Gompertz Kinetics of Soil Microbial Biomass in ... - Science Direct

4 downloads 0 Views 308KB Size Report
biopolymer films of low enthalpy of combustion (starch) in order to increase the carbon conversion into the microbial .... each film (1 cm x 1 cm) each per 75 g aerobic soil. ... The SMB concentration was calculated according to the following Equation (1). ... which was approximately 9.10% less than found for control film.
Available online at www.sciencedirect.com

ScienceDirect Procedia Engineering 148 (2016) 553 – 560

4th International Conference on Process Engineering and Advanced Materials

Gompertz Kinetics of Soil Microbial Biomass in Response to Lignin Reinforcing of Urea-Crosslinked Starch Films Zahid Majeed, Nurlidia Mansor*,Shaja ismail, Ranitha Mathialagan, Zakaria Man Chemical Engineering Department, Universiti Teknologi PETRONAS, Bandar Seri Iskandar, 32610 Tronoh, Perak, Malaysia

Abstract Soil microbial biomass (SMB) is strictly a function of substrate quality and composition under the soil environmental conditions. The current research objective was to determine the gompertz kinetics of SMB in response of predictive variables; 5, 10, 15 and 20% lignin, used to reinforce the urea-crosslinked starch (UcS) films in aerobic soil. Through gompertz kinetics, the specific growth rate (µ max), lag phase (λ), asymptote (A) and inflection (ti) of growth curve were determined for SMB. The results of the study showed that gompertz kinetic predicted reasonably a good agreement between the experimental and predicted soil microbial biomass (R2 > 0.942; Error Mean Square, 0.010-0.089) for all lignin reinforced UcS films in aerobic soil. The value of A was 13.39% higher in soil added with lignin reinforced UcS films than the control film. The average µ max of lignin reinforced UcS films was achieved 3.59 day-1 which was approximately 9.10% less than µ max observed in control film. The average λ was delayed 76.69% more in lignin reinforced UcS films possible due to the lignin's recalcitrant nature to biodegradation by microorganisms. The ti was found 39.53% higher for the lignin reinforced UcS films. The higher oxidation state and higher standard enthalpy of combustion of the UcS increase by addition of the lignin which could cause the higher carbon to microbial biomass conversion. These results predicted by gompertz kinetics have demonstrated the implication of lignin to use in biopolymer films of low enthalpy of combustion (starch) in order to increase the carbon conversion into the microbial biomass rather than rapid conversion into the CO2. © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license © 2016 The Authors. Published by Elsevier Ltd. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of organizing the organizing committee of ICPEAM Peer-review under responsibility of the committee of ICPEAM 2016 2016. Keywords: Gompertz kinetics; Urea-crosslinked starch; Soil microbial biomass; Biodegradability, Aerobic microcosm

*

Corresponding author. Tel.: +0605 3687583 ; fax: +0605 3656176. E-mail address: [email protected]

1877-7058 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of ICPEAM 2016

doi:10.1016/j.proeng.2016.06.510

554

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

1. Introduction Bacteria and fungi comprising more than 90% of the soil microbial biomass, and clear evidence exists that these groups function differently in the decomposition process of natural polymers [1]. Growth yield of microorganisms varies remarkably with the free energy of the oxidation state of the carbon in the polymer. The capabilities of energy assimilation by microorganisms could fluctuate with free energy of the carbon substrate, on which they metabolize during the period of their logistic growth. In the presence of a carbon in polymer with relatively low free-energy content, a large proportion of the carbon converts into CO2 and energy. The energy liberated from catabolic reactions is not substantially utilized for cell growth and proliferation; rather it is used to sustain the living activities of bacterial cells. It has been suggested that the carbon transference between the maintenance catabolism and anabolism are inversely relates to the free-energy content of the organic substrate [2]. Substrate quality could determine the microbial changes in the soil. According to Wickings et al. [3], distinct decomposer communities impose constraints on substrate decomposition regardless of the difference in quality of substrate and stage of decomposition [4], [5]. Jagadamma et al. [6] tested uniformly-labeled 14C substrates (glucose, starch, cinnamic acid and stearic acid) in four different soils (a temperate Mollisol, a tropical Ultisol, a sub-arctic Andisol, and an arctic Gelisol) for the substrates chemistry influence on the microbial decomposition. Cumulative microbial respiration was higher for those soils added with glucose (52 to 60%), starch (39 to 49%). In the work of Hernández and Hobbie [7] microbial respiration responds predictably to the substrate composition and its quantity which maximizes by the addition of labile substrates and greater substrate quantity. Biodegradation models often estimate the maximum rate of substrate decay based on either substrate availability or microbial activity, but in reality, biodegradation is limited by both factors [8]. The rate of biodegradation based on microbial activity are limited by substrate or by microbial activity [9]. One benefit of this approach, mentioned by these researchers that it could be applied to specific chemical constituents of the litter, whose intrinsic decay rate coefficients, guilds characteristics (e.g., metabolic and enzymatic attribute) are very different. The Gompertz model has been the most widely used in describing the microbial growth for its simple formation. Gompertz kinetics, in contrast, works well when the potential of polymer or their blends degradation increases over time scale [10]. In a laboratory, Gompertz kinetics studies of polycaprolactone-starch blend shows a significantly long lag-phase period (λ) and the biodegradation rates slows more as compared to cellulose (control) [11]. When different models were applied to the bacterial growth, Gompertz model has showed a better growth curves fitted than the logistic, linear, quadratic and rth-power. The acceptability of the gompertz's model fitting has more higher values (~70%) compared to the logistic model (~52%) [12]. In general, these models describes only the number of microorganisms and do not include the consumption of substrate and formation of products (e.g., Monod model). Monod introduced its fundamental model that describes the substrate dependent microbial growth under assumption of non-inhibitory effect of substrate or its products. Haldane [13] introduced in Monod’s model a self-inhibitory constant to recognize inhibition of the reaction due to inhibition effect of products in biodegradable substrate kinetics. Substrate consumption and growth-associated product formation can also be quantitatively described based on growth models, as growth is a result of catabolic and anabolic enzymatic activities comes through substrate. Qi et al. [14] provided equation which relates to initial cell density of microbial biomass with substrate-alginate–chitosan–alginate consumption. Starch is a complex of amylose and amylopectin interwoven chains, composed of anhydrous glucose units links primarily through α-D-(1→4) glycosidic bonds. Starch is used as a source of cellular energy by both animals and microorganisms. Glucose molecules are released after breaking of α-D-(1→4) glycosidic bonds by the amylases enzymes secreted by the microorganisms present in the soil. Soil microorganisms utilize these glucose molecules to built up new microbial biomass. Starch hydrophilicity and high rate of biodegradation are limitation to its direct use as a biodegradable matrix particularly where applications are destined to the soil. It has been shown that loading of 10% kraft lignin (commercially known as Indulin AT) reduced the starch hydrophilicity [6]. Lignin is another natural plant based biopolymer and abundantly produced as the waster of pulp industry. Its ability to reduce the starch hydrophobicity, inhibit the microbial activity, radical scavenging properties and reduction of starch biodegradability [15] justified to study the lignin response to microbial biomass growth on the urea-crosslinked starch (UcS) films. Lignin is an amorphous heteropolymer formed by phenyl-propane units, which mainly consist of three types of aromatic units: guaiacyl, syringyl and phenylpropane. Lignin compared to other biopolymers is relatively hydrophobic material. The film forming properties in combination with its hydrophobic and antimicrobial character has highlighted as interesting group of material [5]. Compared to starch, lignin is only biodegradable by

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

special microbial strains predominately by the fungal species (filamentous basidiomycetous fungi). Avella et al. [16] showed that reinforcing PHB with a lignocellulose material (wheat straw) does not affect its biodegradation in longterm soil burial tests. In the present research work, soils were incubated with 5, 10, 15 and 20% lignin reinforced UcS films. The soil microbial biomass response to the lignin in UcS films is predicted through the gompertz kinetics. Through gompertz kinetics, the specific growth rate, lag phase, maximum log growth and inflection time of the microbial growth curvature are determined and compared for different lignin %. To our best of knowledge, current literature do not provide any account of study on the gompertz kinetics of the soil microbial biomass for the lignin reinforced ureacrosslinked starch films in aerobic soil

Nomenclature A L MSE SMB t ti UcS X µ max O e

Asymptote is the maximal value of the soil microbial biomass over the time Lignin Error mean square Soil microbial biomass Time of the soil microbial biomass growth (days) The point of inflection where the microbial curvature of the curve of soil microbial biomass changed from convex to concave (defined as sigmoidal curve shape factor) Urea-crosslinked starch Logarithm of relative population size of soil microbial biomass over the time Maximum specific growth rate (day-1), slope of the tangent line at the point of inflection " ti" Lag time (days) Euler's number

2. Materials and Methods 2.1. Materials Tapioca starch (Kapal ABC, Malaysia) was purchased from the local market. Urea (46% total nitrogen) was received from Petronas Fertilizer [(Kedah) Sdn. Bhd., Malaysia]. Loamy sand soil was collected from Titi Gantung Bota paddy field (4.36oN, 100.84oE), Perak, Malaysia and its chemical and physical characteristics were already cited elsewhere [17]. Alkaline kraft lignin (Indulin AT) was purchased from Sigma Aldrich (USA). All other chemicals of analytical grade ( > 99%) were used in this research. 2.2. Synthesis of Lignin reinforced UcS Films Lignin reinforced UcS Films were synthesized under optimized conditions according to the conditions detailed in the work of Sarwono [18]. Briefly, tapioca starch (5 g) was mixed with 100 mL of deionized water and heated at 80 ºC for 30 min. After this, urea (2 g) and disodium tetra-borate (0.45 g) was added. Then lignin (L) was added at 5, 10, 15 and 20% of initial tapioca starch weight. Reaction mixture was stirred for further 3 hours at 80ºC. The solution of each composition was poured into a square polypropylene plastic bowls (12 cm x 12 cm) and dried at 40oC overnight in a vacuum oven to obtain composite films. Composite films were dried further after setting for 2 hours at a post curing temperature 120 °C in air drying oven to complete cross-linking reactions of urea with starch. Composite films of Urea-crosslinked starch (UcS) was named as UcS5%L, UcS10%L, UcS15%L and UcS20%L. The UcS film which received no lignin served as experimental control.

555

556

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

2.3. Soil microcosm set up Films were amended with aerobic soil with 50% maximum water holding capacity (60% moisture) after soil condition for 10 days to remove the history of any previous substrate amendment. The soil was mixed with 2.54 g of each film (1 cm x 1 cm) each per 75 g aerobic soil. Using a spatula, the films and soil were mixed well and extra care was given to ensure that the films were not exposed to the soil surface. Soil pots were placed under dark conditions at ambient temperature (28 ± 2 oC). Intermittently, the soil pot caps were removed to allow soil oxygenation. Soil pots were collected at 0, 1, 2, 4 days to assess the soil microbial biomass. All the samples were run in duplicate including the soil blanks that did not receive any film. 2.4. Determination of Soil Microbial Biomass Quantification of the soil microbial biomass (SMB) was carried until 4 days of the incubation applying the chloroform-fumigation-extraction method [19]. Moist paper towels were placed in the desiccators to maintain humidity during the fumigation to avoid surface dryness of the samples. Ethanol free CHCl 3, 50 mL was transferred into 100 mL beaker and then 5 g anti-bumping granules were added to boil chloroform calmly. The beaker was placed in the middle of the desiccator along with the soil samples. Moist soil samples approx. 15 g were spread evenly on the petri dishes (Pyrex glass, 40 mm × 12 mm) and placed in a stacked position to allow maximum area for the fumigation (Fig. 1). Beaker containing 50 mL of 1 M NaOH was placed inside the desiccator to trap any evolved CO2 The desiccators was then closed and evacuated for 2–5 min until the CHCl3 boiled vigorously. This process was repeated 2-3 times. The sealed desiccator was kept in the dark for 24 h at 25ºC. After the incubation period (overnight), the vacuum was released under fume hood until complete chloroform residue was removed. Each evacuation was maintained for 2 minutes and air was allowed into the fumigation chamber after each evacuation to remove residual of chloroform.

2.4.1. Chemical Extraction Fumigated and non-fumigated soil samples were extracted with 30 ml of 0.5 M K 2SO4 (2:1) and both samples extracted under similar conditions. A blank sample of 20 mL (0.5 M) K2SO4 was included with each batch. The samples were shaken for 30 minutes and then filtered using Whatman Number 41 (150 mm diameter) filter papers. Soil residues were rinsed with two times with 5 ml of K 2SO2. Final extract volume achieved was 30ml that was stored at 4oC until further analysis.

Fig. 1. Soil fumigation process for soil microbial biomass

557

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

2.4.2. Microbial biomass carbon The extracts of 5 mL was diluted with 25 ml of deionized water and transferred into 40 mL glass vials fixed with silicon septa and were placed in the automatic rack for analysis on Total Organic Carbon Analyzer (Shimadzu, Japan). The SMB concentration was calculated according to the following Equation (1). SMB (μg C g-1 dry soil weight) = (F – NF) / KC

(1)

Where F – NF is the soluble organic C extracted from fumigated (F) soils minus that extracted from non-fumigated (NF) soil and KC = 0.45, is referred to the efficiency of extraction of SMB carbon as the proportion of microbial C evolved as CO2 = 0.45 for 10 days incubations at 25°C [20]. 2.4.3. Gompertz kinetics The gompertz kinetic Equation (2) was used to predict the response of soil microbial biomass measured for the lignin reinforcement of UcS films in aerobic soil by using OriginPro software, version 9.0.0 (OriginLab Corporation, Northampton, MA, USA). Growth of the microbial biomass is considered follow sigmoid growth with distinct specific microbial growth rate (µ max), lag time (O), log maximum growth (A) and curvature of the curve (ti). Therefore, in a critical biodegradation, fixing the initial substrate concentration, and varying the time provides good estimate of microbe’s saturation time for the initial substrate concentration and its composition effect. Therefore, time is more fundamental and reliable predictor of the biomass and substrate relationship in any research work focused on the film's biodegradation. Authors assumed that films over the short period of time does not have any inhibitory effect and starch is the labile component that derives the SMB.

X

­ ªP e º½ A exp ® exp « max (O  t )  1» ¾ ¬ A ¼¿ ¯

(2)

Error mean square (MSE) of the gompertz kinetics was calculated by following equation (3) MSE

Sum (exp erimental values  predicted values )

(3)

deg ree of freedom

All statistical analyses were performed using OriginPro software, version 9.0.0 (OriginLab Corporation, Northampton, MA, USA). 3. Results and Discussion In Fig. 2 is shown the gompertz fitting of initial SMB to different lignin reinforced UcS films in aerobic soil. Reasonably a good fit between the experimental and predicted soil microbial biomass (R2 > 0.942; MSE, 0.0100.089) was observed for the initial SMB growth on lignin reinforced UcS films. The gompertz kinetic parameters predicted are given in Error! Reference source not found.1. The asymptote (A) of log soil microbial biomass exhausted earlier in control film as compared to lignin reinforced UcS films. The value of A was 13.39% higher for the soil added with lignin reinforced UcS films than the control film. The specific soil microbial biomass growth, with an average value of µ max, 3.59 day-1 was achieved for lignin reinforced UcS films which was approximately 9.10% less than found for control film. This is due to the low carbon oxidation state and low standard enthalpy of combustion of the starch in UcS film i.e., ~ 4.15 kcal/g [21] compared to lignin i.e., ~ 6.37 kcal/g [22]. The higher enthalpy values could have caused higher carbon to microbial biomass conversion (lower CO2 release) of lignin reinforced UcS films as compared to control film. The higher enthalpy of combustion of lignin (as carbon substrate) corresponds negatively to the μmax whose values were lower in lignin reinforced UcS film due to higher microbial biomass assimilation efficiencies [2].

558

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

4.5 4.0

Control-Exp Control-Pred Lambda ti

UcS5%L-Exp UcS5%L-Pred Lambda ti

R2 = 0.948 MSE = 0.014

R2 = 0.942 MSE = 0.014

UcS15%L-Exp UcS15%L-Pred Lambda ti

UcS10%L-Exp UcS10%L-Pred Lambda ti

UcS20%L-Exp UcS20%L-Pred Lambda ti

Log Soil Microbial Biomass

3.5 3.0 2.5 2.0

R2 = 0.947 MSE = 0.010

R2 = 0.925 MSE = 0.089

R2 = 0.947 MSE = 0.037

1.5 1.0 0.5 0.0

0 1 2 3 4 5

0 1 2 3 4 5 6 7

0 1 2 3 4 5 6 7 Time (days)

0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7

Fig. 2. Gompertz fitting of soil microbial biomass as a function of lignin reinforced UcS film. The lag period (λ) of the SMB showed an average values of 0.365 day in the case of lignin reinforced UcS films as compared to 0.207 day in control film. This shows that the λ value was increased about 76.69% more by the lignin which is possibly results from the lignin's slow biodegradation properties. The ti, is an estimate of shift in microbial growth curve from linear segment to upper values of A, and its average values was found 0.694 day for the lignin reinforced UcS films as compared with 0.479 day in control film. This predicted that ti value change was 39.53% higher due to the lignin. Table 1. Gompertz kinetic parameter of soil microbial biomass after lignin reinforcement of UcS films. Films UcS

A

μmax (day-1)

λ (day)

ti (day)

2.917

3.950

0.207

0.479

UcS5%L

3.196

4.081

0.175

0.463

UcS10%L

3.313

2.977

0.729

1.139

UcS15%L

2.777

3.585

0.252

0.537

UcS20%L

3.368

3.718

0.307

0.640

Gompertz kinetic estimates for lignin reinforced UcS films showed some inconsistencies in the data (e.g., UcS15%L, Fig. 2) which points to the complex heterogeneity of the biodegradation of the films at the interface of soil and film. Once the lignin reinforced UcS films are added into the aerobic soil, oxygen flux at microsites of soil aggregates also change which could be the reason of inconsistencies in the microbial growth dynamics [23]. The initial density of carbon substrate degrading microbial cells, if low under unfavourable conditions, it could increased the λ with further regression in μmax [10]. This unfavourable effect of the lignin in the UcS film to the soil microbial biomass could not be ignored for the observed deviations in the predicted and experimental data.

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

4. Conclusion This research work concluded that the lignin reinforcing of UcS films has successfully predicted the gompertz kinetic of soil microbial biomass. Lignin addition could reduce the μmax while it increased the parameter A of the SMB. These results has the implication of lignin to use in biopolymer films with low enthalpy of combustion (starch) in order to increase the carbon conversion into the microbial biomass rather than rapid conversion into the CO2. In future, further work could be extended to determine the kinetic rates of microbial growth for different compositions of films specifically vary in their constituent's enthalpy of combustion.

Acknowledgement The authors gratefully acknowledge Ministry of Higher Education (MOHE), Malaysia, for funding under Long Term Research Grant Scheme (LRGS) for the project of OneBaja. We are also grateful to Universiti Teknologi Petronas, Malaysia for providing research facilities and great support.

References [1] M. A. De Graaff, A. T. Classen, H. F. Castro, and C. W. Schadt, Labile soil carbon inputs mediate the soil microbial community composition and plant residue decomposition rates, New Phytol. 188(2010) 1055-1064. [2] E. Chiellini, A. Corti, S. D’Antone, and N. C. Billingham, Microbial biomass yield and turnover in soil biodegradation tests: carbon substrate effects, J. Polym. Environ. 15(2007) 169-178. [3] K. Wickings, A. S. Grandy, S. C. Reed, and C. C. Cleveland, The origin of litter chemical complexity during decomposition, Ecol. lett. 15(2012) 1180-1188. [4] G. Thouand, M. J. Durand, A. Maul, C. Gancet, and H. Blok, New concepts in the evaluation of biodegradation/persistence of chemical substances using a microbial inoculum, Front. Microbiol. 2(2011) 1-6. [5] K. Wickings, A. S. Grandy, S. Reed, and C. Cleveland, Management intensity alters decomposition via biological pathways, Biogeochemistry, 104(2011) 365-379. [6] S. Jagadamma, M. Mayes, J. Steinweg, and S. Schaeffer, Substrate quality alters microbial mineralization of added substrate and soil organic carbon, Biogeosci. Discuss. 11(2014), 445-4482 [7] D. L. Hernández and S. E. Hobbie, The effects of substrate composition, quantity, and diversity on microbial activity, Plant and Soil, 335(2010), 397-411. [8] M. D. Wallenstein and M. N. Weintraub, Emerging tools for measuring and modeling the in situ activity of soil extracellular enzymes, Soil Biol. Biochem. 40(2008), 2098-2106. [9] D. L. Moorhead and R. L. Sinsabaugh, A theoretical model of litter decay and microbial interaction, Ecol. Monogr. 76(2006), 151-174. [10] A. R. Johnsen, P. J. Binning, J. Aamand, N. Badawi, and A. E. Rosenbom, The Gompertz function can coherently describe microbial mineralization of growth-sustaining pesticides, Environ. Sci. Technol., 47(2013), 8508-8514. [11] H. Cho, H. Moon, M. Kim, K. Nam, and J. Kim, Biodegradability and biodegradation rate of poly (caprolactone)-starch blend and poly (butylene succinate) biodegradable polymer under aerobic and anaerobic environment, Waste manage. 31(2011), 475-480. [12] M. Zwietering, I. Jongenburger, F. Rombouts, and K. Van't Riet, Modeling of the bacterial growth curve, Appl. Environ. Microbiol. 56(1990), 1875-1881. [13] J. Andrews, A mathematical model for the continuous culture of microorganisms utilizing inhibitory substance, Biotechnol. Bioeng. 10 (1968), 707-723. [14] W. Qi, J. Ma, W. Yu, Y. Xie, W. Wang and X. Ma, Behavior of microbial growth and metabolism in alginate–chitosan–alginate (ACA) microcapsules. Enzyme Microb. Technol. 38(2006), 697-704. [15] Z. Majeed, N. Mansor, Z. Man and S. Wahid, Lignin reinforcement of urea-crosslinked starch films for reduction of starch biodegradability to improve slow nitrogen release properties under natural aerobic soil condition, e-Polymers. 16(2016), 159-170. [16] M. Avella, G. La Rota, E. Martuscelli, M. Raimo, P. Sadocco, G. Elegir, and R. Riva, Poly (3-hydroxybutyrate-co-3-hydroxyvalerate) and wheat straw fibre composites: thermal, mechanical properties and biodegradation behaviour, J. Mat. Sci. 35(2000), 829-836. [17] Z. Majeed, N. K. Ramli, N. Mansor, and Z. Man, Lignin loading effect on biodegradability and nitrogen release properties of urea modified tapioca starch in wet soil, Key Eng. Mat. 594(2014) 798-802. [18] A. Sarwono, "Lignin reinforced urea crosslinked starch gel as slow releae urea fertilizer," PhD Thesis (2015), Chemical Engineering Department, Universiti Teknologi Petronas, Malaysia. [19] E. Vance, P. Brookes, and D. Jenkinson, An extraction method for measuring soil microbial biomass C, Soil Biol. Biochem. 19(1987), 703707.

559

560

Zahid Majeed et al. / Procedia Engineering 148 (2016) 553 – 560

[20] D. S. Jenkinson and J. N. Ladd, Microbial biomass in soil: measurement and turnover, in: E. A. Paul and J. N. Ladd, (Eds.), Soil Biochemistry, Dekker, New York, pp. 415-471. [21] G. J. Kabo, O. V. Voitkevich, A. V. Blokhin, S. V. Kohut, E. N. Stepurko, and Y. U. Paulechka, Thermodynamic properties of starch and glucose, J. Chem. Thermodyn. 59(2013), 87-93. [22] O. V. Voitkevich, G. J. Kabo, A. V. Blokhin, Y. U. Paulechka, and M. V. Shishonok, Thermodynamic properties of plant biomass components. Heat capacity, combustion energy, and gasification equilibria of lignin, J. Chem. Eng. Data. 57(2012), 1903-1909. [23] S. J. Hall, W. L. Silver, V. I. Timokhin, and K. E. Hammel, Lignin decomposition is sustained under fluctuating redox conditions in humid tropical forest soils, Global Change Biol. 21(2015), 2818–2828.