Soil organic matter dynamics in a North America tallgrass prairie after 9 yr of experimental warming

Transcription

1 iogeosciences, 8, , doi:1.5194/bg uthor(s) 211. CC ttribution 3. License. iogeosciences Soil orgnic mtter dynmics in North meric tllgrss pririe fter 9 yr of experimentl wrming X. Cheng 1,2, Y. Luo 2, X. Xu 2, R. Sherry 2, nd Q. Zhng 1 1 Key Lbortory of qutic otny nd Wtershed Ecology, Wuhn otnicl Grden, The Chinese cdemy of Sciences, Wuhn 4374, Chin 2 Deprtment of otny nd Microbiology, University of Oklhom, 77 Vn Vleet Ovl, Normn, OK 7319, US Received: 1 October 21 Published in iogeosciences Discuss.: 15 November 21 Revised: 24 pril 211 ccepted: 5 My 211 Published: 9 June 211 bstrct. The influence of globl wrming on soil orgnic mtter (SOM) dynmics in terrestril ecosystems remins uncler. In this study, we combined soil frctiontion with isotope nlyses to exmine SOM dynmics fter nine yers of experimentl wrming in North meric tllgrss pririe. Soil smples from the control plots nd the wrmed plots were seprted into four ggregte sizes (>2 µm, 25 2 µm, µm, nd <53 µm), nd three density frctions (free light frction LF, intr-ggregte prticulte orgnic mtter ipom, nd minerl-ssocited orgnic mtter msom). ll frctions were nlyzed for their crbon (C) nd nitrogen (N) content, nd δ 13 C nd δ 15 N vlues. Wrming did not significntly effect soil ggregte distribution nd stbility but incresed C 4 -derived C input into ll frctions with the gretest in LF. Wrming lso stimulted decy rtes of C in whole soil nd ll ggregte sizes. C in LF turned over fster thn tht in ipom in the wrmed soils. The δ 15 N vlues of soil frctions were more enriched in the wrmed soils thn those in the control, indicting tht wrming ccelerted loss of soil N. The δ 15 N vlues chnged from low to high, while C:N rtios chnged from high to low in the order LF, ipom, nd msom due to incresed degree of decomposition nd minerl ssocition. Overll, wrming incresed the input of C 4 -derived C by 11.6 %, which ws offset by the ccelerted loss of soil C. Our results suggest tht globl wrming simultneously stimultes C input vi shift in species composition nd decomposition of SOM, resulting in negligible net chnge in soil C. Correspondence to: X. Cheng (xlcheng@fudn.edu.cn) 1 Introduction Recent Intergovernmentl Pnel on Climte Chnge report (IPCC, 27) predicts globl verge temperture to increse by C during current century. Globl wrming is expected to profoundly impct ecosystem processes such s soil orgnic mtter (SOM) dynmics (e.g., Dvidson nd Jnssens, 26; Von Fischer et l., 28). Crbon (C) in SOM ccounts for 8 % of terrestril C pool nd is regrded s n importnt potentil C sink tht my help offset the greenhouse effect (e.g., Ll, 28; Mi et l., 21). Smll chnges in SOM stock under globl chnge cn potentilly effect tmospheric CO 2 concentrtions (e.g., tjes nd Sombroek, 1997; Mrin-Spiott et l., 29). In ddition, wrming-induced chnges in SOM regulte the vilbility of nitrogen (N) for plnt growth nd ultimtely influence the net primry productivity of terrestril ecosystems. Hence, it is impertive to understnd how globl wrming will effect SOM dynmics. Effects of wrming on SOM dynmics remin widely debted topic (e.g., Pendll et l., 24). For exmple, climtic wrming increses soil temperture nd hence ccelertes orgnic mtter decomposition rtes, leding to loss of soil C nd N (e.g., Rustd et l., 21; Fontine et l., 24). Conversely, some studies hve reported tht wrming leds to increses in soil C nd N becuse of gret increses in biomss nd litter inputs in tundr ecosystems (e.g., Welker et l., 24; Dy et l., 28). These differences re not surprising given response of soils to wrming depends on mny fctors, such s soil moisture nd temperture nd, in prticulr, on plnt species tht provide crbon inputs to soils (e.g., Shw nd Hrte, 21; Fissore et l., 28). Most SOM derives exclusively from the plnt mteril growing on site. Chnges Published by Copernicus Publictions on behlf of the Europen Geosciences Union.

2 1488 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe in vegettion type re thus expected to lter the qulity nd quntity of SOM (Cheng et l., 26; Fissore et l., 28). Recent climtic wrming hs lredy led to drmtic shifts in plnt functionl groups (e.g., C 3 litter with high qulity nd C 4 litter with low qulity), nd this cn effect the ccumultion nd decomposition ptterns of SOM by ltering the quntity nd qulity of plnt mteril entering into soil (Dy et l., 28; Fissore et l., 28). Therefore, understnding the response of SOM to climtic wrming is criticl for ccurte predictions of long-term ecosystem C nd N cycling in future climtic scenrios. However, detecting chnges in the SOM stock of terrestril ecosystem under globl chnge cn be difficult, becuse SOM consists of complex composition with different physicl nd chemicl stbilities (Vn Groenigen et l., 22; Del Gldo et l., 23; Mrin-Spiott et l., 29). To chrcterize chnges in soil C nd SOM dynmics correctly, size nd density frctiontion techniques hve been developed to seprte bulk soil into frctions tht differ in microbil degrdbility nd turnover time (e.g., Jstrow, 1996; Six et l., 2; Mrin-Spiott et l., 29). ggregte size frctiontions hve shown tht C in SOM ssocited with lrger ggregtes hs higher turnover rtes thn C in SOM ssocited with smller ggregtes (Jstrow, 1996; Six et l., 2). Density frctiontions result in light frction, which is composed of physiclly unprotected plnt debris nd it is generlly thought to hve rpid turnover s well s hevy minerl ssocited frction, which remins more reclcitrnt with long-term turnover (lesdent, 1996). Nturl bundnce of stble C isotopes coupled with SOM frctiontion technique offers n pproch to better quntify SOM dynmics when globl chnge induces shift in the dominnt plnt species composition between C 4 nd C 3 (López-Ullo et l., 25; John et l., 25; uerswld et l., 29; Mrin-Spiott et l., 29). Theoreticlly, differences in the nturl stble C isotope signture between C 3 (verge δ 13 C vlue of 27 ) nd C 4 (verge δ 13 C vlue of 11 ) plnts result in SOM with distinct isotopic signtures. Chnges in δ 13 C vlues of SOM over time following chnge in vegettion cn be used to exmine the reltive contribution of C 3 - or C 4 -derived C to SOM formtion (Del Gldo et l., 23; Cheng et l., 26) nd quntify SOM decomposition rtes (e.g., Lio et l., 26). Furthermore, soil δ 15 N vlues reflect the net effect of N-cycling processes s influenced by climte chnge nd species composition (Robinson, 21; Dwson et l., 22; ijoor et l., 28). For instnce, incresed soil temperture hs been suggested to enhnce rtes of N cycling nd loss of N, resulting in 15 N enrichment (ijoor et l., 28). The soil δ 15 N vlues cn be lso used to estimte the degree of SOM decomposition nd humifiction (Krmer et l., 23; Lio et l., 26; Templer et l., 27; Mrin-Spiott et l., 29) In Centrl Oklhom US Gret Plins, long-term, ongoing experimentl wrming nd clipping experiment ws initited on 21 November 1999 in tllgrss pririe (Luo et l., 21), dominted by mixture of C 4 grsses nd few C 3 forbs. Wrming hs resulted in shift towrds more C 4 -grss dominted plnt community nd n increse in boveground biomss s well s boveground net primry productivity (NPP) (Wn et l., 25; Luo et l., 29), nd hence incresed litter input nd ltered litter qulity (n et l., 25; Cheng et l., 21). These chnges provide unique opportunity to utilize the nturl bundnce of δ 13 C nd δ 15 N to evlute chnges in SOM dynmics fter nine yers of experimentl wrming. We hypothesized tht nine yers of wrming would significntly increse SOM storge due to wrming-induced increses in litter input nd chnges in litter qulity (n et l., 25; Cheng et l., 21). To test this hypothesis, we mesured the δ 13 C, δ 15 N, C, nd N concentrtions in ll SOM ggregtes nd density frctions in the tllgrss pririe experiment. The specific objectives of this study were to: (1) evlute the impct of the long-term experimentl wrming on the C nd N pools in SOM frctions; (2) quntify mounts of C derived from C 4 vs. C 3 sources in SOM frctions fter nine yers of experimentl wrming; nd (3) estimte the turnover rte of C in SOM frctions in wrmed soils. 2 Mterils nd methods 2.1 Site description The experiment ws locted on Kessler s Frm Field Lb (formerly Gret Plin piries, N, W), 4 km from the Normn cmpus of the University of Oklhom, US. Detiled description of the site chrcteristics nd design of the experiment hve been reported elsewhere (See Luo et l., 21). riefly, the site is tllgrss pririe primrily dominted by C 4 grsses (Schizchyrium scoprium nd Sorghstrum nutns) nd C 3 forbs (Solidgo rigid nd Solidgo nemorlis). S. scoprium comprises over 4 % of the plntcover, nd S. nutns over 2 % (Sherry nd Luo, unpublished dt). Men nnul temperture is 16. C with monthly men temperture of 3.1 C in Jnury nd 28. C in July. Men nnul precipittion is mm (Oklhom Meteorologicl Survey). The soil is silt lom with 36 % snd, 55 % silt, nd 1 % cly in the top 15 cm. The proportion of cly increses with depth. The soil is prt of the Nsh-Lucien complex, which is chrcterized by low permebility, high vilble wter cpcity, nd deep, modertely penetrble root zone (USD Soil Conservtion Service nd Oklhom griculturl Experiment Sttion, 1963). 2.2 Experimentl design This experiment used pired fctoril design with wrming s the min fctor nested by clipping fctor. Pirs of 2 2 m control nd wrmed plots were replicted six times. One plot hs been subjected to continuous 2 C wrming iogeosciences, 8, , 211

3 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe 1489 Tble 1. The physicl soil frction obtined from trllgrss pririe fter nine yers of wrming nd clipping. Soil frctions nlyzed for C, N concentrtions, δ 13 C, nd δ 15 N re denoted by bold numbers. ggregtes Density frctiontion (1.85 g cm 3 ) Sieving of broken-down ggregtes (1) Mcroggregtes (>2 µm) (1.1) free LF (1.2) hevy frction (1.2.1) msom (<53 µm) (1.2.2) ipom (>53 µm) (2) Mcroggregtes (25 2 µm) (2.1) free LF (2.2) hevy frction (2.2.1) msom (<53 µm) (2.2.2) ipom (>53 µm) (3) Microggregtes (53 25 µm) (3.1) free LF (3.2) hevy frction (3.2.1) msom (<53 µm) (3.2.2) ipom (>53 µm) (4) Microggregtes (<53 µm) LF = Light frction; POM = Prticle orgnic mtter; ipom = intr-ggregte POM; msom = minerl ssocited SOM. since 12 November 1999, while the control hs hd mbient temperture. One cm rdint infrred heter (Klglo Electronics Inc., ethlehem, P, US) with n output of 1 Wtt m 2 ws suspended t 1.5 m bove the ground in ech wrmed plot s the heting device. Reflector surfce of the heters were djusted so s to generte evenly distributed rdint input to soil surfce (Kimbll, 25). s result, temperture increments generted by the infrred heters were reltively even over the entire re of plots nd similr t different soil depths (Wn et l., 25). dummy heter with the sme shpe nd size s the infrred heter ws suspended t the sme height in the control plots to simulte the shding effect of the heter on the plnt cnopy. For ech pired plot, the distnce between wrmed nd control plots ws pproximtely 5 m to void heting of the control plots. The distnce between the pired plots vried from 2 to 6 m. Ech 2 2 m plot ws divided into four 1 1 m subplots. Plnts in the two digonl subplots were clipped t the height of 1 cm bove the ground yerly to remove biomss, usully in ugust. Clipping in this mnner effectively mimics griculturl hy mowing, widely prcticed lnd use in the southern Gret Plins. Usully frmers nd rnchers in the southern Gret Plins mow psture once or twice per yer, depending on rinfll. Clipping lso simultes biomss hrvest for biofuel feedstock production, lthough the study ws not originlly designed to study bioenergy production. The other two digonl subplots were left unclipped. The four tretments in the experiment were unclipped control (UC), clipped control (CC), unclipped wrming (UW), nd clipped wrming (CW). 2.3 Litter nd soil collection, nd soil frctiontion Litter on the soil surfce ws collected in clipped plots in ugust 28. s the unclipped subplots were designed to hve miniml disturbnces over the long term, no plnt mteril ws tken from the unclipped plots. Collected litter ws seprted into C 3 nd C 4 species ccording to morphologicl trits. Litter from C 3 nd C 4 species, nd mixed litter (i.e., mixed C 3 nd C 4 litter) from wrmed plots were selected for nlyses of N nd C concentrtion, nd stble C nd N isotopes (δ 13 C, nd δ 15 N). Soil smples were collected t depth of 2 cm with 4 cm dimeter soil corer in the fll of 28, nine yers fter the wrming begn. Soil smples were ir-dried, fter which lrge roots nd stone were removed by hnd. The method for ggregte seprtion nd size density frctiontions of free frction (LF), intr ggregte prticulte orgnic mtter (ipom), nd minerl-ssocited orgnic mtter (msom) ws dpted from Six et l. (1998). The frctiontion sequences re summrized in Tble 1. Four ggregte sizes were seprted using wet sieving through series of sieves (2, 25, nd 53 µm). 1 g ir dried smple ws submerged for 5 min in room temperture de-ionized wter, on top of the 2 µm sieve. ggregte seprtion ws chieved by mnully moving the sieve up nd down 3 cm with 5 repetitions during period of 2 min. fter the 2-min cycle, the stble >2 µm ggregtes were gently bck-wshed off the sieve into n luminum pn. Floting orgnic mteril (>2 µm) ws discrded, s this is by definition not considered SOM (Six et l., 1998). Wter nd soil tht pssed through the sieve were poured into the next two sieves (one t time) nd the sieving ws repeted in similr fshion, but floting mteril ws retined. Thus, four size frctions were obtined iogeosciences, 8, , 211

4 149 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe (>2 µm, 25 2 µm, µm nd <53 µm). The ggregtes were oven dried t (5 C), weighed, nd stored in glss jrs t room temperture. The density frctiontion ws crried out by using solution of 1.85 g cm 3 sodium polytungstte (SPT), following the method described in Six et l. (1998). subsmple (5 g) of ech oven-dried (11 C) ggregte size frction ws suspended in 35 ml of SPT nd slowly shken by hnd. The mteril remining on the cp nd sides of the centrifuge tube ws wshed into suspension with 1 ml of SPT. fter 2 min of vcuum (138 kp), the smples were centrifuged (125 g) t 2 C for 6 min. The floting mteril (light frction-lf) ws spirted onto 2 µm nylon filter, subjected to multiple wshings with deionized wter to remove SPT, nd dried t 5 C. The hevy frction (HF) ws rinsed twice with 5 ml of deionized wter nd dispersed in.5 % sodium hexmetphosphte by shking for 18 h on reciprocl shker. The dispersed hevy frction ws then pssed through 53 µm sieve nd the mteril remining on the sieve, i.e. the intr-ggregte prticulte orgnic mtter (ipom), ws dried (5 C) nd weighed. 2.4 Crbon, nitrogen, nd isotope nlyses Smples of litter nd soil were dried t 5 C to constnt weight nd then ground to pss through 2-mesh (.84 mm) sieves (Cheng et l., 26). The C nd N concentrtions nd δ 13 C nd δ 15 N were mesured for ll soil frctions nd litter mterils. Subsmples from ll frctions were treted with 1N HCL for 24 h t room temperture to remove ny soil crbontes (Cheng et l., 26). The C nd N concentrtion nd δ 13 C nd δ 15 N of soil nd litter were determined t University of rknss Stble Isotope Lbortory on Finnign Delt + mss spectrometer (Finnign MT, Germny) coupled to Crlo Erb elementl nlyzer (N15 CHN Combustion nlyzer, Crlo Erb Strumentzione, Miln, Itly) vi Finnign Conflo II Interfce. Crbon nd nitrogen contents of SOM frctions were clculted on n rel bsis, correcting for soil depth nd density. The crbon nd nitrogen isotope rtio of the soil frctions ws expressed s: ( / δ h X h ( X = X h 1 1 (1) X l )smple X l )stndrd where X is either crbon or nitrogen, h is the hevier isotope, l is the lighter isotope. oth CO 2 nd N 2 smples were nlyzed reltive to internl, working gs stndrds. Crbon isotope rtios ( 13 C) re expressed reltive to Pee Dee elemnite (δ 13 C =. ); nitrogen stble isotope rtios ( 15 N) re expressed reltive to ir (δ 15 N =. ). Stndrds (cetnilide nd spinch) were nlyzed fter every ten smples; nlyticl precision of the instrument ws ±.13 for δ 13 C nd ±.21 for δ 15 N. Differences in δ 13 C isotope composition due to photosynthetic pthwys llow for the proportion of soil C derived form C 3 or C 4 sources to be clculted using twocomprtment mixing-model (Del Gldo et l., 23; Cheng et l., 26): f = δ X δ δ δ 1% (2) where δ X is the δ 13 C of given frction isolted from the wrmed or control plots, δ nd δ re the isotope vlues of C 3 nd C 4 plnts from these plots, f is the frction of C 3 vegettion, nd f (1 f ) is the proportion derived from C 4 grsses. The frction of new C, f new, derived from the current vegettion in the wrmed soils fter nine yers of wrming is clculted by using the isotope mss blnce method (Mrin- Spiott et l., 29): f new = δ 2 δ δ 1 δ 1% (3) where δ 2 nd δ re δ 13 C vlues for SOM pools in the wrmed nd control plots nd δ 1 is the verge δ 13 C vlue of mixed litter to the SOM pool in the wrmed plots, on the ssumption tht in the pst 9 yr, no shift in rtio between C 3 /C 4 input in the control soil occurred. In Eqs. (2) nd (3), becuse δ (or δ 1 ), δ X (or δ 2 ), nd δ (or δ ) re independently mesured, the stndrd errors (SE) of f ssocited with the use of the mss-blnce pproch cn be clculted using prtil derivtives (Phillips nd Gregg, 21) s: ( ) δf 2 ( ) δf 2 ( ) δf 2 σf 2 = σδ 2 δ + σσ 2 δ X + σδ 2 X δ (4) This cn be reduced to: σ 2 f = 1 (δ X δ ) 2 [ σ 2 δ +f 2 σ 2 δ X +(1 f )σ 2 δ ] where σδ 2,σδ 2 X, nd σδ 2 represent vrinces of the men δ (or δ 1 ), δ X (or δ 2 ), nd δ (or δ ), respectively. The σ f is the SE of the proportion (f ) estimte (Phillips nd Gregg, 21). Furthermore, decomposition rte constnts (k) for old C (i.e. the C of the orgnic mtter previous to wrming) of different frction of SOM in the wrmed plots were clculted using the following eqution (Del Gldo et l., 23): ln(f old ) = kt (6) where f old = (1 f new ) is the proportion of old C, k is the net reltive decomposition rte constnt of old C, nd t is the ge of wrming. (5) iogeosciences, 8, , 211

5 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe Sttistics nlysis of vrince (NOV) of pired split-plot design (one pir of plots being considered block) ws conducted to exmine the effects of wrming on the soil orgnic C nd N contents, the δ 13 C nd δ 15 N vlues, C:N rtios in ll soil frctions, nd the weight distribution. The differences in soil orgnic C nd N contents, the δ 13 C nd δ 15 N vlues, nd C:N rtios between ggregte sizes nd density frctions were nlyzed using one-wy NOV. ll sttisticl nlyses were performed using Stt Soft s Sttistic, sttisticl softwre for Windows (Version 6., SttSoft, Inc., 21). Tble 2. The vlues of δ 13 C nd δ 15 N of litter of C 3, C 4 species nd mixed litter under the wrming tretment. Dt re expressed s men ± SE, n = 6. Different letters indicte sttisticl significnce t P <.5 between the two tretments. Vrible Species Control Wrming δ 13 C ( ) C ± ±.25 C ± ±.13 Mixed litter 16.19± ±.21 b δ 15 N ( ) C ± b C ± ±.17 b Mixed litter 5.81± ±.3 3 Results 3.1 The δ 13 C nd δ 15 N vlues of plnt litter The δ 13 C vlues of C 4 litter hd men vle of 12.74, while C 3 litter hd men vlue of Wrming significntly incresed the δ 13 C vlues of mixed litter on verge by 11.3 % (Tble 2). Wrming significntly enhnced the δ 15 N vlues on verge by 12.2 % for C 4 litter nd by 26.1 % for C 3 litter, wheres wrming slightly incresed the δ 15 N vlues on verge by 8.4 % for mixed litter (Tble 2). 3.2 Whole soil C nd N dynmics Whole soil (i.e., totl soil) orgnic C nd N contents rnged from 2371 to 277 g C m 2, nd 284 to 312 g N m 2, respectively, cross ll tretments. No significnt differences in C nd N content, or C:N rtios mong tretments were found (Tble 3). Nine-yer wrming significntly incresed the δ 13 C signture of SOM for both clipped nd unclipped plots. On verge, the wrmed plot soils were 1.3 more enriched in 13 C thn the control plots. sed on these dt, wrming incresed the frction of C 4 -derived C on verge by 11.6 % (Tble 4). Wrming incresed the δ 15 N vlues of SOM in clipped plots (Tble 4). 3.3 Size distribution, C nd N contents, nd δ 13 C nd δ 15 N of soil ggregtes ggregte distribution ws not significntly effected by wrming or clipping (Fig. 1). Wrming significntly decresed soil orgnic C nd N content in microggregtes (<25 µm) in clipped plots but not in other ggregte size clsses (Tble 4). No significnt differences in C:N rtios were found cross ggregte size nd tretments (Fig. 2). Generlly, mcroggregtes (>25 µm) contined significntly more C nd N (78 84 %) thn microggregtes in ll tretments (Tble 4). Wrming resulted in no significnt increse in the δ 13 C vlues of ll ggregte sizes compred to the control plots (Tble 5), indicting tht wrming possibly stimulted input of C 4 -derived C (Fig. 3). Wrming-induced increses in the frction of C 4 -derived C rnged from 5.3 % to 1.8 % mong ggregte size clsses, with the highest in the >2 µm mcroggregte clss (Fig. 3b). The δ 15 N vlues of ech ggregte size were significntly more enriched in the wrmed plots thn the control (Tble 5). There were no significnt differences in δ 15 N vlues between ggregtes (>53 µm), but microggregtes (<53 µm) hd significntly higher δ 15 N vlue thn ll other ggregte size clsses (Tble 5). 3.4 Density frction: C nd N contents, nd δ 13 C nd δ 15 N in LF, ipom nd msom LF ccounted for the smllest frction of totl SOM, wheres msom ccounted for the lrgest (74 79 %) frction of totl SOM in ll ggregte size clsses cross ll tretments (Tble 4). Wrming significntly decresed soil orgnic C nd N contents in ipom in mcroggregtes (>2 µm) in clipped plots but not in ny other SOM clsses. C nd N content in msom nd ipom significntly decresed with size clss, wheres the highest C nd N contents in LF were found in 2 25 µm mcroggregtes (Tble 4). C:N rtio significntly incresed in LF of µm microggregtes but not in ny other SOM clsses under wrming in comprison to control (Fig. 4). C:N rtios decresed from LF to ipom to msom in ll ggregte clsses cross tretments (Fig. 4). Wrming resulted in n increse in δ 13 C vlues cross ll density frctions in ech ggregte size (Tble 5). The wrming-induced increse ws significnt for δ 13 C vlues from LF in >2 µm mcroggregtes in clipped plots. The δ 13 C vlues were generlly more enriched in msom thn LF nd ipom cross ggregte sizes nd tretments (Tble 5). Wrming-induced increse in C 4 -derived C ws the highest for LF in >2 µm mcroggregtes mong ll ggregtes nd density frctions (Fig. 5b). In generl, wrming stimulted more C 4 -derived C input into LF thn ipom nd msom cross ll ggregte sizes, nd more into lrger thn smller ggregte sizes (Fig. 5b). iogeosciences, 8, , 211

6 1492 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe Tble 3. Soil orgnic C nd N content, 13 C- nd 15 N signture, frction of C 4 -derived C (f ), nd the C:N rtio of whole soils in tllgrss pririe ( 2 cm depth) fter nine yers of wrming nd clipping. Dt re expressed s men ± SE, n = 6. Different letters indicte sttisticl significnce t P <.5 mong the four tretments. bbrevitions: UC, unclipped control; CC, clipped control; UW, unclipped wrming; CW, clipped wrming. Tretment C (g C m 2 ) δ 13 C ( ) f (%) N (g N m 2 ) C:N δ 15 N ( ) UC 2529± ±3.1 b 6.2±7.3 b 299±55 8.4±.7 2.7±1. UW 2693± ± ±6.9 33±56 8.9±2 2.93±.7 CC 2371± ±2.8 b 62.8±5.7 b 312±41 7.6±.6 2.6±.6 b CW 277± ± ± ±48 9.5±1 3.53±.7 Tble 4. Soil orgnic C nd N content of soil frctions under four tretments fter nine yers of wrming nd clipping. Dt re expressed s men ± SE, n = 6. Different letters indicte sttisticl significnce t P <.5 mong the four tretments. See Tble 3 for bbrevitions. Frctions C (g C m 2 ) N (g N m 2 ) UC UW CC CW UC UW CC CW >2 µm 1214± ± ± ± ± ± ± ± 19.3 LF 56 ± 7 43 ± 5 31 ± 4 b 2 ± 3 b 2.4 ± ± ± ±.3 ipom 162 ± ± ± ± 11 b 11.1 ± ± ± ± 1.1 b msom 99 ± ± ± ± ± ± ± ± µm 122 ± ± ± ± ± ± ± ± 22.7 LF 88 ± ± 1 85 ± 9 88 ± ± ± ± ±.8 ipom 51 ± 9 73 ± ± 8 64 ± ± ± ± ±1.1 msom 879 ± ± ± ± ± ± ± ± µm 266 ± 57 b 286 ± 73 b 344 ± ± 63 b 37.4 ± 8.3 b 33.8 ± 1.1 b 44.4 ± ± 7.2 b LF 19 ± 3 11 ± 4 2 ± 3 19 ± ±.3.7 ± ± ±.4 ipom 25 ± 5 31 ± 6 25 ± 7 19 ± ±.4 2. ±.6 2. ± ±.4 msom 218 ± ± ± ± ± ± ± 4.2 b 33.8 ± 3.2 <53 µm 36.4 ± 13 b 45.5 ± 16 b 58.5 ± ± 12 b 5.5 ± 1.2 b 5.8 ±.9 b 7.4 ± ±.8 b LF = Light frction; POM = Prticle orgnic mtter; ipom = intr-ggregte POM; msom = minerl ssocited SOM. In generl, wrming significntly incresed δ 15 N vlues of LF, ipom, nd msom cross ggregte sizes (Tble 5). msom hd the highest δ 15 N vlue nd LF hd the lowest δ 15 N vlue mong density frctions in ll ggregte sizes cross tretments (Tble 5). 3.5 Soil C turnover Experimentl wrming stimulted both new C input from C 4 photosynthesis nd decy rte of old C (Tble 6). New C inputs in whole soil were greter thn those in ll ggregtes. New C inputs were greter in LF thn in ipom, with the gretest in >2 µm mcroggregtes. Overll, new C inputs in soil frctions decresed for smller ggregtes except for msom (Tble 6). ccordingly, decy rtes for old C in whole soil were fster thn those for ll ggregtes. The fstest decy rtes were found in LF in >2 µm mcroggregtes, nd LF hd greter decy rte thn ipom for ll SOM clsses (Tble 6). 4 Discussion Wrming did not significntly increse totl soil C nd N storge, but other spects of SOM dynmics did chnge. Wrming effects t our study site were previously chrcterized by incresed biomss growth nd NPP, shift towrd greter C 4 species dominnce, nd incresed litter input (Wn et l., 25; Luo et l., 29; Cheng et l., 21). Our stble isotope nlysis in this present study confirmed tht the δ 13 C bundnce in SOM in the wrmed soils ws more enriched thn in the control soils (Tble 3), resulting from higher contribution of C 4 residuls. Indeed, wrminginduced increses in C 4 plnts nd decreses in C 3 plnts led to increses in the frction of C 4 -derived C on verge by 11.6 % (Tble 3). However, increses in C inputs nd chnges in SOM qulity fter 9-yr wrming did not significntly increse totl soil orgnic C nd N content (Tble 3; Niu et l., 21). iogeosciences, 8, , 211

7 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe 1493 Weight % () ggregte size clss ( m) 37 UC UW CC CW C C C C > <53 Fig. 1. Weight distribution mong ggregte size clsses under four tretments fter nine yers of wrming nd clipping. Vlues followed by different lowercse letter re significntly different within ggregte size clsses mong Fig. 1. tretments. Vlues followed by different cpitl letter re significntly different mong ggregte size clsses under tretments. bbrevitions: UC, unclipped control; CC, clipped control; UW, unclipped wrming; CW, clipped wrming. C: N rtios ggregte size clsses UC UW CC CW >2 m 2-25 m m <53 m Frction of C 4 -derived C (%) Wrming-induced increses in frction of C 4 -derived C (%) () b b (b) b b ggregte size clss ( m) UC UW CC CW b b >2 μm 2-25 μm μm <53 μm Fig. 3. Frction of C 4 -derivedfig. C of 3. ggregte size clsses under four tretments fter nine yers of wrming nd clipping (), nd wrming-induced increses in the frction of C 4 -derived C of ggregte size clsses in wrmed soils (b). Vles followed by different lowercse letter re significntly 39 different within ggregte size clsses mong tretments. Vlues followed by different cpitl re significntly different mong ggregte size clsses under tretments. C C Fig. 2. C:N rtios of ggregte size clsses under four tretments fter nine yers of wrming nd clipping. See Fig. 1 for the explntion of the symbols. Fig. 2. The min processes tht control soil SOM storge under wrming re determined by blnce between litter input nd soil C respirtion (e.g., Shw nd Hrte, 21; Fissore et l., 28). Our previous study found wrming incresed soil respirtion (Zhou et l., 27), similr to other wrming studies (e.g., Rustd et l., 21; Fontine et l., 24). Furthermore, Wynn nd ird (27) hve found tht the ctive pool of SOM derived from C 4 plnts decomposes fster thn the totl pool of SOM. Wrming-induced increses in C 4 -derived C in SOM pool (Tble 3) likely ccelertes decy rtes of SOM in wrmed soils. 38Thus, unchnged SOM stock in our wrming experiment possibly resulted from concur- rent increses in litter inputs to soil (Luo et l., 29; Cheng et l., 21) nd decomposition rtes (Zhou et l., 27). Similrly, wrming did not significntly chnge soil ggregte size distribution (Fig. 1) nd soil orgnic C nd N contents (Tble 4). lthough wrming induced shift from C 3 to C 4 plnt species, which my effect SOM qulity nd input (Luo et l., 29; Cheng et l., 21), wrming did not effect the level of soil ggregtion (Fig. 1). These results re in greement with Scott (1998), who reported grss species hd no effect on size-distribution of soil ggregtes or orgnic mtter concentrtion. However, the C nd N content of different size frctions re primrily controlled by the mount of ech ggregte size (Elliott, 1986; Vn Groenigen et l., 22). We found tht mcroggregtes (>25 µm) contined significntly more C nd N thn microggregtes (Tble 4), indicting tht orgnic C nd N content generlly decrese iogeosciences, 8, , 211

8 1494 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe Tble 5. δ 13 C nd δ 15 N vlues of soil frctions under four tretments fter nine yers of wrming nd clipping. Dt re expressed s men ± SE, n = 6. Different letters indicte sttisticl significnce t P <.5 mong the four tretments. See Tble 3 for bbrevitions. Frctions δ 13 C ( ) δ 15 N ( ) UC UW CC CW UC UW CC CW >2 µm 18.3± ± ± ± ±.7 b 3.6± ±.9 b 3.37±1.1 LF 19.1±2.6 b 17.1± ±2.9 b 16.2± ±1.1 b 1.19±1.2 b.96±.9 b.72±.7 ipom 19.5± ± ± ± ±.6 b 2.7± ±.6 b 2.21±.7 msom 17.8± ± ± ± ± ± ± ± µm 17.9± ± ± ± ±.7 b 3.4± ±.9 b 3.5±.7 LF 19.6± ± ± ± ±.9 b 1.85±.7 b 1.96±1. b.62±1.2 ipom 18.9± ± ± ± ±.9 b 1.11±1.1 b 1.11±1. b 2.61±.8 msom 17.9±3. 17.± ± ± ±.6 b 3.93±.7 b 3.92±.9 b 4.92± µm 17.7± ± ± ± ±.8 b 3.53± ± ±.8 LF 19.4± ± ± ±1..12±.4.11±.2.32±.8.11±.7 ipom 19.4± ± ± ±2.4.76±.9 b 1.3±1.3 b 1.13±.9 b 2.86±1. msom 17.2± ± ±2. 17.± ±.7 b 4.8±.6 b 3.79±.7 b 5.13±.9 <53 µm 17.3± ± ± ± ± ± ±.9 b 4.54±.8 LF = Light frction; POM = Prticle orgnic mtter; ipom = intr-ggregte POM; msom = minerl ssocited SOM. Tble 6. New C input (f new ), nd decy rte (k, yr 1 ) of old C of soil frctions ( 2 cm) in wrmed soils fter nine yers of experimentl wrming. Frction f new, (%) Decy rte (k) of old C Whole soil 33.7±2.6.46±.3 >2 µm 3.1±3.4.4±.3 LF 36.3± ±.2 ipom 2.4±1.7.25±.3 msom 16.±2.3.19± µm 17.3±19.21±.2 LF 2.±2.7.25±.3 ipom 12.4±1.5.15±.1 msom 22.4±2.8.28± µm 16.2±1.3.2±.1 LF 11.±.9.13±.1 ipom 4.9±.6.6±.1 msom 11.7±1.3.14±.1 <53 µm 1.7±.8.12±.1 LF = Light frction; POM = Prticle orgnic mtter; ipom = intr-ggregte POM; msom = minerl ssocited SOM. with decresing ggregte size (Elliott, 1986; Puget et l., 1995). dditionlly, wrming incresed C 4 -derived C in ll ggregte size with the highest C 4 -derived C in >2 µm mcroggretes (Fig. 3b), supporting the evidence tht new C is incorported more rpidly in corse SOM thn in fine SOM frctions (Desjrdins et l., 24; Schwendenmnn nd Pendll, 26). It is well known tht LF, ipom, nd msom hve different chemicl compositions nd turnover times (Trumbore, 2; Wynn nd ird, 27). Higher C:N rtios of LF reflect more recent litter inputs, while msom hd much lower C:N rtios (Fig. 4). Decresing C:N rtios in soil C frctions hve been ssocited with incresing SOM decomposition nd minerl ssocition (John et l., 25; Mrin-Spiott et l., 29). Moreover, the 15 N vlues enriched from LF to ipom to msom provided further evidence to support the degree of decomposition nd humifiction of SOM. Similr to other studies (Lio et l., 26; Mrin-Spiott et l., 29), the 15 N vlues of the <53 µm microggretes were higher thn other ggregtes (Tble 5). In generl, low 15 N vlues re relted to recent orgnic mtter inputs (litter, roots), wheres high 15 N vlues in silts + clys (<53 µm) re ssocited with incresing SOM trnsformtion nd humifiction. The 15 N vlues of soil frctions were more enriched in wrmed soil thn in the control soil (Tble 5). It suggests tht the nturl bundnce of 15 N in soil becomes enriched in 15 N in the wrmed soil by the process of N loss from soil through incresed minerliztion nd possibly nitrte leching compred to control sites (Rustd et l., 21; ijoor et l., 28). Indeed, wrming resulted in no significnt decreses in totl soil N pools, on verge by 1 % in the clipped plots in our study (Tble 3). SOM decomposition typiclly results in n enrichment in δ 15 N (Lio et l., 26; Mrin-Spiott et l., 29), so higher 15 N for the wrmed soils reltive to the control soils in our study would suggest tht the SOM in the wrmed soils is more decomposed thn the control soils. Even though there were no significnt increses in SOM pools fter nine yers of wrming, isotopic mesurements nd turnover time estimtes suggest different C decy rtes of iogeosciences, 8, , 211

9 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe 1495 C: N rtio ggregte size: b b >2 m 2-25 m b UC UW CC CW m LF ipom msom LF ipom msom LF ipom msom SOM clsses Fig. 4. C:N rtios of LF, ipom, nd msom of ggregte size clsses under four tretments fter nine yers of wrming nd clipping. Vlues followed by different lowercse letter re significntly different within SOM clsses mong tretments of ech ggregte size. Vlues followed by different cpitl re significntly different mong SOM clsses under tretments of ech ggregte size. See Fig. 1 for bbrevitions. Fig. 4. SOM frctions in wrmed soils. Incresed new C inputs from plnt residue could result in fster decomposition of SOM (Dijkstr nd Cheng, 27). The decy rtes for old C in whole soil were fster thn ll ggregtes due to greter new C inputs (Tble 6). This finding supports the evidence tht soil ggregtes physiclly protect certin SOM frctions, resulting in pools with longer turnover times (Six et l., 1998). lthough LF generlly represent only smll proportion of totl soil C (Gregorich et l., 26), chnges in C stocks following chnges in species composition cn be more pronounced in LF compred to bulk soil (Schwendenmnn nd Pendll, 26). Our results showed tht wrming-induced increse in C 4 -derived C in LF ws lrger thn in ipom nd msom in ll ggregte sizes (Fig. 5). Menwhile, orgnic mtter in LF frctions is redily ccessible to microbes, s reflected by their initil rpid loss (Fontine et l., 24; Pendll et l., 24). Wrming cused no significnt decreses in C content in LF in ll SOM size frctions except in 2 25 µm ggregtes in clipped plots (Tble 4), possibly due to rpid loss of lbile substrtes in LF (Tble 6). With incresing degree of decomposition, orgnic mtter my be trnsferred to more stbilized soil frctions. In contrst to rpid decomposition of C 4 -derived C from LF, some C 4 -derived C remined in ipom frctions with slower turnover rtes (Tble 6). ipom nd msom ccounted for the lrgest frction of soil orgnic C nd N contents in ll sized ggregtes (Tble 4). Wrming did not significntly effect soil orgnic C nd N contents in ipom nd msom frctions, supporting the view tht the hevy nd minerl ssocited reclcitrnt frctions tht mintin physicl nd chemicl stbiliztion (e.g., Six et l., 1998). To conclude, we found tht nine yers of experimentl wrming cused no significnt increses in soil orgnic C nd N content in ny soil frction t our site. Wrming did not significntly effect soil ggregte distribution nd stbility. However, wrming did increse C 4 -derived C input into ll frctions, prticulrly in LF of ll ggregte size clsses. Significnt C loss in whole soil nd lbile components of LF under wrming likely offset incresed overll SOM inputs. Under wrming, 15 N vlues of soil frctions were more enriched thn in the controls, indicting incresed N trnsformtion under wrming. C:N rtios nd differences in nturl bundnce of δ 13 C nd δ 15 N in SOM frctions re ssocited with n incresing degree of decomposition cross 4density frctions with incresing minerl ssocition. The δ 13 C vlue of SOM is controlled by multiple fctors, including hydrology, soil temperture, substrte, nd vegettion, but turnover times bsed on nturl bundnce stble isotope methods tend to be more relted to recent C inputs nd C pools ssocited with the C 3 /C 4 vegettion type conversion (Six nd Jstrow, 22). Lck of vribility in the controls in this study might not provide rigorous sttisticl tests of wrming effects on chnges in δ 13 C of SOM. However, environment-induced chnges in δ 13 C of SOM re smll reltive to chnges cused by C 3 /C 4 litter inputs, which do not strongly influence δ 13 C to trce SOM dynmics following chnges in litter inputs (Wedin et l., 1995; Six nd Jstrow, 22). Our results suggest tht shifts in species composition iogeosciences, 8, , 211

10 1496 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe 12 () ggregte size: UC UW CC CW Frction of C 4 -derived C (%) >2 m 2-25 m b b b b b C C C C b b C b b C C b m b b 2 Wrming-induced increses in frction of C 4 -derived C (%) (b) >2 m LF ipom C msom 2-25 m LF ipom msom LF m ipom msom SOM clsses Fig. 5. Frction of C 4 -derived C in the LF, ipom, nd msom of ggregte size clsses under four tretments fter nine yers of wrming nd clipping (), nd wrming-induced increses in the frction of C 4 -derived C in LF, ipom, nd msom of SOM clsses in the wrmed soils (b). See Fig. 4 for the explntion of the symbols. Fig. 5. under wrming could potentilly modify SOM qulity nd decomposition nd consequently effect ecosystem functions. Physicl frctiontion methods combined with isotope nlyses in our study were n ttempt to better understnd SOM dynmics in response to globl wrming by cknowledging tht SOM consists of continuum of substrtes with different turnover times. To ccurtely predict the effects of globl wrming on ecosystem processes, we need ecologicl models nd long-term experiments to project future chnges in ecosystem C nd N cycles in response to multifctor globl chnge. 41cknowledgements. This reserch ws finncilly supported by Ntionl Science Foundtion (NSF) under grnts DE nd DE ; by the Office of Science (ER), Deprtment of Energy, grnts no. DEFG2-6ER64319 nd through the Midwestern Regionl Center of the Ntionl Institute for Climtic Chnge Reserch t Michign Technologicl University, under wrd number DE-FC2-6ER We thnk Meng Lu, Shenfei Fei, Xiolei Yn for ssistnce in the field, Thoms L. Millicn for ssistnce with lbortory nlyses. Edited by: J. Leifeld iogeosciences, 8, , 211

11 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe 1497 References n, Y., Wn, S., Zhou, X., Subed,., Wllce, L. L., nd Luo, Y.: Plnt nitrogen concentrtion, use efficiency, nd contents in tllgrss pririe ecosystem under experimentl wrming, Glob. Chnge iol., 11, , 25. uerswld, K., Wittmer, M. H. O. M., Männel, T. T., i, Y. F., Schäufele, R., nd Schnyder, H.: Lrge regionl-scle vrition in C3/C4 distribution pttern of Inner Mongoli steppe is reveled by grzer wool crbon isotope composition, iogeosciences, 6, , doi:1.5194/bg , 29. lesdent, J.: The significnce of orgnic seprtes to crbon dynmics nd its modelling in some cultivted soils, Eur. J. Soil Sci., 47, , tjes, N. H. nd Sombroek, W. G.: Possibilities for crbon sequestrtion in tropicl nd subtropicl soils, Glob. Chnge iol., 3, , ijoor, N. S., Czimczik, C. I., Ptki, D., nd illings, S..: Effects of temperture nd fertiliztion on nitrogen cycling nd community composition of n urbn lwn, Glob. Chnge iol., 14, , 28. Cheng, X., Luo, Y., Chen, J., Lin, G., Chen, J., nd Li,.: Shortterm C 4 plnt Sprtin lterniflor invsions chnge the soil crbon in C 3 plnt-dominted tidl wetlnds on growing esturine Islnd, Soil iol. iochem., 38, , 26. Cheng, X., Luo, Y., Su,., Zhou, X., Niu, S., Sherry, R. Weng, E., nd Zhng, Q.: Experimentl wrming nd clipping ltered litter crbon nd nitrogen dynmics in tllgrss pririe, gr. Ecosyst. Environ., 138, , 21. Dvidson, E.. nd Jnssens, I..: Temperture sensitivity of soil crbon decomposition nd feedbcks to climte chnge, Nture, 44, , 26. Dy, T.., Ruhlnd, C., nd Xiong, F. S.: Wrming increses boveground plnt biomss nd C stocks in vsculr-plntdominted ntrctic tundr, Glob. Chnge iol., 14, , 28. Dwson, T. E., Mmbelli, S., Plmboeck,. H., Templer, P. H., nd Tu, K. P.: Stble isotopes in plnt ecology, nnu. Rev. Ecol. Syst., 33, , 22. Del Gldo, I., Six, J., Peressotti,., nd Cotrufo, M. F.: ssessing the impct of lnd-use chnge on soil sequestrtion in griculture soils by mens of orgnic mtter frction nd stble C isotopes, Glob. Chnge iol., 9, , 23. Desjrdins, T., rros, E., Srrzin, M., Girrdin, C., nd Mriotti,.: Effects of forest conversion to psture on soil crbon content nd dynmics in rzilin mzoni, gr. Ecosyst. Environ., 13, , 24. Dijkstr, F.. nd Cheng, W.: Interctions between soil nd tree roots ccelerte long-term soil crbon decomposition, Ecol. Lett., 1, , 27. Elliott, E. T.: ggregte structure nd crbon, nitrogen, nd phosphorus in ntive nd cultivted soils, Soil Sci. Soc. m. J., 5, , Fissore, C., Girdin, C. P., Kolk, R., Trettin, C. C., King, G. M., Jurgensen, M. F. rton, C. D., nd Mcdowell, S. D.: Temperture nd vegettion effects on soil orgnic crbon qulity long forest men nnul temperture grdient in North meric, Glob. Chnge iol., 14, , 28. Fontine, S., rdoux, G., bbdie, L., nd Mriotti,.: Crbon input to soil my decrese soil crbon content, Ecol. Lett., 7, , 24. Gregorich, E. G., ere, M. H., Mckim, U. F., nd Skjemstd, J. O.: Chemicl nd biologicl chrcteristics of physiclly uncomplexed orgnic mtter, Soil Sci. Soc. m. J., 7, , 26. Jstrow, J. D.: Soil ggregte formtion nd the ccrul of prticulte nd minerl-ssocited orgnic mtter, Soil iol. iochem., 28, , John,., Ymshit, T., Ludwig,., nd Fless, H.: Storge of orgnic crbon in ggregte nd density frctions of silty soils under different types of lnd use, Geoderm, 128, 63 79, 25. Kimbll,..: Theory nd performnce of n infrred heter for ecosystem wrming, Glob. Chnge iol., 11, , 25. Krmer, M. G., Sollins, P., Sletten, R. S., nd Swrt, P. K.: N isotope frctiontion nd mesures of orgnic mtter ltertion during decomposition, Ecology, 84, , 23. Ll, R.: Sequestrtion of tmospheric CO 2 in globl crbon pools, Energ. Environ. Sci., 1, 86 1, 28. Lio, J. D., outton, T. W., nd Jstrow, J. D.: Orgnic mtter turnover in soil physicl frctions following woody plnt invsion of grsslnd: Evidence from nturl 13 C nd 15 N, Soil iol. iochem., 38, , 26. López-Ullo, M., Veldkmp, E., nd de Koning, H. G. J.: Soil crbon stbiliztion in converted tropicl pstures nd forests depends on soil type, Soil Sci. Soc. m. J., 69, , 25. Luo, Y., Wn, S., Hui, D., nd Wllce, L. L.: cclimtiztion of soil respirtion to wrming in tllgrss pririe, Nture, 413, , 21. Luo, Y., Sherry,., Zhou, X., nd Wn, S.: Terrestril crboncycle feedbck to climte wrming: experimentl evidence on plnt regultion nd impcts of biofuel feedstock hrvest, Glob. Chnge iol. ioen., 1, 62 74, 29. IPCC.: Climte Chnge 27: The Scientific sis, Cmbridge University Press, New York, US, 27. Mi, S. M. F., Ogle, S. M., Cerri, C. E. P., nd Cerri, C. C.: Soil orgnic crbon stock chnge due to lnd use ctivity long the griculturl frontier of the southwestern mzon, rzil, between 197 nd 22, Glob. Chnge iol., 16, , 21. Mrin-Spiott, E., Silver, W. L., Swnston, C. W., nd Ostertg, R.: Soil orgnic mtter dynmics during 8 yers of reforesttion of tropicl pstures. Glob. Chnge iol., 15, , 29. Niu, S., Sherry, R.., Zhou, X., Wn, S., nd Luo, Y.: Nitrogen regultion of the climte-crbon feedbck: evidence from longterm globl chnge experiment, Ecology, 91, , 21. Pendll, E., ridghm, S., Hnson, P. J., Hungte,., Kicklighter, D. W., Johnson, D. W., Lw,. E., Luo, Y., Megonigl, J. P., Olsrud, M., Ryn, M. G., nd Wn, S.: elow-ground process responses to elevted CO 2 nd temperture: discussion of observtions, mesurement methods, nd models, New Phytol., 162, , 24. Phillips, D. L. nd Gregg, J. W.: Uncertinty in source prtitioning using stble istopes, Oecologi, 127, , 21. Puget, P., Chenu, C., nd lesdent, J.: Totl nd young orgnic mtter distributions in ggregte of silty cultivted soils, Eur. J. Soil Sci., 46, , Robinson, D.: 15 N s n integrtor of the nitrogen cycle, Trends Ecol. Evol., 16, , 21. Rustd, L. E., Cmpbell, J. L., Mrion, G. M., Norby, R. J., Mitchell, M. J., Hrtley,. E., Cornelissen, J. H. C., nd Gurevitch, J.: met-nlysis of the response of soil respirtion, net iogeosciences, 8, , 211

12 1498 X. Cheng et l.: Soil orgnic C nd N dynmics in North meric tllgrss pririe nitrogen minerliztion, nd boveground plnt growth to experimentl ecosystem wrming, Oecologi, 126, , 21. Schwendenmnn, L. nd Pendll, E.: Effects of forest conversion into grsslnd on soil ggregte structure nd crbon storge in Pnm: evidence from soil crbon frctiontion nd stble isotopes, Plnt Soil, 288, , 26. Scott, N..: Soil ggregtion nd orgnic mtter minerliztion in forests nd grsslnds: Plnt species effects, Soil Sci. Soc. m. J., 62, , Shw, M. R. nd Hrte, J.: Response of nitrogen cycling to simulted climte chnge: differentil responses long sublpine ecotone, Glob. Chnge iol., 7, , 21. Six, J. nd Jstrow, J. D.: Orgnic mtter turnover, in: Encycloedi of soil science, edited by: Ll, R., Mrcel Dekker, New York, NY, US, , 22. Six, J., Elliott, E. T., Pustin, K., nd Dorn, J. W.: ggregtion nd soil orgnic mtter ccumultion in cultivted nd ntive grss soils, Soil Sci. Soc. m. J., 62, , Six, J., Pustin, K., Elliott, E. T., nd Combrink, C.: Soil structure nd orgnic mtter: I. Distribution of ggregte-size clsses nd ggregte-ssocited crbon, Soil Sci. Soc. m. J., 64, , 2. Trumbore, S. E.: ge of soil orgnic mtter nd soil respirtion: rdiocrbon constrints on belowground C dynmics, Ecol. ppl., 1, , 2. Vn Groenigen, K. J., Hrris,., Horwth, W. R., Hrtwig, U., nd Vn Kessel, C.: Linking sequestrtion of 13 C nd 15 N in ggregtes in psture soil following 8 yers of elevted tmospheric CO 2, Glob. Chnge iol., 8, , 22. Von Fischer, J. C., Tieszen, L. L., nd Schimel, D. S.: Climte controls on C 3 vs. C 4 productivity in North mericn grsslnd from crbon isotope composition of soil orgnic mtter, Glob. Chnge iol., 14, , 28. Wn, S., Hui, D., Wllce, L., nd Luo, Y.: Direct nd indirect effects of experimentl wrming on ecosystem crbon processes in tllgrss pririe, Globl iogeochem. Cy., 19, G214, doi:1.129/24g2315, 25. Wedin, D., Tieszen, L. L., ewey,., nd Pstor, J.: Crbon isotope dynmics during grss decompostion nd soil orgnic mtter formtion, Ecology, 76, , Welker, J. M., Fhnestock, J. T., Henry, G. H. R., O De, K. W., nd Chimner, R..: CO 2 exchnge in three Cndin High rctic ecosystems: response to long-term experimentl wrming, Glob. Chnge iol., 12, , 24. Wynn, J. nd ird, M.: C 4 -derived soil orgnic crbon decomposes fster thn its C 3 counterprt in mixed C 3 /C 4 soils, Glob. Chnge iol., 13, , 27. Zhou, X., Wn, S., nd Luo, Y.: Source components nd internnul vribility of soil CO 2 efflux under experimentl wrming nd clipping in grss ecosystem, Glob. Chnge iol., 13, , 27. iogeosciences, 8, , 211