S

C

Ma

b

a

ARRA

KDFSST

1

ientettee2

inoafba

P

0h

Applied Soil Ecology 62 (2012) 98– 102

Contents lists available at SciVerse ScienceDirect

Applied Soil Ecology

journa l h o me page: www.elsev ier .com/ locate /apsoi l

hort communication

onsiderations in the storage of soil samples for enzyme activity analysis

atthew S. Peoplesa, Roger T. Koidea,b,∗

Department of Horticulture, The Pennsylvania State University, University Park, PA 16802, USAGraduate Program in Ecology, The Pennsylvania State University, University Park, PA 16802, USA

r t i c l e i n f o

rticle history:eceived 15 June 2012eceived in revised form 23 July 2012ccepted 5 August 2012

eywords:ryingreezingoil enzymetorage

a b s t r a c t

The most accurate means of assessing soil enzyme activity involves analyzing fresh soil samples veryrecently taken from the field. Because of the extensive handling times involved, however, such anapproach is frequently impossible when moderate to large numbers of soil samples are involved. There-fore, soil samples are frequently stored, either frozen or dried, prior to analysis. Unfortunately, dependingon soil type and enzyme, both freezing and drying may significantly alter enzyme activity. In many casessuch a storage effect is tolerable as long as the relationships among “treatments” remain unaltered as aconsequence of storage. In this study we used two soils (previously determined to possess significantlydifferent enzyme activities) as proxies for experimental treatments, and calculated the ratio of one soil’sactivity to that of the other for fresh, frozen and dried samples to determine whether soil storage influ-

reatment enced the relationship between the soils. The enzymes included 1,4-�-cellobiohydrolase (CBase), acidphosphatase (Pase), and �-N-acetylglucosaminidase (NAGase). The ratio of activity of the two soils wassignificantly affected by freezing and drying when compared to the fresh control for all three enzymes, butthe ratio generally changed much less as a consequence of freezing than as a consequence of drying. Eachinvestigator should consider whether soil storage affects within-experiment treatment comparisons of

soil enzyme activity.

. Introduction

Enzymes are the means by which soil microorganisms mineral-ze organic matter. Enzymes, therefore, influence many importantcosystem processes including organic matter decomposition,utrient cycling and carbon sequestration, and an assessment ofheir activities affords insight into these processes (Sinsabaught al., 2008). Knowledge of soil enzyme activities has also been usedo gain insight into the effects of various management practices onhe productivity of agricultural systems (Bolton et al., 1985; Dickt al., 1988; Bandick and Dick, 1999) and the function of naturalcosystems (Saiya-Cork and Sinsabaugh, 2002; Sinsabaugh et al.,008; Wallenstein et al., 2009; Weedon et al., 2011).

In most cases it is desirable to assess the activity of soil enzymesn situ, but the required genomic, proteomic or metabolomic tech-ologies are not those that are commonly available to agronomistsr ecologists (Wallenstein and Weintraub, 2008). When in situssessment is impractical, measurement of enzyme activity on

resh soil samples very soon after collection is considered the nextest alternative (DeForest, 2009; Wallenius et al., 2010). Enzymenalyses are laborious, requiring collection, homogenization and

∗ Corresponding author at: Department of Biology, Brigham Young University,rovo, UT 84602, USA. Tel.: +1 802 422 6650.

E-mail address: rxk13@psu.edu (R.T. Koide).

929-1393/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsoil.2012.08.002

© 2012 Elsevier B.V. All rights reserved.

dilution of soil samples, as well as the precise timing of the assess-ment of enzyme activity based on the production of colored orfluorescent products following the addition of substrates (Eivaziand Tabatabai, 1977; Saiya-Cork and Sinsabaugh, 2002). Large num-bers of fresh samples are not easily analyzed in a reasonable periodof time, limiting the number of samples that can be processed and,thus, the number of comparisons that can be made. Consequently,various methods for storing soil samples prior to analysis have beenproposed (Lee et al., 2007; DeForest, 2009).

Unfortunately, common methods of storing soil samplesfrequently result in significant alteration to enzyme activity.Depending on soil and enzyme, both freezing at −20 ◦C and dry-ing at room temperature may significantly alter activity (Walleniuset al., 2010; Abellan et al., 2011). For many projects, however, itis more important to understand the relationships among experi-mental treatments, vegetation types, soils or sampling times thanto determine absolute enzyme activities. For example, if contrastsamong soils were of interest, a soil storage method may be accept-able if it halved the enzyme activities (compared to the freshsamples) for all soils because the relationships among soils wouldbe unaffected. In this communication we report our results ofan experiment in which we compared enzyme activities of two

soils, both currently in agricultural production, for which prelimi-nary tests demonstrated significant differences in the activities of1,4-�-cellobiohydrolase (CBase), acid phosphatase (Pase), and �-N-acetylglucosaminidase (NAGase), enzymes involved in microbial

lied Soil Ecology 62 (2012) 98– 102 99

attes

2

2

pwlhcs4syoefiuhctwa

2

usisshp

12ePhat∼e∼wwpoucecc

(alp

Table 1The effect of soil aliquot weight on the coefficient of variation (SD ÷ mean × 100)among aliquots taken from the same homogenized sample of Rock Springs soil.SD = standard deviation. n = 4.

25 g soil aliquots 100 g soil aliquots

Mean SD CV (%) Mean SD CV (%)

Pase 18.0 1.93 10.7 16.9 0.36 2.1

M.S. Peoples, R.T. Koide / App

cquisition of carbon, phosphorus and nitrogen, respectively. Thewo soils were used as proxies for experimental treatments. Weested whether the relationship between two soil’s activities thatxisted for fresh samples remained unaffected as a consequence oftorage (frozen vs. dried).

. Materials and methods

.1. Soils

Two contrasting agricultural soils were used in this study asroxies for experimental treatments. One soil (Hagerstown seriesith a silt loam surface texture and subsurface textures of silty clay

oam and silty clay, fine, mixed, semiactive, mesic Typic Hapludalfs,ttp://soils.usda.gov/technical/classification/osd/index.html) wasollected from the Pennsylvania State University’s Russell Lar-on Research and Education Center at Rock Springs, PA, USA,0◦42′45.65′′N, 77◦57′26.21′′W. This soil had been planted withoybeans in the year prior to collection and with maize in theear of collection. The second soil (Weikert series, consistingf shallow, well drained soils formed in material that weath-red from interbedded gray and brown acid shale, siltstone, andne-grained sandstone on gently sloping to very steep areas onplands, loamy-skeletal, mixed, active, mesic Lithic Dystrudepts,ttp://soils.usda.gov/technical/classification/osd/index.html) wasollected from another farm designated “Duff farm” operated byhe Pennsylvania State University (40◦39′7.76′′N, 77◦53′38.32′′W)hich had been planted with maize in the year prior to collection

nd with Timothy grass (Phleum pratense L.) in the year of collection.

.2. Determination of aliquot weight

The Rock Springs soil was used in a preliminary study to helps determine aliquot weight for our subsequent work. The soil wasampled to a depth of 15 cm with a 2.2 cm diameter corer. Approx-mately 6 L of each soil were collected and pooled to form a singleample. In the laboratory, the sample was passed through a 2 mmieve to remove large roots and stones. It was then homogenized byand and stored in the refrigerator at 4 ◦C in an enclosed, air-tightlastic bin for 14 and 16 d.

Many researchers typically utilize 1 g (Eivazi and Tabatabai,977; Saiya-Cork and Sinsabaugh, 2002) or, at most, 5 g (DeForest,009) fresh weight aliquots of homogenized soil samples to assessnzyme activity, which can result in a high degree of variation.reliminary analyses of our soil using aliquots from a single,omogenized soil sample indicated that 1 g aliquots resulted in

high degree of variability in enzyme activity, which is consis-ent with that observed by DeForest (2009). We, therefore, tested25 g and ∼100 g (fresh weight) aliquots from a single, homog-nized soil sample. Four ∼25 g aliquots were compared to four100 g aliquots. Each aliquot was added to a food homogenizerith 750 mL water and homogenized at high speed for 30 s. Then,hile the homogenizer was set to low in order to keep the soil sus-ended, either 0.25 mL (100 g aliquots) or 1.00 mL (25 g aliquots)f soil homogenate were transferred to 2 mL microcentrifuge tubessing a pipette tip with the end clipped to accommodate soil parti-les. Two replicate aliquots were taken from each homogenate forach enzyme. Water (0.75 mL) was added to the microfuge tubesontaining 0.25 mL homogenate so the final volume was, in eachase, 1.00 mL.

We assayed for three enzymes including acid phosphatase

Pase, EC 3.1.3.2), 1,4-�-cellobiohydrolase (CBase, EC 3.2.1.91)nd �-N-acetylglucosaminidase (NAGase, EC 2.4.1.255). The MUB-inked substrates used were, respectively, 4-methylumbelliferylhosphate (MUB-P, 69607, Fluka), 4-methylumbelliferyl

CBase 3.51 0.16 4.5 3.58 0.17 4.8NAGase 2.90 0.48 16.7 3.13 0.29 9.2

�-d-cellobioside (MUB-CB, M6018, Sigma), and 4-methylumbelliferyl N-acetyl-�-d-glucosaminide (MUB-NAG,M2133, Sigma). All enzyme substrate solutions (200 �M) werestored at −20 ◦C. Once thawed, substrate solutions were refrig-erated at 4 ◦C for no more than 3 days (MUB-CB, MUB-NAG), or1 day (MUB-P, which is less stable, see DeForest, 2009). Enzymesubstrates (0.50 mL of 200 �M) were added to the homogenatesand the microfuge tubes were placed horizontally on a mixer atlow speed for the duration of the incubation period (45 min forMUB-NAG, 60 min for MUB-CB, 20 min MUB-P).

Upon completion of the incubation, 0.50 mL of 50 mM sodiumhydroxide (NaOH) was added to stop the reaction, bringing the totalvolume in each tube up to 2.00 mL. Tubes were then centrifuged at10,000 × g for 5 min. Because fluorescence may not be temporallystable following addition of NaOH, the time between adding NaOHand reading the plate was always between 8 and 11 min.

Two hundred microliters of supernatant were added to wellsof black, polystyrene 96-well microplates, with one column, eightanalytical replicates, being used per sample (microcentrifuge tube).Four control columns were also filled in strict order as in DeForest(2009). The MUB standard wells contained 100 �L water, 50 �L10 �M �-methylumbelliferone (MUB, M1381, Sigma), and 50 �L50 mM NaOH. The substrate blank, containing 100 �L water, 50 �L200 �M substrate, 50 �L 50 mM NaOH. The soil autofluorescenceblank, containing 50 �L water, 100 �L of sample supernatant (with-out substrate) and 50 �L 50 mM NaOH. The quenching controlcontained 100 �L of sample supernatant (without substrate), 50 �L10 �M MUB, and 50 �L 50 mM NaOH. Fluorescence was deter-mined using a BMG FluoStar unit with a 355 nm excitation filterand 460 nm emission filter. Enzyme activities were calculated fromaverage fluorescence of the eight analytical replicates (with auto-fluorescence and substrate fluorescence subtracted), adjusted forthe quenching coefficient, and divided by incubation time and105 ◦C dry weight of the soil sample. A separate soil sample wasdried at 105 ◦C for the purpose of determining moisture contentfor each soil type, and this was used to calculate dry weights.Enzyme activity was calculated and reported as nmol MUB pro-duced min−1 g−1 soil dry weight.

For Pase and NAGase the coefficient of variation was approxi-mately two to five times larger for the 25 g aliquots than for the100 g aliquots while for CBase the coefficient of variation was simi-lar for 25 and 100 g soil aliquots (Table 1). Therefore, 100 g sampleswere used for all further assays, below.

2.3. Effect of freezing or drying on relationships among soils

Two soils (Duff and Rock Springs, see above) were sampled to adepth of 15 cm with a 2.2 cm diameter corer. Approximately 10 L ofeach soil were collected and pooled to form a single sample of eachsoil. In the laboratory, soils were passed through a 2 mm sieve toremove large roots and stones. Each soil sample was then homog-

enized by hand, and aliquots (exact weights were recorded) wereset aside for immediate analysis (within 3 h of collection) or storedeither uncovered, at room temperature (for the drying treatment)or in plastic bags and frozen at −20 ◦C (for the freezing treatment).

100 M.S. Peoples, R.T. Koide / Applied Soil Ecology 62 (2012) 98– 102

Table 2Table of p values obtained from single factor analyses of variance to determine the effect of treatment (fresh, stored for 14 d, stored for 28 d) on absolute activity of the threeenzymes, two soils (proxies for treatment) and two storage methods. Means of the activities are shown in Fig. 1. n = 4.

Soil type Dried Frozen

Pase Pase CBase NAGase CBase NAGase

0.0000.000

Fpatf−

wsastiw1

Foa9l

Duff <0.0001 <0.0001 <Rock Springs <0.0001 0.0002 <

or each soil, four ∼100 g aliquots of the fresh soil sample were pre-ared as above and assayed for activity of the three enzymes (seebove). The activities of the other ∼100 g aliquots were determinedhat had either been dried at room temperature in uncovered pansor 14 d and 28 d, or that had been stored in sealed containers at20 ◦C for 14 d and 28 d.

Three-factor analyses of variance (treatment × soil × enzyme)ere performed on absolute enzyme activities for dry and frozen

torage methods separately, in which treatments included freshliquots as well as those stored for either 14 or 28 d. Becauseignificant two- and three-factor interactions were observed in

he three-factor analyses of variance of absolute enzyme activ-ties, 12 separate, single-factor analyses were performed in

hich the effect of treatment (fresh and those stored for either4 or 28 d) was analyzed separately for each enzyme-storage

fresh 14 d dry 28 d dry0

3

6

9

12

15

18

fresh 14 d dry 28 d dry

Pas

e ac

tivity

(nm

ol m

in-1

g-1

)

0

20

40

60

80

100DuffRock Springs

fresh 14 d dry 28 d dry0

2

4

6

8

10

12

CB

ase

activ

ity (n

mol

min

-1 g

-1)

NA

Gas

e ac

tivity

(nm

ol m

in-1

g-1

)

a

c

e

A

B B

x

yz

A

BB

x

yy

A

BB

x

y z

ig. 1. Plots of absolute activities of three enzymes (Pase, CBase, and NAGase), for two sf the 12 combinations of enzyme, soil and storage method, the effect of treatment (frectivity, frozen; (c) CBase activity, dried; (d) CBase activity, frozen; (e) NAGase activity, d5% confidence intervals. Different capital letters indicate significant (p < 0.05) difference

etters indicate significant (p < 0.05) differences of means for Rock Springs soil according

1 0.0004 <0.0001 <0.00011 0.0056 0.0052 0.0021

method-soil combination. Means were separated using the 95%confidence intervals.

In order to determine whether either freezing or drying pre-served the relationships of enzyme activities between soils (proxiesfor treatments) that were apparent for fresh aliquots, we examinedthe ratios of activity between Duff soil and Rock Springs soil foreach enzyme when the aliquots were fresh, frozen (14 d and 28 d)and dried (14 d and 28 d). In order to calculate the error associatedwith the various ratios of the soils (ratio of Duff to Rock Springwhile in the fresh, 14 d dry, 28 d dry, 14 d frozen and 28 d frozenstates), the ratios were calculated for all 16 possible combinations

(for example, for fresh Pase, there were four replicate Duff aliquotsand four Rock Springs aliquots). The 16 ratios were then randomlysampled with replacement four at a time 1000 times to bootstrapa frequency distribution of the ratios, and the mean and standard

fresh 14 d frozen 28 d frozen

fresh 14 d frozen 28 d frozen

fresh 14 d frozen 28 d frozen

b

d

f

A

BB

xxy y

A

BBx

y y

A

BC

x

y y

oils (Duff and Rock Springs) and two storage methods (dried and frozen). For eachsh, stored for 14 d, stored for 28 d) was analyzed. (a) Pase activity, dried; (b) Paseried; and (f) NAGase activity, frozen. In each case, n = 4. Error bars represent pooleds of means for Duff soil according to the confidence intervals. Different lower caseto the confidence intervals.

M.S. Peoples, R.T. Koide / Applied Soil Ecology 62 (2012) 98– 102 101

Table 3Table of p values obtained from the single factor analyses of variance using the bootstrapped mean ratios (see Section 2) of Duff to Rock Springs soil enzyme activities (seeSection 2) to determine the effect of treatment (fresh, stored for 14 d, stored for 28 d) on the ratios of three enzymes and two storage methods. Means of the ratios are shownin Fig. 2. n = 4.

Dried Frozen

N

<

dpfFws

3

ia

Ffi(

Enzyme Pase CBase

Duff:Rock Springs <0.0001 <0.0001

eviation were determined. Finally, six single factor analyses wereerformed on the resulting ratios for fresh and stored (14 or 28 d)or each enzyme and storage treatment (dried and frozen), and the

values were calculated assuming n = 4. Data were ln-transformedhen necessary to produce homogeneous variances. Means were

eparated using the 95% confidence intervals.

. Results

The analyses of variance indicated that both drying and freez-ng significantly decreased the absolute value of all three enzymesctivities as compared to fresh soil for both soil types (Table 2,

fresh 14 d dry 28 d dry

Duf

f : R

ock

Spr

ings

Pas

e

0

1

2

3

4

5

fresh 14 d dry 28 d dry0.0

0.5

1.0

1.5

2.0

2.5

fresh 14 d dry 28 d dry0.0

0.5

1.0

1.5

2.0

2.5

3.0

Duf

f : R

ock

Spr

ings

CB

ase

Duf

f : R

ock

Spr

ings

NA

Gas

e

a

c

e

cb

a

b b

a

a

b

a

ig. 2. Plots of the ratio of the enzyme activities of Duff to Rock Springs soils at each timerozen). For each of the six combinations of enzyme and storage method, the effect of treandicate pooled 95% confidence intervals. (a) Pase activity, dried; (b) Pase activity, frozen;f) NAGase activity, frozen. Different lower case letters indicate significant (p < 0.05) diffe

AGase Pase CBase NAGase

0.0001 0.0012 0.0025 0.0005

Fig. 1). Dried aliquots of Duff soil were stable, exhibiting no sig-nificant change in activity between 14 and 28 d. Dried aliquots ofRock Springs soil were not stable in either Pase or NAGase activity,which declined between 14 and 28 d. Frozen aliquots of Duff soilwere stable with the exception of NAGase activity, which increasedfrom 14 to 28 d. Frozen aliquots of Rock Springs soil did not changesignificantly between 14 and 28 d of storage.

The analyses of the ratios of activity (Duff to Rock Springs) indi-

cated that storage significantly altered the relationship betweenthe two soils for each of the enzymes (Table 3, Fig. 2). In gen-eral, freezing had a smaller effect than drying on the relationshipsbetween the two soils. The largest change in the ratio of Duff to

fresh 14 d frozen 28 d frozen

fresh 14 d frozen 28 d frozen

fresh 14 d frozen 28 d frozen

b

d

f

ab

a

a abb

a

bab

for three enzymes (Pase, CBase and NAGase) and two storage methods (dried andtment (fresh, stored for 14 d, stored for 28 d) on the ratio was analyzed. Error bars

(c) CBase activity, dried; (d) CBase activity, frozen; (e) NAGase activity, dried; andrences of means of ratios among treatments according to the confidence intervals.

1 lied So

RstNs

4

ronoiffstDafi

tamhfaaLebss

(ttpoo2bti

tdm

eling the in situ activity of soil extracellular enzymes. Soil Biol. Biochem. 40,

02 M.S. Peoples, R.T. Koide / App

ock Springs enzyme activity (a 181% increase compared to freshamples) occurred with Pase as a result of drying for 28 days. Forhe frozen aliquots the largest change in the ratio occurred withAGase for 14 days of storage, a 22% decrease from original fresh

ample activity.

. Discussion

Despite excellent contributions to the methodology in theecent past (Lee et al., 2007; DeForest, 2009), standardized meth-ds for the handling of soils prior to analysis of soil enzymes haveot been clearly established. Many studies require the collectionf more samples than can be processed in a single day, preclud-ng assessment of activity on fresh soil samples and requiring someorm of storage. While drying and freezing remain among the mostrequently used methods of soil storage prior to enzyme analysis,ome investigators have shown that drying or freezing can leado significant shifts in absolute enzyme activity (Lee et al., 2007;eForest, 2009; Wallenius et al., 2010; Abellan et al., 2011). Welso found that storage of aliquots either in the dry or frozen staterequently resulted in significant changes in absolute enzyme activ-ties when compared to activities of fresh aliquots.

Changes in absolute enzyme activity due to storage may still beolerable in studies in which treatment comparisons are of interests long as storage has the same proportional effect on all treat-ents. Some researchers may implicitly assume that storage does

ave the same proportional effect on all treatments. However, veryew of the previously published investigations of soil enzymesssessed whether relationships among treatments in their enzymectivities changed as a consequence of storage. DeForest (2009) andee et al. (2007) suggested that storage had unique effects on differ-nt soils. In our study we unequivocally showed that relationshipsetween the two soils in the activities of the three enzymes wereignificantly altered by storage of soil both in the frozen and driedtate.

Soil enzymes selectively adhere to organomineral complexesBusto and Perez-Mateos, 2000), and this may serve to increasehe stability of the enzymes (Perezmateos et al., 1991). Thus, soilshat differ in the concentration of organomineral complexes areredicted to vary in their response to freezing or drying. More-ver, management practices that influences the concentration ofrganomineral complexes, such as tillage (Monreal and Bergstrom,000), are also expected to differentially influence soil enzyme sta-ility in storage, even for a single soil type. Thus one might expecthe enzyme activities of different treatments to be differentiallynfluenced by storage.

In our study, the change in the enzyme activity ratio of one soilo the other was generally smaller for frozen samples and larger forried samples. In this case, freezing appears to be a better storageethod than drying. However, our results are not meant to provide

il Ecology 62 (2012) 98– 102

a single method that is appropriate for all situations. We suggestthat in each investigation where storage is employed investigatorsmust possess a clear understanding of the effects of storage in orderto properly interpret treatment effects on soil enzyme activity. Insome cases it may be useful to establish correction factors that canbe applied separately to each treatment prior to comparison.

Acknowledgments

This study was supported, in part, by grants from the USDA NIFAprogram in bioenergy and the Northeast Sun Grant. The authors arevery grateful to Kristin Haider for her editorial comments and herinvaluable help with R.

References

Abellan, M.A., Baena, C.W., Morote, F.A.G., Cordoba, M.I.P., Perez, D.C., 2011. Influenceof the soil storage method on soil enzymatic activities in Mediterranean forestsoils. Forest Syst. 20, 379–388.

Bandick, A.K., Dick, R.P., 1999. Field management effects on soil enzyme activities.Soil Biol. Biochem. 31, 1471–1479.

Bolton Jr., H., Elliott, L., Papendick, R., Bezdicek, D., 1985. Soil microbial biomass andselected soil enzyme activities: effect of fertilization and cropping practices. SoilBiol. Biochem. 17, 297–302.

Busto, M.D., Perez-Mateos, M., 2000. Characterization of beta-d-glucosidaseextracted from soil fractions. Eur. J. Soil Sci. 51, 193–200.

DeForest, J.L., 2009. The influence of time, storage temperature, and substrate age onpotential soil enzyme activity in acidic forest soils using MUB-linked substratesand l-DOPA. Soil Biol. Biochem. 41, 1180–1186.

Dick, R.P., Rasmussen, P.E., Kerle, E.A., 1988. Influence of long-term residue man-agement on soil enzyme activities in relation to soil chemical properties of awheat-fallow system. Biol. Fertil. Soils 6, 159–164.

Eivazi, F., Tabatabai, M., 1977. Phosphatases in soils. Soil Biol. Biochem. 9, 167–172.Lee, Y.B., Lorenz, N., Dick, L.K., Dick, R.P., 2007. Cold storage and pretreatment incu-

bation effects on soil microbial properties. Soil Sci. Soc. Am. J. 71, 1299.Monreal, C.M., Bergstrom, D.W., 2000. Soil enzymatic factors expressing the influ-

ence of land use, tillage system and texture on soil biochemical quality. Can. J.Soil Sci. 80, 419–428.

Perezmateos, M., Busto, M.D., Rad, J.C., 1991. Stability and properties of alkalinephosphate immobilized by a rendzina soil. J. Sci. Food Agric. 55, 229–240.

Saiya-Cork, K.R., Sinsabaugh, R.L., 2002. The effects of long term nitrogen deposi-tion on extracellular enzyme activity in an Acer saccharum forest soil. Soil Biol.Biochem. 34, 1309–1315.

Sinsabaugh, R.L., Lauber, C.L., Weintraub, M.N., Ahmed, B., Allison, S.D., Crenshaw, C.,Contosta, A.R., Cusack, D., Frey, S., Gallo, M.E., Gartner, T.B., Hobbie, S.E., Holland,K., Keeler, B.L., Powers, J.S., Stursova, M., Takacs-Vesbach, C., Waldrop, M.P., Wall-enstein, M.D., Zak, D.R., Zeglin, L.H., 2008. Stoichiometry of soil enzyme activityat global scale. Ecol. Lett. 11, 1252–1264.

Wallenius, K., Rita, H., Simpanen, S., Mikkonen, A., Niemi, R.M., 2010. Sample storagefor soil enzyme activity and bacterial community profiles. J. Microbiol. Methods81, 48–55.

Wallenstein, M.D., Mcmahon, S.K., Schimel, J.P., 2009. Seasonal variation in enzymeactivities and temperature sensitivities in Arctic tundra soils. Glob. Change Biol.15, 1631–1639.

Wallenstein, M.D., Weintraub, M.N., 2008. Emerging tools for measuring and mod-

2098–2106.Weedon, J.T., Aerts, R., Kowalchuk, G.A., van Bodegom, P.M., 2011. Enzymology under

global change: organic nitrogen turnover in alpine and sub-Arctic soils. Biochem.Soc. Trans. 39, 309–314.