Sediment

4.7 ± 0.01

7.2 ± 0.02

Sand

67.3

13.8

Silt

14.0

68.8

Clay

18.7

17.4

Sand

73.4

49.2

Silt

7.9

34.1

Clay

18.7

16.7

Leaves

456 ± 6.2

431 ± 0.8

Soil

23 ± 0.3

56 ± 1.0

Sediment

58 ± 0.1

32 ± 1.0

pH value

Soil Texture [%]

Sediment texture [%]

C content [mg/g]

26

Highlights: -

Reflux, ultrasound, and ‘bomb’ methods for extraction of bound lipids were compared The reflux method gave consistently higher yields independent of extracted material Yield of certain lipids, including cutin acids, may be sensitive to the method used

27

28