Veronika Jílková·Kristyna Dufková·Tomá? Cajthaml

Abstract Temperate coniferous forest soils are considered important sinks of soil organic carbon(C).Fresh C inputs may,however,affect soil microbial activity,leading to increased organic matter decomposition and carbon dioxide production.Litter consists of labile and recalcitrant fractions which are thought to be utilized by distinct microbial communities and at different rates during the growing season.In this study,we incubated the whole litter(LC+RC), the labile (LC) and the recalcitrant (RC)fractions with the coniferous soil at two temperatures representing spring/autumn(10°C)and summer(20°C)for one month.Soil respiration and microbial community composition were regularly determined using phospholipid fatty acids as biomarkers. The LC fraction greatly increased soil respiration at the beginning of the incubation period but this effect was rather short-term.The effect of the RC fraction persisted longer and,together with the LC+RC fraction,respiration increased during the whole incubation period.Decomposition of the RC fraction was more strongly affected by higher temperatures than decomposition of the more labile fractions (LC and LC+RC). However, when we consider the relative increase in soil respiration compared to the dH2O treatment,respiration increased more at a lower temperature,suggesting that available C is more important for microbial metabolism at lower temperatures.Although C was added only once in our study,no changes in microbial community composition were detected,possibly because the microbial community is adapted to relatively low amounts of additional C such as the amounts naturally found in litter.

Keywords Temperate forest·Picea abies·Soil respiration·Hot water-extractable carbon·PLFA(phospholipid fatty acids)

Introduction

Temperate coniferous forest soils are considered important for carbon(C)storage(Lal 2008;Tyrrell et al.2012).Carbon entering the soil mostly originates from the aboveground biomass that is continually shed when needles senesce(Schlesinger and Andrews 2000;Wardle et al.2004),and the amounts of C in the litter usually range from 1.5 to 3 t ha-1y-1in coniferous forests(White 1997).Before the C enters the soil,it is largely transformed by the decomposition community which is expected to greatly differ in the utilization of respective litter fractions(Swift et al.1979).

Litter generally consists of a water-soluble,labile fraction and an insoluble,more recalcitrant fraction;these fractions are assumed to decompose differently based on their availability to decomposers(Berg and McClaugherty 2008).The labile fraction formed by sugars,phenolics and/or amino acids represents a rather marginal percentage(3-5% )of C(Don and Kalbitz 2005;Kammer et al.2012),and is thought to be rapidly depleted after litter fall(Qualls and Haines 1991;Zsolnay and Steindl 1991).The more recalcitrant fraction,consisting of structural tissues formed mostly of cellulose and lignin,is harder to decompose and thus persists longer in the soil(Berg and McClaugherty 2008).

Because of the low pH and high lignin and phenol contents, coniferous forest soils are not a favourable environment for soil fauna and the decomposer community is largely dominated by microorganisms(Persson et al.1980;Brady and Weil 2002).It has been observed that litter inputs into the soil induce shifts in the soil microbial community(Frankland 1998;Manzoni et al.2012),and that these shifts might be associated with preferential utilization of the labile and recalcitrant fractions by distinct components of microbial communities (Paterson et al.2008).Bacteria,which are known to be most effective in labile C utilization(Paterson et al.2008,2011;Koranda et al.2014;Jílkováet al.2018),are assumed to predominate in the decomposition of the labile fraction,while fungi are better equipped for recalcitrant C utilization and predominate in the decomposition of the recalcitrant fraction.Microbial C decomposition has also been shown to be temperature-sensitive(Paul and Clark 1996).This sensitivity greatly depends on substrate availability where decomposition of more complex substrates with higher activation energy increases more at higher temperatures compared to less complex substrates(von Lützow and K?gel-Knabner 2009).

However,not much is known about how the respective coniferous litter fractions influence soil respiration and microbial community composition and how this is affected by temperature during the growing season.This factor might be especially important in view of future climate changes when carbon dioxide(CO2)concentrations and temperatures are expected to increase and cause changes in litter fraction composition and utilization by the soil microbial community(Raich et al.2006;Wang et al.2013).In this study,we incubated the whole litter,the labile and the recalcitrant fractions with the coniferous forest soil at two temperatures representing spring/autumn(10°C)and summer(20°C),based on our previous study(Jílkováet al.2015).Incubation proceeded for one month to reveal the immediate responses of microbial activity on soil respiration and microbial community composition using phospholipid fatty acids (PLFA) as microbial biomarker contents. Moreover, hot water-extractable C(HWC)was used to determine the pool of mineralisable C in the soil(Ghani et al.2003;Cepákováet al.2016).Our hypotheses were that:(1)the addition of C will support microbial activity;(2)the labile fraction will be rapidly depleted and will support bacterial growth;(3)the recalcitrant fraction will persist longer and will support fungal growth;and,(4)microbial decomposition of the recalcitrant fraction will be more temperature-sensitive than that of the labile fraction.

Materials and methods

Soil and litter collection and preparation

Soil and litter were collected from a coniferous temperate forest on the southern slope of Klet’Mountain in Southern Bohemia(Czech Republic)at 700 m a.s.l.in March 2017.The stand was entirely of Norway spruce(Picea abies(L.)H.Karst.).The soil was collected from the mineral A horizon with a soil corer(dia.46 mm)to a depth of 5 cm.Spruce litter was collected with a shovel(ca.100 mL)from the forest floor.Twenty samples of soil and litter were taken over a total area of ca.one ha.The samples were thoroughly mixed to form two composite samples,one for the soil and one for the litter.Fresh soil was sieved through a 2-mm sieve and stored for several days at 4°C until incubation.The litter was dried at 40°C,sieved through a 5-mm sieve to separate twigs and cones,milled to particles ＜1 mm and used for labile C(LC)and recalcitrant C(RC)extraction.The initial properties of both substrates were determined(Table 1).

Extraction of LC and RC fractions

The extraction method of Paterson et al.(2008,2011)was followed.A 1.0 g milled litter was placed into 50-mL centrifuge tubes and 20 mL of deionized H2O(dH2O)were added.The tubes were shaken at 2000 rpm for 30 s and placed in an 80°C water bath for 15 min.The tubes were again shaken at 2000 rpm for 30 s, centrifuged at 3000 rpm for 5 min,the supernatant decanted and retained,and the procedure repeated.The combined supernatant was filtered(0.45 μm)and both the supernatant(LC)and pellet(RC)were freeze-dried.Both fractions as well as unextracted milled litter(LC+RC)were stored at 20°C until the start of incubation.The amount of C in the LC and RC fractions was determined as total organic C(TOC)for the LC fraction after dilution by 5 mL of dH2O and as total C(TC)for the RC fraction.

Table 1 Initial properties of soil and litter

Incubation

An amount of 30 g of the fresh soil was weighed into 100-mL glass vessels and pre-incubated at 10°C and 20°C for 14 days in the dark.The soil was then treated with 5 mL of dH2O,the LC fraction diluted with 5 mL of dH2O(26.3 mg of C added),the RC fraction diluted with 5 mL of dH2O(434 mg of C added),and 1.0 g of the LC+RC fraction diluted with 5 mL of dH2O and(460.3 of C added).Each treatment was represented by eight replicate vessels,giving a total of 64 vessels(four treatments x two temperatures x eight replicates).Three empty vessels were also incubated at each temperature to determine the CO2concentration of the air.All vessels were incubated at 10°C and 20°C for 30 days.As the vessels were not tightly closed to allow for air exchange,evaporated H2O was replenished every week.

During the incubation period,respiration was determined initially and on days 1,3,7,13 and 30.Before a gas sample was collected,the air in the vessel was mixed several times using a 10-mL syringe.A 5-mL volume of gas was then withdrawn using a syringe and stored in a 3-mL evacuated glass tube(Exetainer?,Labco Limited,UK).The gas samples were analysed in the laboratory within 24 h with an HP 5890 gas chromatograph.CO2concentrations were determined with a thermal conductivity detector at 100°C using helium as the carrier.On days 1,7,13 and 30,the soil was destructively harvested(two replicates each time)and subjected to HWC and PLFA analyses.

Substrate analyses

The total C(Ctot)in the dry-crushed initial substrates was analysed using an EA 1108 elemental analyser(Carlo Erba Instruments,UK).The organic matter(OM)content was determined as loss on ignition at 550°C for 5 h.The dry weight was assessed after drying at 105°C for 12 h.For HWC determination,samples were placed in deionized water(1:10 substrate:water)at 80°C for 18 h,and then shaken on a horizontal shaker for 10 min(Sparling et al.1998).HWC suspensions were passed through a paper filter and the filtrate was immediately analysed for C content using a TOC-LCPH/CPNanalyser(Shimadzu).

Fresh samples were freeze-dried and used for PLFA analysis.The phospholipid fatty acids were extracted with a chloroform-methanol-phosphate buffer(1:2:0.8)(?najdr et al.2008).Phospholipids were separated using solidphase extraction cartridges(LiChrolut Si 60,Merck),and the samples subjected to mild alkaline methanolysis.The free methyl esters of the phospholipid fatty acids were analysed by gas chromatography-mass spectrometry(450-GC,240-MS ion trap detector,Varian,Walnut Creek,CA,USA)as previously described(Jílkováet al.2015).The fungal biomass was quantified based on the 18:2ω6,9 content;the bacterial biomass was quantified as the sum of i14:0,i15:0,a15:0,16:1ω7,16:1ω9,10Me-16:0,i16:0,i17:0,a17:0,cy17:0,17:0,10Me-17:0,18:1ω7,10Me-18:0,cy19:0(actinobacteria 10Me-16:0,10Me-17:0,10Me-18:0,G+bacteria i14:0,i15:0,a15:0,i16:0,i17:0,a17:0 and Gbacteria 16:1ω7,16:1ω9,18:1ω7,cy17:0,cy19:0).

Statistical analyses

Data on respired CO2and microbial biomarker contents during incubation were analysed by repeated-measures ANOVA.Total cumulative respired CO2and HWC contents at the end of the incubation period were analysed by a general linear model.If statistically significant,the Tukey post hoc test was used to assess differences between the treatments.If required,the data were log-transformed to maintain homogeneity of variance.The Statistica 12 software(StatSoft Inc.2013)was used for these statistical analyses.

Results

Respiration was significantly affected by the treatments but changed during the incubation period (F=168.1,p ＜0.001)(Fig.1).At the beginning of incubation,respiration was greatly enhanced in the LC and LC+RC treatments(day 0;370% and 410% increase compared to the dH2O(W)treatment,respectively),but the effect of LC decreased rapidly during incubation.The effect of RC persisted longer,and together with LC+RC, became prevalent at the end of the incubation period(day 30;123% and 133% ,respectively).In general,respiration was 2.5 times higher at 20°C than at 10°C(F=1356,p ＜0.001).The effect of treatments was also affected by temperature(F=42.2,p ＜0.001),as there was a short delay in respiration decrease in the LC treatment at 10°C compared to 20°C.Total cumulative respiration also differed between the two temperatures as greater effects of the treatments were found at 10°C than at 20°C(F=88.0,p ＜0.001)(Fig.1).

Fig.1 Cumulative respiration of forest soil at 10°C and 20°C after addition of water(W),labile carbon(LC),recalcitrant carbon(RC),and labile+recalcitrant carbon(LC+RC).The values are the mean±SD.The percentages show a CO2 increase in the respective treatments compared to W treatment(control)

The HWC(hot water-extractable C)content was not only 4% higher at 10°C compared to 20°C(F=13.17,p ＜0.01),but also increased from W to LC to RC and to LC+RC treatments(F=54.30,p ＜0.001)(Fig.2).The microbial biomarker contents differed only slightly among the treatments(Table 2,Fig.3).At the start of incubation,almost no differences were detected in the microbial biomarker contents.Only at the end of the incubation period did the abundance of fungi,Actinobacteria and G-bacteria increase in the LC+RC treatment.However,the microbial biomarker contents differed between temperatures as fungal and bacterial abundance was higher at 10°C than at 20°C.The fungal:bacterial ratio shows that bacteria predominate at both temperatures and in all the treatments.

Discussion

Fresh C input into soils corresponds to the input of available C which can be readily exploited by the soil microbial community compared to the more stable soil organic matter already present in the soil(Kuzyakov et al.2000).Therefore,fresh C input in form of litter increases microbial activity,as shown previously for broadleaf(Kalbitz et al.2007;Kammer et al.2012)and coniferous forests(Crow et al.2009).A similar result was found in our study where all the C fractions in the coniferous litter supported microbial activity. The effect of C input was already apparent after 1 day of incubation in all the treatments,indicating that all the fractions were more readily available to the soil microbial community than the organic matter already present in the soil.These rapid responses to C inputs are consistent with the fact that the microbial biomass is generally C-limited in mineral soil horizons(Wardle 1992;Ekschmitt et al.2005).

Fig.2 Hot water-extractable C(HWC)at the end of the incubation at 10°C and 20°C

Table 2 Results of repeated-measures ANOVA for the effects of treatment(W vs.LC vs.RC vs.LC+RC),temperature(10°C vs.20°C),and their interaction on substrate properties

Fig.3 Microbial biomarker contents in the forest soil during the incubation at 10°C and 20°C for day 1,3,7,and 30

However,the effect of the respective fractions on soil respiration differed in magnitude and persistence.The LC fraction,which is highly labile and mobile,constitutes a readily available source of C for respiration(Marschner and Kalbitz 2003),as experimentally shown for broadleaf and coniferous litter(Joly et al.2016).In our study,the LC fraction caused a higher increase in soil respiration than the RC fraction,but this effect was rather short-term.Respiration in the LC treatment decreased below the respiration in the RC treatment after several days of incubation.No difference in soil respiration could be found between the W treatment and the LC treatment at the end of the 1-month incubation, suggesting that the LC fraction had been depleted by the end of the experiment.This was also supported by the lower HWC content in the LC treatment compared to the RC and LC+RC treatment at the end of the incubation.Paterson et al.(2008,2011)also found rapid mineralisation of the LC fraction compared to the RC fraction of a grass species(Lolium perenne L.).Litter leachates,described here as the LC fraction,are a mixture of labile compounds which are quickly utilized when leached(Qualls and Haines 1991;Zsolnay and Steindl 1991;Joly et al.2016).

Immediately after the addition,the RC fraction caused a smaller increase in soil respiration compared to the LC and LC+RC fractions,indicating that this fraction consists of more recalcitrant compounds and mineralisation may be dependent on the presence of specific extracellular enzymes to breakdown the substrate(Valá?ková et al.2007). However, a microbial community apparently equipped with the necessary enzymes utilized the RC fraction,and the effect persisted longer than for the rapidly depleted LC fraction,leading to a greater increase in respiration in relation to the overall effect on cumulative soil respiration.The reason for this might be that needles have a thick epidermis and hypodermis which impede leaching of the inner tissues,and therefore these available compounds remain in the needle during extraction and could be released only during decomposition when the structural tissues decompose(Don and Kalbitz 2005).However,the LC+RC fraction had the greatest overall effect on soil respiration,as it combines both the readily available and the more recalcitrant compounds.Under natural conditions,leaching of the LC fraction is a continuous process as opposed to the single addition in our experiment,which is supported by respiration in the LC+RC treatment being increased for the whole incubation period compared to the other treatments.

The effect of the respective fractions on soil respiration differed between spring/autumn and summer temperatures.The temperature sensitivity of C decomposition depends on the substrate availability(von Lützow and K?gel-Knabner 2009)and thus differed among the treatments in our study.Recalcitrant C with a higher activation energy has been reported to be more sensitive to increases in temperature than labile C (Karhu et al. 2010; Xu et al. 2014).Accordingly,the total cumulative respiration in the W and RC treatments was increased three fold at 20°C compared to that at 10°C,whereas the LC and LC+RC treatments showed only twice as high respiration.However,when we compare the total cumulative respiration in C-addition treatments with the W treatment,a greater increase in respiration occurs at the lower temperature than at the higher temperature.Available C is apparently important for microbial metabolism when temperatures are low and not yet optimal for microbial activity and growth(Paul and Clark 1996),and thus C input in the form of litter is crucial for increasing microbial activity during the spring and autumn but is less important in the summer,as already inferred in our previous study(Jílkováet al.2018).

A single large addition of C with distinct availability and quality can result in successions and changes within the microbial community(Frankland 1998;Manzoni et al.2012).However,smaller additions might not cause significant community changes and would be processed by organisms already adapted to do this (Paterson et al.2008,2011).The second possibility apparently applies to our study,as almost no changes in the microbial community structure were found during the course of the incubation.The amount of C added as litter in our study(2.3 t C ha-1)equalled the amount of C in the litter shed in coniferous forests per year(1.5 to 3 t C ha-1;White 1997),when we consider the soil surface in incubation vessels(20 cm2),the amount of litter added(1.0 g)and the C content in the litter(460 mg g-1).We also need to consider that not only the fresh litter contributes to the C input in coniferous forest soils,as the old litter is also utilized by the soil microbial community,and thus the addition of the whole-year litter at once still seems to be a rational option.We thus expect that the microbial community present in the soil was adapted to these amounts of C inputs and no adaptations of the community composition were necessary to utilize the added C.Despite there being no differences among the C treatments, the abundances of microbial biomarker contents were greater at 10°C than at 20°C.This again supports the idea that available C is important for microbial metabolism at lower temperatures(Paul and Clark 1996),leading to increases in microbial activity and abundance when available C is supplied.

The length of the incubation period was chosen to detect the immediate responses of the soil microbial community to changes in the C input.We are aware that 1 month is a short time for the recalcitrant C to be decomposed and that the C utilized in the RC fraction most likely originated in relatively available substrates, although it was not extractable from the litter,which then led to the lack of differences in the microbial community composition.A longer incubation period is required to see successional changes in the microbial community during the decomposition of the recalcitrant C.

Conclusion

Litter represents an important source of carbon for coniferous forest soils,which are considered to be C-limited(Wardle 1992;Ekschmitt et al.2005).Respective litter fractions affect soil respiration differently,with the labile fraction being rapidly depleted and the more recalcitrant fraction having a more persistent effect.However,both these fractions greatly increased soil respiration,especially at the lower temperature(10°C),which occurs for most of the year.These changes were detected after only 1 month of incubation.This has further implications in view of future climate changes,when increased net primary production is anticipated,leading to increased plant litter inputs into the soil(Raich et al.2006;Wang et al.2013).

AcknowledgementsThis study was supported by the Czech Academy of Sciences(L200961602;MSM200961606;Otev?ená věda,fellowship No.1.062)and by the European Regional Development Fund-Project‘‘Research of key soil-water ecosystem interactions at the SoWa Research Infrastructure’’(No.CZ.02.1.01/0.0/0.0/16_013/0001782).Part of the equipment used for this study was purchased from the Operational Programme Prague-Competitiveness(Project CZ.2.16/3.1.00/21516).The authors wish to thank Kate?ina Jandováfor total carbon analyses of the initial forest soil and litter and ?árka and Gerrit Angst for helpful comments on the manuscript.