gpo

A.vers

a r t i c l e i n f o

Article history:Received 19 March 2012Received in revised form9 August 2012Accepted 13 August 2012Available online 10 September 2012

Keywords:

Fierer and Schimel, 2002). Maintenance of cell turgor, which isvital to microbial cell growth and survival, is strongly regulated byextracellular water dynamics (Bremer and Krmer, 2000; Schimelet al., 2007). Numerous hypotheses have emerged on the adap-tation strategies that soil microorganisms utilize to cope with

culture have been shown to accumulate inorganic and/or organicosmolytes in their cytoplasm tomaintain cell turgor. The osmolytesinclude K ions and a group of organic solutes like glutamate,proline, peptides, N-acetylated amino acids (amino acids and theirderivatives), sucrose, trehalose (carbohydrates), polyols, glycinebetaine, carnitine (quaternary amines) and tetrahydropyramidineslike ectoines (Killham and Firestone, 1984; Csonka, 1989; Blombergand Adler, 1992; Galinski and Truper, 1994; Kempf and Bremer,1998; Poolman and Glaasker, 1998) rich in C and N. Among these

* Corresponding author. Tel.: 1 540 231 2547; fax: 1 540 231 3083.

Contents lists available at

Soil Biology &

els

Soil Biology & Biochemistry 57 (2013) 644e653E-mail address: markwill@vt.edu (M.A. Williams).1. Introduction

The uctuations in soil water potential caused by episodicdryerewet events in terrestrial ecosystems exert physiologicaland energetic challenges to microbial communities. Extremelylarge uxes of C, important to the global C cycle, have also beenlinked to soil drying and re-wetting in ecosystems (Birch, 1958;

declining and low water potentials. Perhaps the most commonhypothesis, supported by observations of soil C ux and microbialbiomass dynamics, is that microbial cells accumulate and releaseintracellular osmolytes to rapidly respond to water dynamics(Brown, 1976; Harris, 1981; Miller and Wood, 1996; Mikha et al.,2005).

Microorganisms exposed to lowosmotic potentials in laboratoryOsmolytesCompatible solutesMatric water potentialSoilEPSGCeMS0038-0717/$ e see front matter 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.soilbio.2012.08.014a b s t r a c t

Numerous studies have observed the physiological responses of soil microorganisms to water stresscaused by soil drying, however, only a few have attempted to assess the microbial response in soil in situ.An experiment was conducted to analyze the change in extractable metabolites, particularly sugars andamino acids, in soil and the associated microbial community at various intensities of soil desiccation.Water potential was manipulated in two soils, Marietta and Sumter, representing relatively moist anddrought-prone water regimes, respectively. The matric potential of the soils was maintained relativelymoist at 0.03 MPa or lowered to 1.5, 4.5, 10, 20 and 40 MPa by air drying over w3 days. Wehypothesized that microbial communities inhabiting the drought-prone Sumter would accumulate moreosmolytes, and that the soil with a relatively moist water regime, the Marietta, may have communitiesless adaptable to water stress, have fewer osmolytes, and show evidence for greater microbial turnoverand death. However, there was no evidence that the soils responded to drying by accumulating osmo-lytes or that there was greater microbial turnover and death related to soil type. Microbial communitystructure did change with drying, however, with greater fungal-to-bacterial biomass in the Sumter butnot in Marietta soil. A signicant increase of w10e25% in phenol sulfuric acid analyzable sugars (PSA-sugars) at intermediate levels (4.5 MPa) of drying was observed compared to dryer and more moistconditions. However, the GCeMS derived quantities of polyols (glucitol, inositol and xylitol), sugars, andamino acids showed few strong and consistent patterns with level of desiccation. These results providesome of the rst evidence that microbial communities in soil in situ do not strongly rely on these basicosmolytes to cope with typical soil water decits. In natural soils, we propose that microbial commu-nities respond differently to water decits perhaps through re-allocation of C to cell wall mucilage,exopolysaccharides (EPS), and phospholipids, than organisms in culture, perhaps a consequence of lowenergy and limiting supplies of N.

2012 Elsevier Ltd. All rights reserved.bNatural Products Utilization Research, University of Mississippi, 38677, USAMicrobial community response to varyinA test of the osmolyte accumulation hy

Madhavi L. Kakumanu a, Charles L. Cantrell b, MarkaRhizosphere and Soil Microbial Ecology Laboratory, Virginia Polytechnic and State Uni

journal homepage: www.All rights reserved.magnitudes of desiccation in soil:thesis

Williams a,*

ity, 301 Latham Hall, Blacksburg, VA 24060, USA

SciVerse ScienceDirect

Biochemistry

evier .com/locate/soi lbio

y &organic solutes, fungi tend to accumulate polyols whereas bacteriautilize amino acids and sugars to cope with water decit (Brown,1976). When exposed to hypo-osmotic conditions, microorgan-isms expel the accumulated solutes extracellularly to maintainequilibrium (Tschichholz and Trper, 1990; Halverson et al., 2000).

The sudden ush of C and N mineralization following therewetting of dry soil has been reported by many researchers andmore recently was hypothesized as representing microbial releaseof intracellular osmolytes (Birch, 1958; Sorensen, 1974; Schimelet al., 1999; Franzluebbers et al., 2000; Chowdhury et al., 2011).Although isotopic studies have revealed that at least part of the Creleased during the short-term pulse following the rewetting of drysoil is microbial in origin (Kieft et al., 1987; Van Gestel et al., 1992;Magid et al., 1999), it is unclear whether microbial cell lysis, theregulated expulsion of intracellular microbial osmolytes, or othermechanisms are responsible for the pulse.

The response of soil microorganisms to low water potential hasbeen studied by exposing soil isolates to salt induced hyperosmoticstress and through desiccation on a simulated soil matrix (Killhamand Firestone, 1984, 1984b; Schimel et al., 1989; Roberson andFirestone, 1992; Halverson et al., 2000). Several of these studiesobserved increased levels of amino acids, sugars and the accumu-lation of extracellular polysaccharides in response to decliningwater potential. Enhanced pools of cytoplasmic C and N contentwith water potential decit were also reported. However, to ourknowledge no study has attempted to directly measure the pool ofsoil microbial compatible solutes in response to drying under in situsoil conditions. Laboratory cultures are so vastly different from soilconditions that studies under these latter scenarios are greatlyneeded to understandmicrobial adaptations towater stress and thecontrols of soil C and nutrient dynamics in ecosystems.

An experiment was designed to test the physiological andstructural response of soil microbial communities to drying acrossa water decit gradient using two soils that developed withincontrasting water regimes. The rst and primary objective was todetermine if microorganisms adapt to soil drying by accumulatingcompatible solutes. It was hypothesized that as the soils were dried,greater amounts of organic osmolytes would be detected. We ex-pected an increase in fungal polyols, and bacterial derived sugarsand amino acids in response to water decit. The second objectivewas to assess whether microbial communities in two differentsoil types with contrasting water regime histories would responddifferently to drying and whether changes in fungal to bacterialratios would coincide with the composition of fungal and bacterialosmolyte pools. It was hypothesized that microbial communitiesinhabiting the drought-prone soil would accumulate more osmo-lytes, and that the soil with a relatively moist water regime mayhave communities less adaptable to water stress, have fewerosmolytes, and show evidence for greater microbial turnover anddeath.

2. Materials and methods

2.1. Site description

The experiment was conducted on two soils, the wetter lowlandMarietta and dryer upland Sumter series located near MississippiState University, Mississippi, USA (33 280 N and 088 470 W) in Fall2009. These soils have been previously sampled across a 5 countyarea in northern Mississippi and were shown to have similar pH,texture, color, and hue as predicted by the soil type. For the purposesof the work in this paper, a 100-m transect across a 2e5 Ha forestwas used within each of the respective soil types to collect soil toa 10-cm depth. Five-kg of soil was collected every 20-m fromwithin

M.L. Kakumanu et al. / Soil Biologa central location of each forest. At sampling, the soils wererelatively moist with water content of 34e36% (w150 kPa). Thesoils were sieved to 4 mm and thoroughly cleaned of obvious plantlitter and rocks, and refrigerated at 4 C until use.

Rainfall across the area, for both soil types, averages 51 inchesand the mean annual temperature is 17.7 C. The Marietta soils are(ne-loamy, siliceous, active, thermic Fluvaquentic Eutrudepts)derived from deep alluvial deposits near streams in the blacklandprairie region of Mississippi. These soils drain areas of the mixeduplands of the Southern Coastal Plain. The soils are subject tofrequent ooding. Mottles and stains starting at the depth of 10-cmand shallow water table are indicators of the generally moist waterstatus of these soils. The site was forested with >50-y old decid-uous vegetation dominated by Carya illinoinensis. The C:N contentof the Marietta soils was 2.35% and 0.17% respectively with pH of6.2. The Sumter soil is silty clay, with medium granular structure,moderately deep, well drained, with rapid runoff. These uplandsoils were formed in marly clays and chalk of the blackland prairies(ne-silty, carbonatic, thermic Rendollic Eutrudepts) and drain tolowlands (e.g. Marietta). The water table is deep and the perme-ability of the soil is slow. The soil has a pH of 6.3 with C and Ncontent of 2.56 and 0.15% respectively. Total soil organic C and Ncontents were measured on a Vario MAX CNS macro elementalanalyzer (Elementar Americas, Inc., Mt. Laurel, NJ). Soil pH wasmeasured after shaking the soil with 0.01 M CaCl2 (1:1, mass:volume) suspension for 30 min.

2.2. Experimental setup

A laboratory experiment was conducted to study the physio-logical response of soil microbial communities to water potentialdecits (matric) caused by air drying. Keeping in view of thetextural differences in the soil types and their water holdingcapacities, we used water potential to measure the drying effecton the soil microbial communities. Each treatment (soil, n 2;water potential, n 6) was replicated three times, thus resultingin a total of 36 separate samples. Each sample consisted of 10 g(dry weight) of well homogenized soil weighed into 150 mlvolume specimen cups. This experimental setup was repeated 5to have enough material for all the analysis (e.g. Phospholipidfatty acids (PLFA), osmolytes, respiration, microbial C, N andsoluble C). Keeping the soil mass low within each specimen cupalso allowed for control and homogeneity of the soil dryingprocess. The water content of all the samples was adjusted to theirrespective eld capacities (0.03 MPa) by adding sterile distilledwater. Soils were pre-incubated at room temperature (22 C)for ve days to reduce disturbance effects related to sampling,sieving, and storage.

The pre-incubated soils were either maintained moist(0.03 MPa) or slowly air dried to ve different water potentialsof 1.5, 4.5, 10, 20, and 40 MPa at 22 C. For three consec-utive days, soils were dried for 6e12 h each day until reachingtarget water potential. The soils took approximately 16, 22, 29, 33and 36 drying hours to reach the water potentials of 1.5 MPa,4.5 MPa, 10 MPa, 20 MPa, 40 MPa, respectively. The relationbetween the soil water content and water potential were analyzedprior to the experiment bylter papermethod (McInnes et al.,1994).As well, the water potentials of the soils were cross-validated andmonitored using a WP4 dewpoint potentiameter by (Decagon, Inc.,Pullman, WA).

The soils remained at respective water potentials for 24 h beforeany further analysis. PLFA analysis was done to understand thestructural and physiological changes in the community withdrying. The respective soil samples were stored at 80 C untilanalysis. All the other extractions except for the PLFAs were done

Biochemistry 57 (2013) 644e653 645the following day. Microbial osmolytes are known to comprise

y &amino acids (e.g. proline, glutamic acid), sugars (e.g. sucrose,trehalose) and polyols (e.g. sorbitol, mannitol). The physiologicaland biomass responses of microbial communities were assessed byninhydrin analysis (amino acids) and phenol sulfuric acid (PSA)analysis for sugars. Metabolite composition was analyzed usingGCeMS.

2.3. Extraction of metabolites from soil and associatedmicroorganisms

Microbial metabolites from soil were extracted using a mixtureof chloroform and 0.01 M K2SO4 (1:4; v/v). Soluble organics wereextracted using 0.01 M K2SO4 only. The principle behind theextraction of microbial metabolites was based on chloroform lysesof microbial cells and the consequent release and dissolution ofintracellular cytosol into the 0.01 M K2SO4. We utilized the chlo-roform slurry method for extraction of microbial metabolitesinstead of a traditional fumigationeextraction assay for tworeasons. First, fumigation of dried soils has been shown to giveerratic results (Sparling and West, 1989) that would then make itdifcult to compare between drying treatments. Second, theactivity of hydrolytic enzymes on proteins and polysaccharides insoils during fumigation (Renella et al., 2002) would result in thecontamination of the cytosolic pool of organics and exaggerateconcentrations of sugars and amino acids.

For microbial metabolite extraction, 10 ml chloroform wasadded to 10 g (dry weight) of soil that was transferred to separate160 ml serum bottle. After one minute, the K2SO4 solution wasadded to bring each bottle up to 40 ml of 0.01 M K2SO4 andvigorously agitated for 2 h (250 rpm) on an orbital shaker. Solubleorganics were extracted identically, but without chloroform. Toseparate the chloroform and aqueous phase, the samples wereallowed to settle for 15 min and then centrifuged at 1500 rpm for10 min on IEC bucket centrifuge. The aqueous supernatant wastaken and ltered through whatman #1 lter paper and thesolution was lyophilized and stored at 80 C until furtheranalysis.

For measuring microbial and soluble C and N concentrations,soil samples were extracted, ltered, and analyzed on a ShimadzuTOC analyzer. Microbial C and N were determined by subtractingsoluble C and N values from chloroform-K2SO4 extracts. Unlike thefumigation method (Vance et al., 1987) no correction factor wasapplied to the values.

2.4. Analysis of soil extracts by colorimetric methods

Reducing sugars in the soil extracts were analyzed by the PSAmethod (Dubois et al., 1956). Briey, 50 ml of 80% phenol solutionfollowed by addition 5 ml of concentrated H2SO4 (w18 M) solutionwas added to a small volume of soil extract. The mixture wasallowed to stand at room temperature for 45 min. The absorbancewas measured at 480 nm on UV spectrometer. Absorbance valueswere used to calculate the concentration of reducing sugars basedon an 8-point standard curve of glucose.

Amino acids were determined using the ninhydrin method(Stevenson, 1982). Though ninhydrin also reacts with peptides,proteins, ammonium and other compounds with free a-aminogroups it is generally considered sensitive and useful for quanti-cation of amino acids (Jorgenson and Brooks, 1990). The soilextracts along with 0.5 ml of citric acid and 2 ml of ninhydrinreagent were incubated at 100 C for 25 min. The solution wascooled, diluted with 5 ml of 50% ethanol, and the absorbance wasmeasured at 570 nm on UV spectrometer. A standard curve wasmade by measuring the absorbance at different concentrations of

M.L. Kakumanu et al. / Soil Biolog646L-Leucine-N.2.5. Analysis of extractable metabolites by Gas ChromatographyeMass Spectroscopy

The sugars and amino acids in the extracts were characterizedusing GCeMS. Prior to analysis the samples were derivatized.Derivatization was carried out in 1 ml reaction vials treated with5% dimethyldichlorosilane in toluene. For analysis of sugars, thesoil extracts were derivatized by N, O-bis-(trimethylsilyl) trifur-oacetamide (BSTFA) solution. Approximately 250 ml of the aliquotsof soil extract were taken in silylated reaction vials and driedcompletely under nitrogen. The dried residues were converted totrimethylsilyl derivatives by adding BSTFA containing 1% trime-thylchlorosilane (TMCS) and pyridine in 2:1 ratio and incubatingthem for 3 h at 70 C (Medeiros et al., 2006). The samples wereallowed to stay overnight at room temperature and completelydried under pure nitrogen. The dried residues were redissolved in110 ml of hexane and sent for GC analysis.

Derivatization of amino acids was done as per the method givenby Fan (1996). Approximately 500 ml of soil extracts were taken andthe pH was lowered to 2 by adding equal volumes 1 M HCl inreaction vials. The solution was dried completely under purenitrogen. Dried extracts were sonicated with 1:1 mixture of N-Methyl-N-(Tert-Butyldimethylsilyl) triuoroacetamide (MTBSTFA)and acetonitrile for 2 h at 60 C. The solution was left at roomtemperature overnight and dried under nitrogen. The derivatizedextracts were redissolved in 110 ml of hexane before GC analysis.

The samples were analyzed on Varian CP-3800 Gas Chromato-graph coupled to a Varian Saturn 2000 MS/MS. The GC was equip-pedwith a DB-5 fused silica capillary column (30m 0.25mm,withlm thickness of 0.25 mm) operated using the following conditions:injector temperature, 240 C, column temperature, 60e280 C at8 C/min then held at 280 C for 5 min; carrier gas, He; injectionvolume, 1 mL (splitless). The MS mass ranged from 40 to 650 m/z,lament delay of 3 min, target TIC of 20,000, a prescan ionizationtime of 100 ms, an ion trap temperature of 150 C, manifoldtemperature of 60 C, and a transfer line temperature of 170 C.

Individual sugars were identied by comparison of mass spectrawith literature and library data, comparison of mass spectra and GCretention times with those authentic standards and/or interpreta-tion of mass spectrometric fragmentation patterns. Standard solu-tions of glucose and other compounds like trehalose, sorbitol,sucrose, proline, glutamine which are commonly expected asmicrobial osmolytes, were analyzed (Williams and Xia, 2009).Compounds were quantied using total ion current (TIC) peak areaand converted to compound mass using calibration curves of theexternal standards (glucose for monosaccharides, sorbitol for sugaralcohols and sucrose for disaccharides).

2.6. PLFA analysis

Total lipids were extracted according to the procedure of Whiteand Ringelberg (1998) as modied by Butler et al. (2003). Briey,w10 g (dry weight) of soil samples from 3 replications of all thetreatments were transferred to sterilized 160 ml serum bottles. Thesoils were extracted overnight using amixture of 50mMphosphatebuffer (pH 7.1), chloroform and methanol (0.8:1:2). The sampleswere centrifuged the following day at 1000 rpm for 5 min andltered usingWhatman # 1 lter paper. The ltrate was added with3 M NaCl solution and a pinch of Na2SO4 salt and allowed to standforw8 h for phase separation. The chloroform phase was collectedinto separate glass tubes and dried completely under stream ofnitrogen. Dried lipids were redissolved and fractionated intoneutral, glyco- and phospholipids using silicic acid bonded phaseextraction columns (Supelco, cat. No. 505048). The neutral and

Biochemistry 57 (2013) 644e653glyco lipids were eluted using chloroform and acetone respectively.

PS

A an

alyzab

le C

(

g g

-1

so

il)

80

100

120

140

160

Marietta

Sumter

Water Potential (MPa)

-0.03 -1.5 -4.5 -10 -20 -40

To

tal P

LF

A- C

(

g g

-1

so

il)

16

20

24

28

32

36

a

b

*

*

**

*

*

*

Fig. 1. Concentrations of (a) reducing sugar carbon (mg g1soil) in chloroformeK2SO4extracts as determined by the phenol sulfuric acid (PSA) method and (b) total PLFAcarbon (nmol g1soil) at different levels of water potentials in two soils (Marietta andSumter). Total PLFA-C was quantied using the detector response to 16:0 methyl esterto calculate the concentrations of all fatty acids. Error bars represent the mean stan-dard error between the replicates. Asterisks (*) denote signicant difference betweenthe corresponding drying treatment and the control (0.03 MPa) within each soil(n 3; P < 0.05). Differences between soils were signicantly different (n 3;P < 0.05).

Water Potential (MPa)

-0.03 -1.5 -4.5 -10 -20 -40

Fu

ng

al/B

acterial R

atio

0.14

0.16

0.18

0.20

0.22

0.24

0.26

0.28

0.30

Marietta

Sumter

**

**

Fig. 2. The changes in fungal/bacterial ratio across the water potential gradient in twosoils (Marietta and Sumter). Error bars represent the standard error between thereplicates. Asterisks (*) denote signicant difference between the drying treatments

y & Biochemistry 57 (2013) 644e653 647Phospholipids were nally eluted with methanol into separatetubes and completely dried under nitrogen. The dried phospho-lipids fractionwas transformed into fatty acids methyl esters underalkaline conditions and extracted twice in 1:4 chloroformehexanesolution. The chloroformehexane mixture was completely evapo-rated under stream of ultra high purity 99.999% N2 gas and theresidue was resuspended in 500 ml of hexane for GC analysis.

The fatty acids were quantied and detected on Agilent 6890Series gas chromatograph (Santa Clara, CA) equipped with a ameionization detector, an Ultra-2 column (19091B-102; 0.2 mm by25 m), and controlled by a computer loaded with ChemStation andSherlock software. Ultra high purity H2 was the carrier gas ata column head pressure of 20 kPa, septum purge of 5 ml min1,a split ratio of 40:1, injection temperature of 300 C, injectionvolume of 2 ml. The oven temperature ramps from 170 C to 288 Cat 28 C min-1 and the analysis time of each sample was 6 min.Peak identication was carried out by the Microbial IdenticationSystem (MIDI, Inc.) following calibrationwith a standardmixture of17 fatty acid methyl esters (1300A calibration mix).

The PLFAs i15:0, a15:0, i16:0, a16:0, i17:0, a17:0 (gram-positive),16:1u9, 16:1u7, 18:1u7 and cy19:0 (gram-negative) were consid-ered as bacterial biomarkers, 10:Me 16:0 and 10:Me 18:0 for acti-nomycetes and 18:1u9 and 18:2u6 as fungal biomarkers (Frostegrdand Bth, 1996; Zhang et al., 2005; Liang et al., 2008). The ratio offungal to bacteria biomarker fatty acids were used to indicatechange in the fungal to bacterial biomass ratio (Bossio et al., 1998).

2.7. Statistical analysis

Two-way ANOVA was done on all the data to test for effects ofsoil type, water potential and their interaction. However, irre-spective of the interaction effects we looked at each factor sepa-rately using one-way ANOVA (Randomized complete block designwith factors: soil type, water potential and 3 replications) and thetreatments were considered signicant at p< 0.05 (Microsoft Excel,2007; SAS 9.2, Sysstat software, 2008). Data were transformedwhen necessary to meet the assumptions of normality. We usedPC-ORD version 4 software (MJM Software, Gleneden Beach, OR)for ordination and multivariate analysis of the data as per McCuneand Grace (2002).

3. Results

3.1. Quantication of sugars in microbial extracts

Phenol sulfuric acid analyzable sugars (PSA-sugars) in the(chloroform labile) soil extracts indicated that Marietta had signif-icantly greater (w30%) amounts of PSA-sugars than Sumter soil(Fig. 1a). Moreover, PSA concentrations in the two soils respondedsimilarly towater potential decit. For instance, inMarietta soil PSA-sugars increased signicantly (w20e25%) up to 10 MPa, whileSumter had w15% greater PSA-sugar in moderately (4.5 MPa)dry soil comparedwith the amounts inmoist and extremely dry soil.

3.2. PLFAs and fungal to bacterial ratio

The overall abundance of PLFA in Marietta compared to Sumter(Fig. 1b) were in agreement with differences in the size of the PSA-sugar pool, however, the general response to drying tended tocontrast with the results of PSA-sugar. Indeed, Marietta tended toshow declines and Sumter showed signicant increases in PLFA dueto drying.

There was no signicant effect of water potential decit on thefungal to bacterial ratio in Marietta; however, there was a signi-

M.L. Kakumanu et al. / Soil Biologcant increase with drying in the Sumter soil (Fig. 2). Sumter hadand the control (0.03 MPa) within each soil (n 3; P < 0.05). Differences betweensoils were signicantly different (n 3; P < 0.05).

signicantly greater fungal to bacterial ratios thanMarietta soil. Theincrease in fungal to bacterial ratio with soil drying is consistentwith higher microbial C:N ratio at dry treatments in Sumter. Therelative abundance of PLFA biomarkers indicative of gram-positivebacteria and actinomycetes showed decreases with water decitwhereas fungal biomarkers increased across the water potentialgradient of the two soils (Fig. 3).

3.3. Microbial C, N, and soluble C

Microbial C and Nwere both signicantly greater in theMariettathan Sumter. In general, microbial biomass C was not affectedby drying (Fig. 4), however, there was an increase at 20 MPacompared to 0.03 MPa in Sumter. Microbial N in Sumter soildecreased when dried to40MPa but otherwise was unaffected bydrying. In the Marietta soil, in contrast, microbial N tended todecrease duringmoderate drying compared tomoist and extremelydry soil. Sumter had greater levels of soluble C compared to Mar-ietta soil. Soluble C concentrations across the drying gradientchanged similarly for both soil types (Fig. 5) increasing signicantlyat the dryer end of the gradient. Microbial C:N ratios tended tomaintain values ofw12, however, when soils dried to the two drieststates, ratios increased considerably in the drought-prone Sumterbut not themesicMarietta soil (Fig. 6). Our soil respiration data (notshown) indicated a very typical response to drying and rewettingwhereby the respiration declined several fold between moist and

n 5

10

15

20

-1.5

-4.5

-10

-20

-40

a Marietta

-5

* *

*

0

60

80

100

120

140

Marietta

Sumter

Water Potential (MPa)

-0.03 -1.5 -4.5 -10 -20 -40

M

icro

bial b

io

mass carb

on

an

d n

itro

gen

(

g g

-1

so

il)

0

5

10

15

20

*

Carbon

b

a

Nitrogen

* *

*

Fig. 4. The amount of microbial biomass (a) carbon (b) nitrogen extracted from twosoils (Marietta and Sumter), at different levels of water potential. Microbial C and Nwas calculated as (C and N in choloroformeK2SO4 extracts) minus (C and N in solubleextracts). No conversion factor was applied. Error bars represent the standard errorbetween the replicates. Asterisks (*) denote signicant difference between the dryingtreatments and the control (0.03 MPa) treatment within each soil (n 3; P < 0.05).Differences between soils were signicantly different (n 3; P < 0.05).

*

*

*

Water Potential (MPa)

-0.03 -1.5 -4.5 -10 -20 -40

So

lu

ble

C

arb

on

(

g g

-1

so

il)

10

15

20

25

Marietta

Sumter

*

*

*

Fig. 5. The amount of soluble carbon extracted from two soils (Marietta and Sumter),at different levels of water potentials. Error bars represent the standard error betweenthe replicates. Asterisks (*) denote signicant difference between the drying treat-

M.L. Kakumanu et al. / Soil Biology & Biochemistry 57 (2013) 644e653648-10 Gram- Gram- Actinomycetes Fungi

Positive Negative

Fig. 3. Percent variation in the relative abundance of PLFA biomarkers (indicative ofbacteria and fungi) across the water potential gradient of 1.5, 4.5, 10, 20and 40 MPa in a) Marietta b) Sumter soils. The bars below and above the zero%V

ariatio

n in

m

icro

bial co

mp

ositio

-10

-5

0

0

5

10

15

b Sumterindicate the percent decrease and increase in the biomarkers at respective waterpotentials relative to moist control (0.03 MPa).

ments and the control (0.03 MPa) treatment within each soil (n 3; P < 0.05).Differences between soils were signicantly different (n 3; P < 0.05).

sugars (monosaccharides and disaccharides) and polyols. Thecompounds that were conrmed include glucose, galactose, trihy-droxy butyric acid, arabinose, glycerol, glucitol, xylitol, inositol,myo-inositol, turanose and sucrose. The total sugars detected byGCeMS varied from approximately 20 mg g1soil to 120 mg g1soil,an amount w10e50% the size of the total pool of sugars (Table 1).

The changes in the concentration of glucose, polyols and othersaccharides at different intensities of matric stress in Mariettaand Sumter soil are shown in Fig. 8. Glucose was the mostabundant monosaccharide found in all treatments and accountedfor 45e60% of total GCeMS metabolites. Compounds like glyc-erol, galactose, glucose, glucitol, myo-inositol and turanose werealso found in all treatments. However, certain compounds werenot consistently present along the moisture regime. For example,inositol was found only in Marietta soil and not Sumter. Similarly,

Water Potential (MPa)

-0.03 -1.5 -4.5 -10 -20 -40

C/N

ra

tio

o

f m

ic

ro

bia

l

bio

ma

ss

0

10

20

30

40

Marietta

Sumter

*

*

Fig. 6. The C:N ratio of microbial extracts across the matric water potential gradient intwo soils (Marietta and Sumter). Error bars represent the standard error between thereplicates. Asterisks (*) denote signicant difference between the drying treatmentsand the control (0.03 MPa) treatment within each soil (n 3; P < 0.05). Differences

M.L. Kakumanu et al. / Soil Biology & Biochemistry 57 (2013) 644e653 649very dry soil and a 20e50% increase in CO2 upon rewettingcompared to continuously moist soil.

3.4. Ninhydrin reactive nitrogen (NRN)

Ninhydrin reactive-N (NRN) in chloroformeK2SO4 extracts atvarious intensities of matric stress showed different dynamics inMarietta and Sumter soil (Fig. 7). The initial concentration of NRN inMarietta (7.46 mg g1 soil) was considerably greater than in Sumtersoil (4.62 mg g1), resembling the relative differences observed forthe PSA-sugars. Moreover, the amounts of NRN declined signi-cantly in Marietta but not in the Sumter soil as a result of soildrying.

3.5. GCeMS characterization of soil microbial extracts

The GCeMS analysis of trimethylsilyl (TMS) derivatives of

between soils were signicantly different except at 4.5 MPa (n 3; P < 0.05).chloroform labile soil extracts showed the presence of several

Water Potential (MPa)

-0.03 -1.5 -4.5 -10 -20 -40

Nin

hyd

rin

R

eactive N

(

g g

-1

so

il)

3

4

5

6

7

8

9

10

Marietta

Sumter

*

* * *

*

Fig. 7. Concentrations (mg g1 soil) of Ninhydrin Reactive Nitrogen (NRN) by chloro-formeK2SO4 extracts at different water potentials in (Marietta and Sumter) soil. Errorbar represents the mean standard error between the replicates and Asterisks (*) denotesignicant difference between the corresponding drying treatment and the control(0.03 MPa) within each soil (n 3; P < 0.05).sugars like arabinose, fructose and polyols like xylitol werefound in detectable amounts only in some replications of driedtreatments.

The composition and the relative abundance of the saccharidesand polyols vary with the soil type (Fig. 8). In this regard, Sumterhad signicantly greater concentrations of cellular metabolites.In general, drying tended to reduce the metabolite content inSumter soil and tended to change little in the Marietta soil exceptfor an increase at intermediate levels of drying (4.5 MPa).Metabolites were thus a greater proportion of total biomass inSumter compared to Marietta soil.

Amino acids, the other important group of osmolytes werefound only in low concentrations in continuously moist treatmentsbut were not detected in dry soils. We identied the amino acidsvaline, proline, leucine, isoleucine, glutamine, glutamic acids, andsome fatty acids in continuously moist treatments. Both NRN andspectroscopic NMR analysis supported the relatively low abun-dance of amino acids in the extracts (data not shown).

4. Discussion

The primary aim of the research was to test the microbialosmolyte accumulation hypothesis (OAH) in soil. We hypothesizedthat microbial communities inhabiting the drought-prone soilwould accumulate more osmolytes, and that the soil with a rela-tively moist water regime may have communities less adaptable towater stress, have fewer osmolytes, and show evidence for greatermicrobial turnover and death. Sugars, polyols and amino acids havebeen reported as common osmolytes that accumulate withinmicroorganisms; however, these reports are overwhelmingly basedon microbial responses to salt stress when grown in cultures(Galinski and Truper, 1994; Kempf and Bremer, 1998; Poolman andGlaasker, 1998). Generally speaking, the results from this study did

Table 1Total amount of sugars (mg g1 soil) identied in GCeMS analysis in Marietta andSumter soils at different water potentials.

Treatment Marietta Sumter

Sugars(mg g1 soil)

% Detecteda Sugars(mg g1 soil)

% Detecteda

0.03 34.3 (15.9) 14.3 (6.8) 93.5 (13.8) 50.5 (9.6)1.5 37.8 (23.1) 15.2 (9.6) 46.2 (10.6) 26.6 (5.4)4.5 69.4 (16.3) 15.0 (5.1) 46.1 (6.9) 21.1 (2.61)10 37.7 (4.1) 12.2 (1.2) 58.0 (18.5) 30.4 (10.2)20 21.4 (1.1) 7.7 (2.7) 39.5 (14.8) 23.2 (8.8)40 21.2 (7.5) 7.7 (2.7) 27.0 (3.4) 16.4 (2.0)

Values are means and the gures in the parenthesis indicate the standard error within the treatments.

a Amount of sugars detected in GCMS% detected Sugars detected in PSA analysis

100

-1

s

y &0

10

g

m

etab

olites g

20

30

40

50

60 b Sumter

*

*20

30

40

50

60

Glucose

Polyols

Other sugars

oil

a Marietta

M.L. Kakumanu et al. / Soil Biolog650not support the OAH as a primary mechanism for microbial adap-tation to soil drying. Specic pools of metabolites such as sugarsand alcohols did not increase consistently in soil microorganismsdue to drying. However, two other insights are important to note.First, the pool of microbial metabolites was relatively large,accounting for a large proportion (20e50%) of the total microbialbiomass C (not including the use of k-factors to estimatemicrobial biomass) and suggesting that these molecules area physiologically important pool of high energy compounds.Second, there was considerable change in the pools of microbialderived PLFA, microbial C, and PSA-sugars that may be importantindications of community structure and intracellular re-allocationsor C resulting from microbial adaptation to soil drying.

4.1. Metabolite composition

The composition of the metabolite pool was consistent with thetypes of compounds expected to allow for microbial cytoplasmicadjustment to water potential decits (Brown and Simpson, 1972;Hocking, 1986; Al-Hamdani and Cooke, 1987; Kelly and Budd, 1991;Witteveen and Visser, 1995; Shen et al., 1999). Yet, it is important toalso note that these compounds are very common components ofcellular metabolism (Fig. 8) and energy cycling. The amounts ofmetabolites, if utilized for physiological adaptation to water decit,would have increased and have occurred in near molar concen-trations within microbial cells due to desiccation, thus making their

Water potential (MPa)

0

10

-0.03 -1.5 -4.5 -10 -20 -40

Fig. 8. The amount of metabolite (mg g1 soil) pools detected by GCeMS at different ofwater potentials in (a) Marietta and (b) Sumter soils. The metabolites include glucose,polyols (sum of glycerol, xylitol, glucitol, inositol, myo-inositol) and others (sum oftrihydroxy butyric acid, galactose, turanose). Error bar represents the mean standarderrors (n 3).detection straightforward (Yancey et al., 1982; Schimel et al., 1989).Overall, the results suggest that microorganisms are not accumu-lating large concentrations of typical osmolytes during soil drying(Fig. 8), and thus do not support OAH as a primary mechanism ofmicrobial adaptation to soil drying.

It was expected that drought-prone Sumter soils wouldrespond differently to drying than the more mesic Marietta soil. Itwas therefore interesting that the Sumter soil which containedhalf the microbial biomass had 2 the concentration of metabo-lites relative to the Marietta even when soils were moist (Fig. 8).The microbial metabolite response to drying was also differentbetween the two soils but was inconsistent with themain idea thatthe drought-prone Sumter would accumulate more osmolyteswith drying and that the Marietta soil would show more evidenceof microbial death and lysis (Fig. 1a,b; Fig. 4). Indeed, related to theissue of microbial turnover, bulk pools of PLFA and PSA did notchange in ways consistent with microbial lysis, and it is thusconcluded that microbial lysis in soils did not result during soildrying.

If soil microbial communities are not primarily reliant uponosmolyte accumulation, but are nevertheless resistant and adapt-able to drying, then other mechanisms are needed to explain soilmicrobial responses to desiccation. Rather than having unlimitedresources like that typically found during early microbial growth inculture, soil microorganisms under nutrient limited and oligotro-phic energy conditions may not be able to physiologically afford thecosts of intracellular accumulation of osmolytes. In contrast, moreefcient utilization (Tiemann and Billings, 2011) and re-allocationof resources from within microbial cells may be a potentiallyimportant mechanism for efciently coping with water decit insoils. For example, the production of exopolysaccharides (EPS) hasbeen shown to be an important mechanism for bacterial adaptationto desiccation and other environmental stresses (Roberson andFirestone, 1992).

4.2. Microbial pools of PLFA, sugars & NRN

The PSA-sugar concentrations along the drying gradient tendedto increase in both soil types during moderate levels of drying(Fig. 1a). The PSA methodology measures both monomeric andoligosaccharide sugars and thus is not a specic determinate ofmonomer-type sugars that have so often been shown to play a rolein osmolyte accumulation (Safarik and Santruckova, 1992; Gallardoand Schlesinger, 1995). It is clear, however, that microbial cells areadapting to soil drying either by increasing microbial biomass orspecic pools of cell biomass, a result that is often reported to occurduring desiccation in soil (e.g. Lundquist et al., 1999; Williams andXia, 2009; Schimel et al., 2010). Similarly, it is notable that thedroughty Sumter soil had increased levels of NRN at moderatelevels of drying and that the Marietta soil showed signicantdeclines in NRN as result of drying. The ninhydrin method,however, much like the PSAmethodmust be interpreted cautiouslybecause it is not specic to known osmloytes, and is also sensitiveto peptides, proteins, and ammonium, for example. The concen-trations of NRN are also rather low and the changes due to dryingwere relatively small compared to the size of the water potentialdecit (Fig. 7). While the greater PSA and NRN concentrations atintermediate levels of drying could be construed as consistent witha microbial response to desiccation, the small changes observedwould explain only a small part of the response needed to conrma general OAH hypothesis (Killham and Firestone, 1984; Schimelet al., 2007).

The results presented here contrast with the 11e60% increase inN and amino acids by cultivated soil microorganisms (Killham and

Biochemistry 57 (2013) 644e653Firestone, 1984b; Schimel et al., 1989) in response to matric decit.

y &However, unlike that of most culture conditions, C and N arelimiting nutrients for microbial biomass production in most non-agricultural soils and may help to explain the small pool ofamino-type N found in our study. Moreover, under dry conditionsthe thinning of water lms and the discontinuity between lms canresult in substantial reductions in substrate diffusion (Schimelet al., 1989) and nutrient availability to microorganisms. Undersuch conditions the limitations in nutrient (e.g. nitrogen) andenergy availability could lower the capacity for an organism toproduce/accumulate appropriate concentrations of osmolytes tocounterbalance cellular water loss (Stark and Firestone, 1995).Could microorganisms in C and nutrient limited soil environmentshave evolved different mechanisms to cope with soil drying (andre-wetting) than mechanisms observed to occur by microbes innutrient-rich laboratory culture (Tschichholz and Trper, 1990;Schleyer et al., 1993)?

Microbial respiration rates in this study (data not shown)declined in a typical manner with degree of drying and followingthe re-wetting of soil (e.g. Birch, 1958; Williams and Xia, 2009). Ifother microbial responses are similarly tied to desiccation level,then it might be expected that intracellular osmolyte moleculeconcentrations might show similar patterns of change to desicca-tion intensity. Indeed, the ush of soluble C (Fig. 5) releasedfollowing re-wetting of soil has been widely hypothesized torepresent the intracellular release of microbial osmolytes and alsotends to correlate well with the degree of drying (Schimel et al.,1999; Williams and Xia, 2009; Chowdhury et al., 2011), however,the chemical composition of this pool has not been observed tocontain high concentrations of putative osmolytes (Williams andXia, 2009).

The increase in PSA-sugars and NRN in the moderately drycompared to both dry and moist soils may be interpreted to arisefrom completely different microbial adaptations. Under extremelydry conditions (20 and 40 MPa) the costs of cell maintenancerise and the availability of substrate to sustain microbial acclima-tion would tend to decline, perhaps favoring lower cost solutionssuch as re-allocation of resources or transition into inactive anddormant states. Cell maintenance activities under low waterpotential, for example, are used to maintain cell wall uidity,metabolite regulation, and reallocation of resources (Potts et al.,2005). In this regard, the dynamics of sugar and nitrogen provideclues to understanding how soil communities respond to desic-cation and that may not be necessarily tied to osmolyte accumu-lation. The shifting of resources, such as PLFA, microbial C, and Nwithin a desiccation resistant microbial community suggests thatmicrobes could be re-allocating resources to cope with soil drying(Singh et al., 2002).

It has been hypothesized that the increases in microbial C thatoften occur during soil dryingmay be the result of re-distribution ofcellular C from structural to cytoplasmic cellular material (Schimelet al., 1989; Williams and Xia, 2009). However, the re-allocation inthe opposite direction, from cytoplasmic to structural cell wallmaterial may also occur (Kieft et al., 1994; Singh et al., 2002). In thisregard, it is interesting that the abundance of PLFA showedopposing trends to PSA-sugar abundance, for example, increasingthe PLFA concentration in the drought-prone Sumter and decliningin themoremesic Marietta soil in response to drought (Fig. 1b). Thedifferent response of the cellular sugar (carbohydrate) and PLFApool sizes further conrm the different community responses, butare also consistent with a cellular re-allocation hypothesis inresponse to water decit. The reallocation of resources betweenhigh energy carbohydrates and wall membranes may be animportant means of cellular adaptation to desiccation for microbesin soils (Roberson and Firestone, 1992; Potts, 2001; Gustavs et al.,

M.L. Kakumanu et al. / Soil Biolog2009; Hocking et al., 2012).4.3. Effect of water decit on microbial communities and fungal/bacterial (F/B) ratio

The response of the microbial community to desiccation maysimultaneously result in physiological re-allocation betweencellular pools and compositional community change. Shifts incommunity composition were much larger for the Sumter than theMarietta (Fig. 3) soil, indicating that the soils contain communitieswith different desiccation coping mechanisms. In particular, thegreater community dynamics (greater F/B ratio) in the drought-prone Sumter may suggest the presence of adaptations that allowgreater levels of activity and functioning during desiccation, anobservation consistent with the idea that fungi are more likely toremain active at very low water potentials than bacteria (Harris,1981; Shipton and Burggraaf, 1982). The more mesic and lessdrought-prone Marietta soil communities may adapt to drying, incontrast, by reducing activity and overall community change. Thisinterpretation is consistent with the 50e75% reduction inmicrobialrespiration in Marietta relative to the Sumter soil (data not shown).

Related to the EPS hypothesis, it was observed that desiccationincreased the F/B ratio of Sumter but not the Marietta soil. Bacteriaare well known for EPS production and fungi are likely to modifycell wall composition during stress periods (Csonka and Hanson,1991; Blomberg and Adler, 1992; Kieft et al., 1994; Sajbidor, 1997;Vargas, 2005). In this regard, the greater F/B ratio and increasingpool size of membrane PLFA is consistent with fungal cellmembrane based responses to water potential decit in thedrought-prone Sumter. Though the relatively high activity of fungimay increase cellular turnover and perhaps result in the loss ofsome modular biomass and hyphae when dry soil is re-wetted(Williams, 2007), it also provides the opportunity for foraging byhyphae and access to resources with less competition from rela-tively localized and sessile taxa, such as bacteria.

Change in microbial C and N pool sizes to desiccation can also beexplained by community changes. Because fungi generally requireless N than bacteria to build cellular biomass, the increasing PLFA-based fungal to bacterial biomass ratio (Fig. 2) is consistent with thegreater microbial C:N ratio with drying in the Sumter soil (Williamsand Xia, 2009). Decreasing the water potential, furthermore, mayalso increase cell wall growth relative to cytoplasmic growth, thusincreasing the C/N ratio of fungal organisms (Paustian andSchnrer, 1987) and provide an explanation for the increasingmicrobial C:N ratios with drying. This mechanism assumes limitingresources are re-directed to hyphal growth to enhance resourceexploitation. The difference in microbial N demand in fungicompared to bacteria could thus effect microbial physiologicalresponses and explain some of the different community responsesto drying between soils.

Among the bacterial groups, gram-positive bacteria, and tosome extent actinomycetes showed declines in their relativeabundance, an observation that goes against the idea that thesebacteria, because of their enhanced peptidoglycan layer should bemore resistant to drying (Potts, 1994). Indeed, recent studies haveshown opposite trends, that gram-negative are more well adaptedthan gram-positive bacteria to water stress (Williams, 2007;Williams and Rice, 2007; Aanderud and Lennon, 2011, Fig. 3.). It issimilarly possible that changes in methyl-branched PLFA, forexample, are indicative of physiological change in cell membranesused to cope with drying (Nordstrm, 1993).

Though fungi beneted relative to bacterial biomass as a resultof desiccation in the drought-prone Sumter soil, there was again noindication that specic fungal osmolytes, such as alcohols, wereproduced to adapt to drying soil. In contrast, fungi may alter theallocation of energy and C to cell wall production, and especially

Biochemistry 57 (2013) 644e653 651change fatty acid synthesis during desiccation (Fig. 2; Blomberg

y &and Adler, 1992). Bacteria, in contrast, may respond to desiccationby increasing the allocation of C to EPS on the outer surface of thecell wall (Roberson and Firestone, 1992), an interpretation consis-tent with increases in PSA-sugars in the Marietta soil (Fig. 1a).

5. Conclusion

The study was focused on testing the osmolyte accumulationhypothesis (OAH) in soil by characterizing the concentrations ofcommonmicrobial osmolytes across a gradient of desiccation stressunder in situ soil conditions. Eleven putative osmolytes (sugars andalcohols) were characterized but showed no consistent increaseswith water potential decit that would link their production tomicrobial adaptation to water decit. In contrast, the dynamics inmicrobial PLFA, microbial C, and PSA-sugars indicate that micro-organisms are responding to water decit. It is hypothesized thatmicrobes are adapting to desiccation, perhaps through the re-allocation of molecular resources to the cell wall (fungi) andextracellular cell wall EPS (bacteria). Some of the variability inmicrobial responses to soil drying may also be explained bydifferences in microbial community composition, structure, andactivity. The utilization of EPS production by bacteria to adapt todesiccation has been described previously, however, in a native soilcontext it has not been explicitly tested as a means of microbialadaptation. Compared to OAH, these hypothesized mechanisms ofmicrobial response may be more consistent with the oligotrophicenergy and nutrient conditions typically found in soils. Theresearch suggests that current conceptual models of microbialadaptation to water decit need re-evaluation.

Acknowledgments

Research was supported in part by NSFePCE and DOE awards toM.A. Williams

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.soilbio.2012.08.014.

References

Aanderud, Z.T., Lennon, J.T., 2011. Validation of heavy-water stable isotope probingfor the characterization of rapidly responding soil bacteria. Applied and Envi-ronmental Microbiology 77, 4589e4596.

Al-Hamdani, A., Cooke, R., 1987. Effects of water potential on accumulation andexudation of carbohydrates and glycerol during sclerotium formation andmyceliogenic germination in Sclerotinia sclerotiorum. Transactions of the BritishMycological Society 89, 51e60.

Birch, H., 1958. The effect of soil drying on humus decomposition and nitrogenavailability. Plant and Soil 10, 9e31.

Blomberg, A., Adler, L., 1992. Physiology of osmotolerance in fungi. Advances inMicrobial Physiology 33, 145e212.

Bossio, D., Scow, K., Gunapala, N., Graham, K., 1998. Determinants of soil microbialcommunities: effects of agricultural management, season, and soil type onphospholipid fatty acid proles. Microbial Ecology 36, 1e12.

Bremer, E., Krmer, R., 2000. Coping with osmotic challenges: osmoregulationthrough accumulation and release of compatible solutes in bacteria. BacterialStress Responses, 79e97.

Brown, A., 1976. Microbial water stress. Microbiology and Molecular BiologyReviews 40, 803.

Brown, A., Simpson, J.R., 1972. Water relations of sugar-tolerant yeasts: the role ofintracellular polyols. Journal of General Microbiology 72, 589e591.

Butler, J.L., Williams, M.A., Bottomley, P.J., Myrold, D.D., 2003. Microbial communitydynamics associated with rhizosphere carbon ow. Applied and EnvironmentalMicrobiology 69, 6793e6800.

Chowdhury, N., Burns, R.G., Marschner, P., 2011. Recovery of soil respiration afterdrying. Plant and Soil 348, 269e279.

M.L. Kakumanu et al. / Soil Biolog652Csonka, L.N., 1989. Physiological and genetic responses of bacteria to osmotic stress.Microbiology and Molecular Biology Reviews 53, 121e147.Csonka, L.N., Hanson, A.D., 1991. Prokaryotic osmoregulation: genetics and physi-ology. Annual Reviews in Microbiology 45, 569e606.

Dubois, M., Gilles, K., Hamilton, J., Rebers, P., Smith, F., 1956. New colorimetricmethods of sugar analysis VII. The phenolesulfuric acid reaction for carbohy-drate. Analytical Chemistry 28, 350e356.

Fan, T., 1996. Metabolite proling by one-and two-dimensional NMR analysis ofcomplex mixtures. Progress in Nuclear Magnetic Resonance Spectroscopy 28,161e219.

Fierer, N., Schimel, J., 2002. Effects of drying-rewetting frequency on soil carbon andnitrogen transformations. Soil Biology and Biochemistry 34, 777e787.

Franzluebbers, A., Haney, R., Honeycutt, C., Schomberg, H., Hons, F., 2000. Flush ofcarbon dioxide following rewetting of dried soil relates to active organic pools.Soil Science Society of America Journal 64, 613e623.

Frostegrd, ., Bth, E., 1996. The use of phospholipid fatty acid analysis to estimatebacterial and fungal biomass in soil. Biology and Fertility of Soils 22, 59e65.

Galinski, E.A., Truper, H.G., 1994. Microbial behaviour in salt-stressed ecosystems.FEMS Microbiology Reviews 15, 95e108.

Gallardo, A., Schlesinger, W.H., 1995. Factors determining soil microbial biomassand nutrient immobilization in desert soils. Biogeochemistry 28, 55e68.

Gustavs, L., Eggert, A., Michalik, D., Karsten, U., 2009. Physiological and biochemicalresponses of green microalgae from different habitats to osmotic and matricstress. Protoplasma, 1e12.

Halverson, L.J.J., Firestone, T.M., Mary, K., 2000. Release of intracellular solutes byfour soil bacteria exposed to dilution stress. Soil Science Society of AmericaJournal 64, 1630e1637.

Harris, R., 1981. Effect of water potential on microbial growth and activity. WaterPotential Relations in Soil Microbiology, 23e95.

Hocking, A.D., 1986. Effects of water activity and culture age on the glycerol accu-mulation patterns of ve fungi. Microbiology 132, 269e275.

Hocking, J., Priyadarshini, R., Takacs, C.N., Costa, T., Dye, N.A., Shapiro, L.,Vollmer, W., Jacobs-Wagner, C., 2012. Osmolality-dependent relocation ofpenicillin-binding protein PBP2 to the division site in Caulobacter crescentus.Journal of Bacteriology 194, 3116e3127.

Jorgenson, R., Brooks, P., 1990. Ninhydrin positive nitrogen measurements ofmicrobial biomass in 0.5 K2SO4 soil extracts. Soil Biology and Biochemistry 22,1023e1027.

Kelly, D.J.A., Budd, K., 1991. Polyol metabolism and osmotic adjustment in themycelial ascomycete Neocosmospora vasinfecta (EF Smith). ExperimentalMycology 15, 55e64.

Kempf, B., Bremer, E., 1998. Uptake and synthesis of compatible solutes as microbialstress responses to high-osmolality environments. Archives of Microbiology170, 319e330.

Kieft, T.L., Soroker, E., Firestone, M.K., 1987. Microbial biomass response to a rapidincrease in water potential when dry soil is wetted. Soil Biology andBiochemistry 19, 119e126.

Kieft, T., Ringelberg, D., White, D., 1994. Changes in ester-linked phospholipid fattyacid proles of subsurface bacteria during starvation and desiccation ina porous medium. Applied and Environmental Microbiology 60, 3292e3299.

Killham, K., Firestone, M., 1984. Salt stress control of intracellular solutes in strep-tomycetes indigenous to saline soils. Applied and Environmental Microbiology47, 301e306.

Killham, K., Firestone, M., 1984b. Proline transport increases growth efciency insalt-stressed Streptomyces griseus. Applied and Environmental Microbiology 48,239e241.

Liang, C., Fujinuma, R., Balser, T.C., 2008. Comparing PLFA and amino sugars formicrobial analysis in an Upper Michigan old growth forest. Soil Biology andBiochemistry 40, 2063e2065.

Lundquist, E., Jackson, L., Scow, K., 1999. Wetedry cycles affect dissolved organiccarbon in two California agricultural soils. Soil Biology and Biochemistry 31,1031e1038.

Magid, J., Kjrgaard, C., Gorissen, A., Kuikman, P., 1999. Drying and rewetting ofa loamy sand soil did not increase the turnover of native organic matter, butretarded the decomposition of added 14C-labelled plant material. Soil Biologyand Biochemistry 31, 595e602.

McCune, B., Grace, J.B., 2002. Analysis of Ecological Communities. MjM SoftwareDesign, Gleneden Beach, OR.

McInnes, K., Weaver, R., Savage, M., 1994. Soil water potential. Methods of SoilAnalysis, Part 2, 53e58.

Medeiros, P.M., Fernandes, M.F., Dick, R.P., Simoneit, B.R.T., 2006. Seasonal variationsin sugar contents and microbial community in a ryegrass soil. Chemosphere 65,832e839.

Mikha, M., Rice, C., Milliken, G., 2005. Carbon and nitrogen mineralization asaffected by drying and wetting cycles. Soil Biology and Biochemistry 37, 339e347.

Miller, K.J., Wood, J.M., 1996. Osmoadaptation by rhizosphere bacteria. AnnualReviews in Microbiology 50, 101e136.

Nordstrm, K.M., 1993. Effect of temperature on fatty acid composition of a whiteThermus strain. Applied and Environmental Microbiology 59, 1975e1976.

Paustian, K., Schnrer, J., 1987. Fungal growth response to carbon and nitrogenlimitation: application of a model to laboratory and eld data. Soil Biology andBiochemistry 19, 621e629.

Poolman, B., Glaasker, E., 1998. Regulation of compatible solute accumulation inbacteria. Molecular Microbiology 29, 397e407.

Biochemistry 57 (2013) 644e653Potts, M., 1994. Desiccation tolerance of prokaryotes. Microbiological Reviews 58,755e805.

Potts, M., 2001. Desiccation tolerance: a simple process? TRENDS in Microbiology 9,553e559.

Potts, M., Slaughter, S.M., Hunneke, F.U., Garst, J.F., Helm, R.F., 2005. Desiccationtolerance of prokaryotes: application of principles to human cells. Integrativeand Comparative Biology 45, 800e809.

Renella, G., Landi, L., Nannipieri, P., 2002. Hydrolase activities during and after thechloroform fumigation of soil as affected by protease activity. Soil Biology andBiochemistry 34, 51e60.

Roberson, E.B., Firestone, M.K., 1992. Relationship between desiccation and exo-polysaccharide production in a soil Pseudomonas sp. Applied and Environ-mental Microbiology 58, 1284e1291.

Safark, I.V.O., Santruckov, H., 1992. Direct determination of total soil carbohydratecontent. Plant and Soil 143, 109e114.

Sajbidor, J., 1997. Effect of some environmental factors on the content andcomposition of microbial membrane lipids. Critical Reviews in Biotechnology17, 87e103.

Schimel, J.P., Scott, W.J., Killham, K., 1989. Changes in cytoplasmic carbon andnitrogen pools in a soil bacterium and a fungus in response to salt stress.Applied and Environmental Microbiology 55, 1635e1637.

Schimel, J., Gulledge, J., Clein-Curley, J., Lindstrom, J., Braddock, J., 1999. Moistureeffects on microbial activity and community structure in decomposing birchlitter in the Alaskan taiga. Soil Biology and Biochemistry 31, 831e838.

Schimel, J., Balser, T., Wallenstein, M., 2007. Microbial stress-response physiologyand its implications for ecosystem function. Ecology 88, 1386e1394.

Schimel, J., Boot, C., Holden, P., Roux-Michollet, D., Parker, S., Schaeffer, S., Treseder,K., 2010. The Biogeochemistry of Drought. 19th World Congress of Soil Science,Soil Solutions for a Changing World, pp. 55e58.

Schleyer, M., Schmid, R., Bakker, E.P., 1993. Transient, specic and extremely rapidrelease of osmolytes from growing cells of Escherichia coli K-12 exposed tohypoosmotic shock. Archives of Microbiology 160, 424e431.

Shen, B., Hohmann, S., Jensen, R.G., 1999. Roles of sugar alcohols in osmotic stressadaptation. Replacement of glycerol by mannitol and sorbitol in yeast. PlantPhysiology 121, 45e52.

Shipton, W., Burggraaf, A.J.P., 1982. Frankia growth and activity as inuenced bywater potential. Plant and Soil 69, 293e297.

Singh, S.C., Sinha, R.P., Hader, D., 2002. Role of lipids and fatty acids in stresstolerance in cyanobacteria. Acta Protozoologica 41, 297e308.

Sorensen, L., 1974. Rate of decomposition of organic matter in soil as inuenced byrepeated air dryingerewetting and repeated additions of organic material. SoilBiology and Biochemistry 6, 287e292.

Sparling, G.P., West, A.W., 1989. Importance of soil water content when estimatingsoil microbial C, N and P by the fumigationeextraction methods. Soil Biologyand Biochemistry 21, 245e253.

Stark, J.M., Firestone, M.K., 1995. Mechanisms for soil moisture effects on activity ofnitrifying bacteria. Applied and Environmental Microbiology 61, 218e221.

Stevenson, F., 1982. Nitrogen-organic forms. In: Page, A.L. (Ed.), Methods of SoilAnalysis, Part 2. Chemical and Microbiological Properties, American Society ofAgronomy, Madison, pp. 625e641.

Tiemann, L.K., Billings, S.A., 2011. Changes in variability of soil moisture altermicrobial community C and N resource use. Soil Biology and Biochemistry 43,1837e1847.

Tschichholz, I., Trper, H.G., 1990. Fate of compatible solutes during dilution stressin Ectothiorhodospira halochloris. FEMS Microbiology Letters 73, 181e185.

Van Gestel, M., Ladd, J., Amato, M., 1992. Microbial biomass responses to seasonalchange and imposed drying regimes at increasing depths of undisturbed topsoilproles. Soil Biology and Biochemistry 24, 103e111.

Vance, E., Brookes, P., Jenkinson, D., 1987. An extraction method for measuring soilmicrobial biomass C. Soil Biology and Biochemistry 19, 703e707.

Vargas, C., Kallimanis, A., Koukkou, A.I., Calderon, M.I., Canovas, D., Iglesias-Guerra, F.,Drainas, C., Ventosa, A., Nieto, J.J., 2005. Contribution of chemical changes inmembrane lipids to the osmoadaptation of the halophilic bacterium Chromo-halobacter salexigens. Systematic and Applied Microbiology 28, 571e581.

White, D.C., Ringelberg, D.B., 1998. Signature Lipid Biomarker Analysis. OxfordUniversity Press, New York.

Williams, M., 2007. Response of microbial communities to water stress in irrigatedand drought-prone tallgrass prairie soils. Soil Biology and Biochemistry 39,2750e2757.

Williams, M., Rice, C., 2007. Seven years of enhanced water availability inuencesthe physiological, structural, and functional attributes of a soil microbialcommunity. Applied Soil Ecology 35, 535e545.

Williams, M., Xia, K., 2009. Characterization of the water soluble soil organic poolfollowing the rewetting of dry soil in a drought-prone tallgrass prairie. SoilBiology and Biochemistry 41, 21e28.

Witteveen, C.F.B., Visser, J., 1995. Polyol pools in Aspergillus niger. FEMS Microbi-ology Letters 134, 57e62.

Yancey, P.H., Clark, M.E., Hand, S.C., Bowlus, R.D., Somero, G.N., 1982. Living withwater stress: evolution of osmolyte systems. Science 217, 1214e1222.

Zhang, W., Parker, K., Luo, Y., Wan, S., Wallace, L., Hu, S., 2005. Soil microbialresponses to experimental warming and clipping in a tallgrass prairie. GlobalChange Biology 11, 266e277.

M.L. Kakumanu et al. / Soil Biology & Biochemistry 57 (2013) 644e653 653

Microbial community response to varying magnitudes of desiccation in soil: A test of the osmolyte accumulation hypothesis1. Introduction2. Materials and methods2.1. Site description2.2. Experimental setup2.3. Extraction of metabolites from soil and associated microorganisms2.4. Analysis of soil extracts by colorimetric methods2.5. Analysis of extractable metabolites by Gas ChromatographyMass Spectroscopy2.6. PLFA analysis2.7. Statistical analysis

3. Results3.1. Quantification of sugars in microbial extracts3.2. PLFAs and fungal to bacterial ratio3.3. Microbial C, N, and soluble C3.4. Ninhydrin reactive nitrogen (NRN)3.5. GCMS characterization of soil microbial extracts

4. Discussion4.1. Metabolite composition4.2. Microbial pools of PLFA, sugars & NRN4.3. Effect of water deficit on microbial communities and fungal/bacterial (F/B) ratio

5. ConclusionAcknowledgmentsAppendix A. Supplementary dataReferences