Biogeochemistry (2017) 132:293 306 DOI 10.7/s10533-017-0302-4 Differences in sptil nd temporl root lifespn of three Stip grsslnds in northern Chin W. M. Bi. M. Zhou. Y. Fng. W. H. Zhng Received: 20 April 2016 / Accepted: 27 Jnury 2017 / Pulished online: 4 Ferury 2017 Springer Interntionl Pulishing Switzerlnd 2017 Astrct Lifespn of fine root plys n importnt role in regulting cron (C) cycling in terrestril ecosystems. Determintion of root lifespn nd elucidtion of its regultory mechnism in different plnt communities re essentil for ccurte prediction of C cycling from ecosystem to regionl scles. There re three mjor types of grsslnds in the temperte steppes of Inner Mongoli, ech dominted y different species of common grss: Stip krylovi, Stip grndis, nd Stip reviflor. There hve een no studies to compre the root dynmics mong the three types of grsslnds. In the present study, we determined root lifespn of the three grsslnds using the rhizotron technique. We found tht root lifespn differed sustntilly mong the three mjor types of grsslnds within the temperte steppes of Inner Mongoli, such tht root lifespn of S. reviflor [ S. grndis [ S. krylovii grsslnds. Root lifespn cross the three types of grsslnds displyed similr temporl pttern, such tht root lifespn followed the order of utumn-orn roots [ summer-orn roots [ - spring-orn roots. The sptil nd temporl differences in root lifespn cross the three types of grsslnds were minly relted to contents of solule sugrs in roots of the dominnt species nd BNPP/ ANPP rtio of the communities. The differences in root lifespn cross the mjor types of grsslnds nd different sesons highlight the potentil importnce of tking these differences into ccount in models of future cron cycling nd climte chnge. Keywords Inner Mongoli grsslnds Rhizotron Root longevity Root turnover Temperte steppe Responsile Editor: Ssh C. Reed. W. M. Bi M. Zhou Y. Fng W. H. Zhng (&) Stte Key Lortory of Vegettion nd Environmentl Chnge, Institute of Botny, The Chinese Acdemy of Sciences, Beijing, Chin e-mil: whzhng@ics.c.cn W. M. Bi W. H. Zhng Reserch Network of Glol Chnge Biology, Beijing Institutes of Life Science, Chinese Acdemy of Sciences, Beijing, Chin M. Zhou Y. Fng W. H. Zhng Grdute School of Chinese Acdemy of Sciences, Yuqunlu, Beijing 049, Chin Introduction Roots re mjor sources of the orgnic cron pool in soils. Approximtely hlf of the nnul photosynthetic products of plnts cn e llocted into roots (Vogt et l. 1986; Hendrick nd Pregitzer 1993; Peek 2007). Root turnover is n importnt trit tht links plnts to the orgnic cron pool in soils (Loy et l. 2004; Strnd et l. 2008). Thus, knowledge out the dynmic processes ssocited with root turnover is of importnce for our understnding of cron
294 Biogeochemistry (2017) 132:293 306 (C) cycling within the terrestril ecosystems (Aerts et l. 1992; Nory et l. 2004). However, these processes remin poorly understood (Iversen nd Brien 2010; Brdgett et l. 2014), mking the modeling-sed predictions of C cycling within terrestril ecosystems uncertin (McCormck et l. 2013; Wrren et l. 2015). Root turnover is dependent upon root lifespn. However, root lifespn is highly vrile within nd cross iomes, nd mong species in given ecosystem, nd root lifespn for n individul plnt species cn lso vry in different sptil nd temporl scles (Eissenstt nd Yni 1997; Burton et l. 2000; Withington et l. 2006; Bi et l. 2008; Peek 2007; Guo et l. 2008; McCormck nd Guo 2014). Moreover, root lifespn lso vries most strongly y root order within given plnt (Guo et l. 2008). Severl extrinsic nd intrinsic fctors hve een shown to hve impcts on root lifespn, including extrinsic fctors such s temperture, precipittion, soil moisture nd nutrient vililities in soils. In generl, root lifespn is often reduced y high temperture due to enhnced respirtion rte (Mjdi nd Ohrvik 2004). Plnt growth is often limited y wter vilility, such tht moderte increse or reduction in wter supply cn prolong root lifespn due to enhnced plnt productivity (Joslin et l. 2000). Root lifespn is lso often reduced with incresing vilility of nutrients s plnts cn cquire more nutrients y enhncing root turnover (Vn der Krift nd Berendse 2002). Intrinsic fctors tht ffect root turnover include root nutrient contents, root orders, specific root length, root solule sugr contents nd the rtio of elowground net primry production (BNPP) to oveground NPP (ANPP) (Peek 2007; Chen nd Brssrd 2013; McCormck nd Guo 2014). Roots with higher nutrient contents, lower orders of rnching system, smller root dimeter nd greter specific root length often exhiit shorter lifespn (McCormck et l. 2012; Weemstr et l. 2016). Moreover, emerging evidence lso suggests tht fine roots of the 1st 3rd order, which hve greter cpcity to tke up wter nd nutrients, cn e llocted more C, thus exhiiting longer root lifespn (Hendrick nd Pregitzer 1993; Eissenstt nd Yni 1997; Frrr nd Jones 2000). Once nutrients in the rhizosphere of fine roots re depleted, less C is llocted to roots due to reduced nutrient cquisition, leding to roots ging nd eventul deth (Hendrick nd Pregitzer 1993; Eissenstt nd Yni 1997; Frrr nd Jones 2000). The control of root lifespn y lloction of photossimiltes cn e ccounted for y the theory of cost-enefits nlysis (Eissenstt nd Yni 1997; Frnklin et l. 2012; Reich et l. 2014). However, there hve een only few studies to directly evlute how these fctors interctively ffect root lifespn t community level in generl, nd in the temperte steppe in prticulr. Under nturl conditions, different types of vegettion occur cross different edphic nd climtic conditions. Therefore, root lifespn of different plnt communities my reflect the overll iotic, edphic nd climtic chrcteristics t regionl scle. Yes such vrition cn lso occur within n ecosystem, nevertheless, few studies hve quntified root lifespn nd compred the sptil differences in lifespn mong different types of plnt communities within given ecosystem (Eissenstt nd Yni 1997; Gill nd Jckson 2000; Guo et l. 2008; Chen et l. 2013; McCormck et l. 2013). Moreover, single vlue of root lifespn for given community within n ecosystem hs often een used to predict the ecosystem C cycling y modeling, yet lifespn cn vry intr-nnully ccording to seson (López et l. 2001; Anderson et l. 2003; Bi et l. 2008). Such vrition is missing from model projections of future cron cycling, which my mke them significntly less ccurte (McCormck et l. 2013; Wrren et l. 2015). In prt models cnnot include such informtion due to lck of pproprite dt, nd quntittive determintion of lifespn of fine roots is techniclly chllenging ecuse of difficulties ssocited with direct oservtions of roots in situ under field conditions. Therefore, knowledge on how the iologicl, edphic nd climtic fctors impct root lifespn in regionl nd temporl scles is scrce in the literture. The Inner Mongoli temperte steppes, n integrl prt of Eursin steppes, re mjor temperte grsslnds in northern Chin, with totl re of 5.85 9 10 6 h. The Eursin steppes, northern Americn Pririe, southern Americn Pmps nd Africn Svnn re min grsslnds cross the gloe. Despite covering vst res on the erth, there re limited dt on root lifespn nd turnover of grsslnds so fr. Stip grndis, S. krylovii nd S. reviflor grsslnds re three mjor types of grsslnds in the Inner Mongoli steppes, ccounting for 30.1, 18.6 nd 13.3% of totl re in Inner Mongoli, respectively. The three Stip communities differ in their productivity, dominnt
Biogeochemistry (2017) 132:293 306 295 species, soil trits, nd climtic fetures such s nnul precipittion nd temperture. However, little is known whether nd how the root metrics of the three types of grsslnds differ. The development of rhizotron nd minirhizotron technologies mkes it possile to directly monitor dynmics of individul roots in situ t the ecosystem level (Tylor et l. 1990; Mjdi et l. 2005; McCormck nd Guo 2014), thus llowing us to evlute the roles of extrinsic nd intrinsic fctors in the regultion of root lifespn cross different types of grsslnds. To unrvel the overll ptterns of root longevity of Inner Mongoli steppes nd understnd the role of extrinsic nd intrinsic fctors in the control of root longevity, we compred root lifespn of the three types of grsslnds in the temperte steppes of Inner Mongoli y rhizotron. We further explored the temporl ptterns of root lifespn of the three grsslnds y mesuring lifespns of roots produced in different sesons. Given the differences in precipittion, dominnt species nd edphic trits mong the three types of grsslnds, we evluted whether root lifespns of the three grsslnds differ, nd explored the mechnisms underlying the differences. Mterils nd methods Study site Stip krylovii grsslnd is one of the representtives of the typicl sutypes of grsslnds mong different types of grsslnds in the Inner Mongoli temperte steppes. This type of grsslnd occurs in the middle nd westwrd region of Inner Mongoli Autonomous Region with totl re of 1.09 9 10 6 h, ccounting for 18.6% of the totl re of Inner Mongoli temperte steppes. Our study site ws in the Duolun Restortion Ecology Experimenttion nd Demonstrtion Sttion, Institute of Botny, the Chinese Acdemy of Sciences (DREEDS, 116817 0 E, 422 0 N; 1324 m.s.l.), which is typicl S. krylovii grsslnd (Fig. 1). The re is locted in the temperte climtic zone, nd its men nnul temperture is 2.1 C. Men nnul precipittion is 382.2 mm. Vegettion in this re is typicl steppe community nd the dominting species re perennils, including S. krylovii nd Artemisi frigid (Yng et l. 2011). Fig. 1 Loctions of the Duolun Restortion Ecology Experimenttion nd Demonstrtion Sttion, Institute of Botny, the Chinese Acdemy of Sciences (DREEDS, 116817 E, 422 N; 1324 m.s.l.), the Inner Mongoli Grsslnd Ecosystem Reserch Sttion, the Chinese Acdemy of Sciences (IMGERS, 116842 0 E, 43838 0 N; 1 m.s.l.) nd the Inner Mongoli Acdemy of Agriculture nd Animl Husndry Reserch Sttion (IMAHRS, 111853 0 E, 41847 0 N; 1450 m.s.l.) in n Inner Mongoli steppe in northern Chin. The three loctions tht the root lifespn ws determined re representtives of the three types of grsslnds within Inner Mongoli steppes Stip grndis grsslnd is zonl representtive of the typicl sutypes of grsslnds in the temperte steppes, nd locted in the middle of Inner Mongoli temperte steppes with n re of 1.76 9 10 6 h, ccounting for 30.1% of totl re of the grsslnds in Inner Mongoli. The S. grndis grsslnd used in the present study is locted in Inner Mongoli (116842 0 E, 43838 0 N; 1 m.s.l.; Fig. 1). This re is chrcterized y continentl, semi-rid climte, nd its men nnul temperture is 0.7 C. Men nnul precipittion is 335 mm. Dominnt species in the typicl steppe community re the perennil unchgrss, including S. grndis, nd Leymus chinensis (Bi et l. 2004). Stip reviflor grsslnd is one of the zonl sutypes of the desert steppe in the Inner Mongoli temperte steppes. The S. reviflor grsslnd used in the present study is locted in the westwrd region of the Inner Mongoli temperte steppes (111853 0 E, 41847 0 N; 1450 m.s.l.; Fig. 1). The men nnul temperture is 3.4 C. Men nnul precipittion is 2 mm. Dominnt species in the S. reviflor grsslnd re perennil S. reviflor nd Artemisi frigid (Li et l. 2008).
296 Biogeochemistry (2017) 132:293 306 Experimentl design In Septemer 2010, the representtive plnt communities tht hve een excluded from grzing since 2000 in these three studied sites were selected to set the experimentl plots. The plots were 6 m 9 8 m nd seprted y seprted y 2-m-wide uffer strips with six replicted plots for ech type of grsslnd. Mesurement of root dynmics On Septemer 3, 2010, one glss root-window ws instlled in ech plot. The glss window (0.4-cm thick) of 35 cm in height nd 20 cm in length ws instlled verticlly into the soil. On ech glss window, 35 9 20 cm pnel (with 5-cm distnce to the ottom nd 5 cm to the right nd left sides) ws seprted into three 10 9 10 cm squres y crving the glss. In order to minimize the impcts of light on root growth, the upper edge of the glss window ws instlled under the soil surfce nd piece of drk iron (20 cm in length, 1.5 cm in redth nd 0.5 mm in thickness) ws covered on the top of the glss. A hole ws dug in ech plot with verticl profile for the glss window to e instlled. The glss window ws put tightly to the trench nd fixed with one iron stick t ech side. After the root-windows were inserted, soil ws ckfilled s tightly s originlly. The soil ws closely ttched to the glss throughout the whole study period. A digitl cmer (Cnon G5) ws used to monitor root growth. Oservtions egn on My 10, 2011, which ws out 8 months fter the instlltion of glss windows, nd lsted until 20 Octoer 2012 with smpling intervls of 15 d. On ech smpling dte, the soil on one side of the glss window ws removed nd the screen ws clened with tissue pper. One digitl picture ws tken for ech of the three 10 cm 9 10 cm numered squres. Removed soil ws crefully ckfilled gin fter pictures were tken s descried previously (Bi et l. 2008, 2012). Removl nd ckfilling soils during picturing occurred outside of the rhizotron, thus roots inside of the rhizotron seprte y glss were not distured. Appernce nd disppernce of roots were nlyzed y the softwre of Mpinfo Professionl (5.0; Pitney Bowes Mpinfo Corportion, New York, US) s descried previously (Bi et l. 2008, 2012). For the initilly collected imges, ech root ws ssigned with n identifiction numer nd distinguished it s living or ded sed on its colors s descried in our previous pper (Bi et l. 2008). For the following imge sets, the trcings from the previous dte were compred with the new imges, thus llowing previously existing roots to e identified. Newly emerged roots were lso identified nd numered. Roots tht hd disppered t susequent imges were ssumed to e ded nd decomposed. Complete records were kept for ll roots, even for those tht were clssified s ded. In the present study, roots monitored re first nd second order roots ccording to the clssifiction y Pregitzer et l. (2002) with dimeters less thn 1 mm. Mesurements of oveground net primry productivity (ANPP) Aoveground iomss ws clipped t the ground level in rndomly selected qudrt (1 9 1 m) in ech plot in mid-august ech yer, nd oth living oveground iomss nd stnding litter of the sme species were included. Plnt smples were oven-dried t 70 C for 48 h nd then seprtely weighed for determintion of ANPP. Mesurement of soil trits In ech plot, three fresh soil smples t the soil depths of 0 30 cm were tken in the mid-august of 2012, nd inorgnic N (NH 4? -N, NO 3 - -N) concentrtions were mesured using continuous-flow ion uto-nlyzer (Sclr SAN plus segmented flow nlyzer, the Netherlnds). A portion of ech soil smple ws ground with mill nd pssed through 0.18-mm sieve for determintion of orgnic C concentrtions. The orgnic cron concentrtion ws determined using CHNOS Elementl Anlyzer (Vrio EL III). To determine the ville phosphorus contents, 0 30 cm soil ws digested with NHCO 3 nd then ws determined y molydenum-stiium colorimetry method with UV Visile spectrophotometer (UV- 2550, SHIMADZU Corportion, Chin). Potssium in soil (0 30 cm) ws extrcted y 1 M NH 4 OAc (ph 7.0) t 1:10 rtio (w/v) for 30 min. The extrcted solution ws filtered nd used to determine K? content y ICP-OES. Soil ulk density ws mesured using the soil cores (volume cm 3 ) otined from the three lyers in ech plots. Soils (0 30 cm) ner the rhizotron were smpled y soil core (3 cm dimeter) nd used for determintion of grvimetricl soil wter contents
Biogeochemistry (2017) 132:293 306 297 fter collecting root imges. A soil core (3-cm dimeter) of fresh soil from 0 to 30 cm soil lyer ws rndomly smpled in August of 2012, then irdried soil ws pssed through 2-mm sieve for determintion of soil ph. Soil ph ws determined with Russell RL060P portle ph meter (Thermo Electron Corportion 166 Cummings Center, USA), nd the wter/soil rtio ws 1:2.5. Mesurements of solule sugrs in roots In ech plot, we collected roots of plnts community within the soil lyer of 0 30 cm in Sept., 2012. Roots with dimeter \1 mm were selected nd dried for further determintion of solule sugr contents y the nthrone method with UV-VIS7500 spectrophotometer (Techcomp, Shnghi, Chin) s descried y Rnwl nd Miller (2008). Mesurements of BNPP nd root N content A root in-growth method ws used to estimte elowground net productivity (BNPP) in 0 30 cm soils y the end of growing sesons in lte August in ech yer. To determine root N content, living roots tken y soil-smpling pplince (30 cm in length, 12.5 cm in dimeter) on lte August were wshed nd dried in n oven. N contents in roots with dimeters less thn 1 mm were mesured colorimetriclly y the Kjeldhl cid-digestion method with n Alpkem utonlyser (Kjektec System 1026 distilling unit, Sweden) fter extrction with sulphuric cid. Mesurements of root trits For S. krylovii, S. grndis nd S. reviflor, t lest 20 well-developed individuls were crefully dug out with pickxe to depth of 20 cm nd root systems were immeditely put in continers filled with wter. After seprted from shoots, roots were crefully clened with fine strem of wter to remove soil, orgnic mtter prticles. The first-order roots s descried y (Pregitzer et l. 2002) were seprted from the root systems. In order to ensure sufficient quntity to weigh, the first-order roots from 3 or 4 intct root systems were mixed into one susmple. Ech susmple ws firstly scnned with Cnon desk scnner (Cnon LiDE 220, Tokyo, Jpn), nd then oven dried t 60 C for 48 h nd weighed. From the scnned imges, the verge root dimeter nd root totl length were determined using WinRHIZO softwre (Regent Instruments Inc., Queec City, QC, Cnd). The specific root length (SRL) ws clculted s the root totl length divided y its dry mss. Dt nlysis nd sttistics Root lifespn ws clculted s the dte on which roots were oserved s lck or disppered minus the dte on which roots were initilly oserved on the window. The dte of root ppernce or disppernce ws estimted s the dte midwy etween the smpling periods ecuse they might hve occurred on ny dy during the pprox. 15-d smpling intervl etween two consecutive oservtions (López et l. 2001; Anderson et l. 2003). A totl of 22949 roots were used to evlute the effects of grsslnd types on root longevity. We selected totl of 13,307 new roots germinted in spring (1 30 My), summer (1 31 July) nd utumn (20 August 10 Septemer) of 2011 nd 2012 to nlyze the effects of different sesons on their survivl rtes nd longevity of the three grsslnds. We clculted their men longevity from survivl curve using the Kpln Meier method with SPSS (12.0) softwre, nd compred the root survivl rtes y Log-rnk test. The one-wy ANOVA ws used to exmine the influences of different grsslnd types on soil prmeters, oveground net primry production, nd root solule sugr content nd root longevity. Liner regressions were used to determine the correltion etween root longevity nd oveground net primry production, root solule sugr content, soil wter content, soil orgnic mtter, soil inorgnic N content, soil ville phosphorus contents (AP), soil ville potssium contents (AK), ulk density, soil ph, root N content, root dimeter, specific root length nd BNPP/ ANPP. Stepwise multiple liner regressions were further used to identify the most importnt fctor ffecting root longevity fter the different grsslnd types. All sttisticl nlyses were conducted with SAS softwre (SAS Institute Inc., Cry, NC, USA).
298 Biogeochemistry (2017) 132:293 306 Tle 1 Chrcteristics of 0 10 cm soils in S. krylovii, S. grndis nd S. reviflor grsslnds in inner Mongoli temperte steppes Grsslnd types ph Soil orgnic mtter (g kg -1 ) NH 4?? NO 3 - content (mg kg -1 ) AP (mg kg -1 ) AK (mg kg -1 ) Bulk density (g cm -3 ) Soil wter content (%) S. krylovii 6.94 ± 0.02 43.79 ± 1.45 8.69 ± 0.38 8.11 ± 0.31 200.0 ± 12.31 1.22 ± 0.08 7.52 ± 0.37 S. grndis 7.26 ± 0.14 27.89 ± 1.75 8.99 ± 0.37 5.36 ± 0.42.9 ± 10.07 1.39 ± 0.05 12.56 ± 0.30 S.reviflor 8.06 ± 0.03 21.54 ± 0.78c 6.12 ± 0.38 7.00 ± 0.43 162.8 ± 6.86 1.38 ± 0.05 5.72 ± 0.22c Dt re mens ± stndrd errors (n = 6) Different letters ove the column men significnt difference etween different grsslnd types plots t P \ 0.05 AP Aville phosphorus content in soil, AK Aville potssium content in soil Results Soil metrics in the three grsslnds Soil ph vlues in the three types of grsslnds of S. krylovii, S. grndis nd S. reviflor grsslnds rnged from 6.9 to 8.1, with the soil ph in S. reviflor grsslnd nd S. krylovii grsslnd highest nd lowest, respectively (Tle 1). Soil orgnic mtter in S. krylovii grsslnd ws highest, wheres soil orgnic mtter in S. reviflor grsslnd ws the lowest (P \ 0.05). There ws no significnt difference in soil inorgnic nitrogen (N) contents etween the S. krylovii nd S. grndis grsslnd (P [ 0.05), while the inorgnic N content in S. reviflor grsslnd ws significntly lower thn tht in S. krylovii nd S. grndis grsslnd (P \ 0.01). The ville phosphorus (P) contents in soil of S. grndis grsslnd were significntly lower thn those of S. krylovii nd S. reviflor grsslnds (P \ 0.05). In contrst to soil P nd inorgnic N contents, there were no significnt differences in oth soil potssium contents nd soil ulk density mong the three types of grsslnds (P [ 0.05). During the experimentl period, soil wter contents in S. grndis grsslnd were significntly higher thn in S. krylovii nd S. reviflor grsslnds, nd soil wter contents in S. reviflor grsslnd were the lowest mong the three grsslnds (P \ 0.01; Tle 1). Root lifespn in the three types of grsslnds In the Inner Mongoli temperte steppes, survivl rtes nd root lifespn of plnt communities in the three types of grsslnds differed significntly (Fig. 2). Among the three types of grsslnds studied, the survivl rtes nd root lifespn of S. krylovii grsslnd were the lowest (P = 0.021), followed y those of S. grndis grsslnd, while the survivl rtes nd root lifespn of S. reviflor grsslnd were the highest (P = 0.0001). Accordingly, the men root lifespn in the S. krylovii, S. grndis nd S. reviflor grsslnds ws estimted to e 98, 125 nd 146 dys, respectively (Fig. 2; Tle 2). Lifespn of roots produced in different sesons Roots produced in the three growing sesons in the three types of grsslnds exhiited similr survivl rtes nd lifespn (Fig. 3). Log-rnk test reveled tht survivl rtes nd root lifespn for the roots produced in spring were significntly shorter thn those produced in summer nd utumn (P \ 0.01; Fig. 3). In contrst, survivl rtes nd lifespn for those roots produced in utumn were significntly longer (P \ 0.01; Fig. 3c) thn those orn in summer nd spring. The survivl rtes nd lifespn of those roots orn in summer were etween those of roots orn in spring nd in utumn (Fig. 3). For those roots produced in spring, no significnt difference in root lifespn etween S. krylovii nd S. grndis grsslnd ws detected, while root lifespn in the two types of grsslnds were oth significntly shorter thn tht of S. reviflor grsslnd (P \ 0.01). For those roots produced in summer nd utumn, root lifespn in S. krylovii grsslnd ws the shortest (P \ 0.05; Fig. 3c), followed y those roots in S. grndis grsslnd (P \ 0.01). The root lifespn in S. reviflor grsslnd ws the longest (P \ 0.001).
Biogeochemistry (2017) 132:293 306 299 Fig. 2 Survivl curves nd longevity of root cross different grsslnd types t soil depths of 0 30 cm during the growing sesons of 2011 nd 2012. Dt shown in the figure were sed on totl numer of 10690 nd 8364 nd 3895 individul roots for S. krylovii, S. grndis nd S. reviflor grsslnd, respectively. Men lifespn ws generted using the Kpln Meier method. Survivl differences of different grsslnd types were compred using Logrnk test. Dt re men ± SE of six plots. The missing smpling dtes were winter months Proportion of roots surviving 1.2 1.0 0.8 0.6 0.4 0.2 0.0 S. krylovii S. grndis S. reviflor Root longevity (d) c 60 0 SK SG SB Grsslnd types 0 60 240 300 360 420 Time fter first ppernce (d) Tle 2 Root longevity nd 95% confidence intervl (CI) cross different grsslnd types nd different sesons t soil depths of 0 30 cm during the growing sesons of 2011 nd 2012 Root longevity (d) Stip krylovii Stip grndis Stip reviflor Men 95% CI Men 95% CI Men 95% CI Lower Upper Lower Upper Lower Upper Overll 98 95 102 125 131 146 139 153 Spring-orn 59 58 61 54 48 60 78 70 86 Summer-orn 98 88 108 115 108 156 144 169 Autumn-orn 139 131 148 166 157 176 195 186 204 ANPP, contents of solule sugrs, root dimeter nd specific root length in the three types of grsslnds ANPP in the S. krylovii grsslnd ws the highest, wheres ANPP in the S. reviflor grsslnd ws the lowest (P \ 0.001). But there ws no significnt difference in ANPP etween the S. reviflor nd S. grndis grsslnds (Fig. 4). Among the three types of grsslnds, solule sugr contents in roots of S. krylovii grsslnd were the lowest, followed y those of S. grndis grsslnd, while the root solule sugr contents of S. reviflor grsslnd were the highest (P \ 0.001) (Fig. 4). Among the three types of grsslnds, root dimeters of S. krylovii were the lowest, followed y those of S. grndis, while the root dimeters of S. reviflor were the highest (P \ 0.01) (Fig. 5). In contrst, specific root length of S. krylovii ws the highest, followed y tht of S. grndis, while the specific root length of S. reviflor ws the lowest (P \ 0.01) (Fig. 5). Correltions etween root lifespn nd iotic nd iotic fctors To elucidte the mechnisms y which the root lifespn differed mong the three types of grsslnds, we evluted the roles of soil orgnic mtter, soil inorgnic N contents, ove ground net primry production (ANPP), nd root solule sugr contents of dominnt species in the three types of grsslnds.
300 Biogeochemistry (2017) 132:293 306 Fig. 3 Survivl curves nd longevity of roots orn in spring (), summer () nd utumn (c) in S. krylovii, S. grndis nd S. reviflor grsslnds t soil depth of 0 30 cm. Dt were otined from totl numer of 4983, 3939 nd 4385 individul roots orn in spring, summer nd utumn in 2011 nd 2012. Survivl differences in different grsslnd types were compred using Log-rnk test. Dt re men ± SE (n = 6) Proportion of roots surviving 1.2 () Spring-orn roots S. krylovii S. grndis 1.0 0.8 0.6 0.4 0.2 S. reviflor 200 150 50 0 SK SG SB Grsslnd types 0.0 0 40 Root longevity (d) 1.2 () Summer-orn roots Proportion of roots surviving 1.0 0.8 0.6 0.4 0.2 Root longevity (d) 200 150 50 0 c SK SG SB Grsslnd types 0.0 0 90 270 360 450 Proportion of roots surviving 1.2 (c) Autumn-orn roots 1.0 0.8 0.6 0.4 0.2 Root longevity (d) 200 150 50 0 c SK SG SB Grsslnd types 0.0 0 90 270 360 450 Time fter first ppernce (d)
Biogeochemistry (2017) 132:293 306 301 450 () 0.3 () ANPP (g m -2 ) 360 270 90 Root dimeter (mm) 0.2 0.1 c 0 0.0 Root solule sugr content (%) 10 8 6 4 2 0 () c As shown in Fig. 6, cross different grsslnds in the two growing sesons, root lifespn ws negtively correlted with soil orgnic mtter (r 2 = 0.55), soil inorgnic N concentrtion (r 2 = 0.44), ANPP (r 2 = 0.55) nd root N contents (r 2 = 0.44). In contrst, root lifespn ws positively correlted with solule sugr contents in roots (r 2 = 0.57) nd BNPP/ ANPP rtio (r 2 = 0.68). No significnt correltions etween root longevity nd soil ph, ville P nd K contents in soils, soil ulk density nd soil wter content were detected (P [ 0.05). Root lifespn ws positively correlted with root dimeter (r 2 = 0.37) nd negtively correlted with specific root length (r 2 = 0.44). Stepwise multiple regression nlyses reveled tht 66.13% of the sptil vrition in root longevity cn e explined y root nonstructurl crohydrte content (prtil r 2 = 0.3166, P = 0.0003) nd rtio of BNPP to ANPP (BNPP/ ANPP) (prtil r 2 = 0.3447, P = 0.0001). S. krylovii S.grndis S. reviflor Grsslnd types Fig. 4 Aoveground net primry production (ANPP, ) nd root solule sugr content () in S. krylovii, S. grndis nd S. reviflor grsslnd. Dt re men ± SE (n = 6) Specific root length (m g -1 ) 50 40 30 20 10 0 () Discussion To the est of our knowledge, there hve een no previous studies tht directly mesure nd compre root lifespn cross different grsslnd types. Our results show tht root lifespn in the three types of grsslnds in the Inner Mongoli steppes differed significntly (Fig. 2), suggesting tht within the temperte grsslnd ecosystem in Inner Mongoli, root turnover in different types of grsslnds cn differ sustntilly. These results highlight tht the prediction of C cycling y modeling under the widely used ssumption of uniform root turnover mong the different types of grsslnds should e treted with cution. For exmple, we estimted root turnover rte of 2.17, 1.89 nd 1.36 yr -1 for S. krylovii, S. grndis nd S. reviflor grsslnds, respectively using root length production nd root length stnding crop. The estimted root turnover rte for the Inner Mongoli S. krylovii S. grndis S. reviflor Grsslnd types Fig. 5 Comprison of root dimeter () nd specific root length (SRL, ) of first order roots mong S. krylovii, S. grndis nd S. reviflor. Dt re men ± SE (n = 6) c
302 Biogeochemistry (2017) 132:293 306 () S. krylovii S. grndis S.reviflor () r 2 = 0.5479 P = 0.0004 r 2 = 0.4442 P = 0.0025 15 20 25 30 35 40 45 50 (c) Soil orgnic mtter (g kg -1 ) 4.0 5.5 7.0 8.5 10.0 11.5 Inorgnic N content (mg kg -1 ) (d) Root longevity (d) r 2 = 0.5507 P = 0.0004 r 2 = 0.571 P = 0.0003 200 240 2 320 360 400 440 ANPP (g m -2 ) (e) 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 Root soule sugr content (%) (f) r 2 =0.4368 P = 0.0028 r 2 = 0.6843 P = 0.0001 60 (g) 0.9 1.2 1.5 1.8 2.1 Root N content (%) 1 2 3 4 5 6 BNPP/ANPP (h) r 2 = 0.374 P = 0.007 0.15 0.20 0.25 0.30 Root dimeter (mm) r 2 = 0.4443 P = 0.0025 30 33 36 39 42 45 SRL (m g -1 )
Biogeochemistry (2017) 132:293 306 303 Fig. 6 The correltions etween root longevity nd soil orgnic mtter (SOM, ), soil inorgnic N content (), oveground net primry production (ANPP, c) nd contents of root solule sugrs (d), root N content (e), BNPP/ANPP (f), root dimeter (g) nd SRL (h) ins. krylovii, S. grndis nd S. reviflor grsslnds with dt otined from 18 plots steppes cn differ y up to 0.81 yr -1. In contrst to our results, Chen et l. (2000) reported root turnover rte of 0.49 yr -1 of similr grsslnds in Inner Mongoli steppes y the trditionl soil core method. A few studies hve evluted root lifespn nd turnover rtes of grsslnd species y the minirhizotron method so fr. A men root turnover rte of 0.86 yr -1 of shortgrss steppe of estern Colordo ws otined using minirhizotron y Gill et l. (2002). However, men root turnover rte of 0.236 yr -1 ws otined y the trditionl soil core techniques for the identicl shortgrss steppe (Gill et l. 2002). These results indicte tht root turnover rtes otined y different methods differ sustntilly. Such vrition in root turnover rtes mke it difficult to ccurtely predict C cycling of grsslnd ecosystems t the glol scle nd point to need to resolve methodology for estimting root turnover. Root lifespn, root turnover rte nd C lloction re direct prmeters in mny models. However, few studies hve mesured nd compred root lifespn cross different grsslnds nd/or methodologies t regionl scle. In forest ecosystem, McCormck et l. (2013) reported the root lifespn of temperte tree species cross the estern prt of the United Sttes t the regionl scle. They found tht root lifespn in the North Centrl sttes is shorter, wheres it is longer in the southestern sttes. They lso found distinct ptterns for the root lifespn cross the studied res, nd ttriuted their oservtions to reltively high undnce of certin tree species (McCormck et l. 2013). One importnt finding in the present study is tht root lifespn cross different types of grsslnds t the regionl scle differed sustntilly. A numer of extrinsic nd intrinsic fctors my e involved in the control of the root lifespn. These include the climtic fctors, temperture nd rinfll, edphic fctors, for exmple, soil inorgnic N, nd intrinsic fctors such s plnt species composition, solule sugrs concentrtions nd N contents of plnt roots. These fctors cn directly nd/or indirectly determine root lifespn y ffecting growth nd deth of plnt roots (Peek 2007; Chen nd Brssrd 2013; McCormck nd Guo 2014). Thus, the fctors involved in the control of root lifespn for different plnt communities cross regionl scle within the grsslnd ecosystem re more complex thn those for single plnt community in n individul site due to interweving effects of these fctors on root metrics. Our multivrite nlysis reveled tht the effects of mny environmentl cues on root lifespn, including root solule sugrs contents nd BNPP/ANPP. In the Inner Mongoli temperte steppes, we found tht mong three different types of grsslnds, root lifespn in the three grsslnds followed the orders of S. reviflor [ S. grndis [ S. krylovii grsslnds (Fig. 2). The differences in vegettion chrcteristics, climtic nd edphic fctors mong the three grsslnds my underpin the oserved differences in their root longevity. We further explored the mechnisms underlying the differences. Our results showed tht the overll root lifespn in Inner Mongoli temperte steppes ws negtively correlted with soil orgnic mtter concentrtion, inorgnic N, nd ANPP of grsslnds, while root lifespn exhiited significntly positive correltion with contents of solule sugrs in roots of the dominnt species nd BNPP/ ANPP (cf. Figure 6). It hs een demonstrted tht the vilility of soil N is n importnt fctor controlling root growth nd lifespn (Burton et l. 2000), such tht higher soil N vilility leds to shorter root lifespn due to enhnced root respirtion resulting from greter ccumultion of N in roots. For exmple, root lifespn cn e longer when grown in infertile soils compred to tht grown in nutrient-rich soils, llowing roots to conserve more nutrients (Eissenstt nd Yni 1997; Vn der Krift nd Berendse 2002; West et l. 2003; McCormck nd Guo 2014). Our results re consistent with this frmework. For exmple, we found tht root lifespn in the S. reviflor grsslnd ws the longest mong the three grsslnds, while soil orgnic mtter nd soil inorgnic N contents in the S. reviflor grsslnd were the lowest (Tle 1; Fig. 2). In ddition, the differences in root lifespn mong the three types of grsslnds my lso e explined y their different ptterns of C lloction. It hs een well documented tht root lifespn is closely relted to C supply from shoot to root (Eissenstt et l. 2000; Anderson et l. 2003). There re reports showing tht decrese in C lloction to root shortens the root lifespn (Eissenstt nd Yni 1997; Bi et l.
304 Biogeochemistry (2017) 132:293 306 2008, 2010). Reich et l. (2014) reported tht plnts cn llocte more iomss into roots under cold nd drought conditions. Our results lso showed tht the contents of solule sugrs in roots of the dominnt species, nd the rtio of BNPP/ANPP in S. reviflor grsslnd with less fertile soils in the driest climte were higher compred to the other two types of grsslnds (Fig. 4). This my ccount for their longer lifespn. The longer root lifespn of S. reviflor grsslnd my lso help these plnts conserve more nutrients. In contrst, S. krylovii nd S. grndis grsslnds often occur in reltively wetter climtes with more fertile soils thn S. reviflor grsslnds. Therefore, less C would e llocted to the elowground in S. reviflor grsslnd, leding to shorter root lifespn thn tht of S. reviflor grsslnd. The shorter root lifespn in the S. krylovii nd S. grndis grsslnds could mximize the enefits for cquisition of resources with less C investment (Eissenstt nd Yni 1997; Frnklin et l. 2012). Our results re in line with this hypothesis (Fig. 6c, d). Moreover, root trits such s dimeter nd root specific length re closely relted to root lifespn. For exmple, it hs een widely reported tht root lifespn is prolonged with incresing root dimeter, while high specific root length often concurs with short root lifespn (Weemstr et l. 2016). In the present study, we oserved tht root lifespn cross the three grsslnds ws positively correlted with root dimeters, while it ws negtively correlted with specific root length (Fig. 6g, h), highlighting the conservtive nture of root longevity cross different ecosystems. These fetures re lso in greement with the root economic spectrum (Mommer nd Weemstr 2012; Weemstr et l. 2016). In ddition to the intrinsic fctors, the environmentl fctors cn lso hve importnt impcts on root lifespn (Chen nd Brssrd 2013; McCormck nd Guo 2014). In our studies, the men nnul precipittion tken over multiple yers in the three types of grsslnds differed with the order of S. krylovii (382 mm) [ S. grndis (335 mm) [ S. reviflor grsslnds (2 mm). However, cross the experimentl period, the men soil wter contents in the S. grndis grsslnd were ctully the highest. This my result from unevenness in precipittion inter- nnully, nd my ccount for the oservtion of no significnt correltion etween root lifespn nd soil wter contents. The men nnul temperture in S. reviflor grsslnd ws the highest mong the three grsslnds. Given the negtive correltion etween root lifespn nd temperture (Mjdi nd Ohrvik 2004; Lepplmmi-kujnsuu et l. 2014), shorter root lifespn in S. reviflor grsslnd is expected. However, we found tht root lifespn in the S. reviflor grsslnd ws longest mong the three types of grsslnds. These results suggest tht the nnul tmosphere temperture my not e key determinnt for root lifespn in the Inner Mongoli steppe. We further conducted stepwise multiple regression nlyses, nd found tht the chnges in root solule sugrs contents nd BNPP/ANPP rtio ccounted for 66% of vrition in the root lifespn mong different types of grsslnd, suggesting tht the intrinsic fctors re of importnce for controlling root lifespn cross the three types of grsslnds in northern Chin. It hs een reported tht root lifespn is closely relted to the sesons when roots were initilly produced (López et l. 2001; Anderson et l. 2003; Guo et l. 2008). We thus compred lifespn for roots orn in different sesons (spring, summer nd utumn) mong the three types of grsslnds. We found consistent pttern for lifespn of roots orn in different sesons mong the three types of grsslnds: lifespn of roots orn in spring ws the shortest, while lifespn of roots orn in utumn ws the longest (Fig. 3). The differences in lifespn of roots produced in different sesons cn e relted to their differences in functions. For instnce, those roots produced in spring my minly involve uptke of wter nd nutrients, while the roots orn in utumn my e minly used to store nutrients nd produce new lterl roots (López et l. 2001). We further oserved tht lifespn for the roots orn in the sme seson exhiited significnt difference mong the three types of grsslnds. For exmple, lifespn for those roots produced in spring of S. krylovii nd S. grndis grsslnds ws comprle, ut it ws significntly shorter thn lifespn of roots in S. reviflor grsslnd initited in spring (cf. Figure 3). Similr to the roots orn in spring, lifespns of roots produced in summer nd utumn cross the three grsslnds exhiited the sme ptterns: root lifespn in S. reviflor nd S. krylovii grsslnds ws longest nd shortest, respectively (cf. Figure 3, c). These results my suggest tht root lifespn in different types of grsslnds is minly determined y lifespn of roots orn in summer nd
Biogeochemistry (2017) 132:293 306 305 utumn. These findings lso indicte tht lifespn for roots orn in the different sesons hs different sensitivity to environmentl fctors. Our previous study found tht N ddition hs different effects on lifespn of roots orn in spring, summer nd utumn, nd tht cn e explined y C lloction nd chnges in temperture (Bi et l. 2008, 2010, 2012). These oservtions indicte tht the impcts of the different intrinsic nd extrinsic fctors on the root lifespn orn in different sesons re complex, ut the mechnisms underlying the differences in root lifespn cross the different sesons remin unknown. Severl fctors such s soil wter contents, temperture, nd sttus of crohydrtes nd nutrients in roots my e expected to e involved in determintion of the differences in root lifespn (Bddeley nd Wtson 2005). Given tht the root lifespns in the three types of grsslnds were relted to the contents of solule sugrs in roots, it is conceivle tht crohydrte contents in roots orn in different sesons my differ, leding to the different responses of their lifespn in different types of grsslnds. In ddition, phenologicl mechnisms of plnt species my lso ffect the production nd deth rtes of roots during different sesons. Thus, future studies focusing on how oth the intrinsic nd extrinsic fctors drive the vriility in root lifespn cross lrge rnge of plnt species re wrrnted. In summry, we report, for the first time, tht root lifespn differs sustntilly mong the three mjor types of grsslnds within the temperte steppes of Inner Mongoli, such tht root lifespn of S. reviflor [ S. grndis [ S. krylovii grsslnds. Root lifespn cross the three types of grsslnds in the Inner Mongolin temperte steppes displys similr temporl pttern, such tht lifespn of the roots orn in spring nd utumn is shortest nd longest, respectively, wheres lifespn of summer-orn roots is etween tht of roots produced in spring nd utumn. The sptil nd temporl differences in root lifespn cross the three types of grsslnds re minly relted to the content of solule sugrs in roots of the dominnt species. The differences in root lifespn cross the three Stip grsslnds nd different sesons indicte the potentil importnce of including these differences in the considertion nd prediction of C cycling within these grsslnd ecosystems. Acknowledgements We would like to thnk G Wng nd Y Dong for their help in field nd lortory work. This reserch ws supported y the Ntionl Nturl Science Foundtion of Chin (31370468, 31670481) nd the Stte Key Bsic Reserch Development Progrm of Chin (2013CB956304). We thnk the editor nd nonymous reviewers for their constrictive suggestions tht gretly contriuted to improving the mnuscript. References Aerts R, Bkker C, de Cluwe H (1992) Root turnover s determinnt of the cycling C, N, nd P in dry hethlnd ecosystem. Biogeochemistry 15:175 190 Anderson LJ, Coms LH, Lkso AN, Eissenstt DM (2003) 0 Multiple risk fctors in root survivorship: 4-yer study in Concord grpe. New Phytol 158:489 501 Bddeley JA, Wtson CA (2005) Influences of root dimeter, tree ge, soil depth nd seson on fine root survivorship in Prunus vium. Plnt Soil 276:15 22 Bi YF, Hn XG, Wu JG, Chen ZZ, Li LH (2004) Ecosystem stility nd compenstory effects in the Inner Mongoli grsslnd. Nture 431:181 184 Bi WM, Wng ZW, Chen QS, Zhng WH, Li LH (2008) Sptil nd temporl effects of nitrogen ddition on root life spn of Leymus chinensis in typicl steppe of Inner Mongoli. Funct Ecol 22:583 591 Bi WM, Xun F, Li Y, Zhng WH, Li LH (2010) Rhizome severing increses root lifespn of Leymus chinensis in typicl steppe of Inner Mongoli. PLoS ONE 5:e12125 Bi WM, Xi JY, Wn SQ, Zhng WH, Li LH (2012) Dy nd night wrming hve different effect on root lifespn. Biogeosciences 9:375 384 Brdgett RD, Mommer L, De Vries FT (2014) Going underground: root trits s drivers of ecosystem processes. Trend Ecol Evol 29:692 699 Burton AJ, Pregitzer KS, Hendrick RL (2000) Reltionships etween fine root dynmics nd nitrogen vilility in Michign northern hrdwood forests. Oecologi 125:389 399 Chen HYH, Brssrd BW (2013) Intrinsic nd extrinsic controls of fine root life spn. Crit Rev Plnt Sci 32:151 161 Chen ZZ, Wng SP (2000) Typicl steppe ecosystems of Chin. Science Press, Beijing Chen L, Zeng H, Eissenstt DM, Guo DL (2013) Vrition of first-order root trits cross climtic grdients nd evolutionry trends in geologicl time. Glo Ecol Biogeogr 22:846 856 Eissenstt DM, Yni RD (1997) The ecology of root life spn. Adv Ecol Res 27:1 60 Eissenstt DM, Wells CE, Yni RD, Whiteck JL (2000) Building roots in chnging environment: implictions for root longevity. New Phytol 147:33 42 Frrr JF, Jones DL (2000) The control of cron cquisition y roots. New Phytol 147:43 53 Frnklin O, Johnsson J, Dewr R, DieckmnnU McMurtrie RE, Brnnstrom A, Dyzinski R (2012) Modeling cron lloction in trees: serch for principles. Tree Physiol 32:648 666 Gill RA, Jckson RB (2000) Glol ptterns of root turnover for terrestril ecosystems. New Phytol 147:13 31
306 Biogeochemistry (2017) 132:293 306 Gill RA, Burke IC, Luenroth WK, Milchuns DG (2002) Longevity nd turnover of roots in the shortgrss steppe: influence of dimeter nd depth. Plnt Ecol 159:241 251 Guo D, Mitchell RJ, Withington JM, Fn PP, Hendricks JJ (2008) Endogenous nd exogenous controls of root life spn, mortlity nd nitrogen flux in longlef pine forest: root rnch order predomintes. J Ecol 96:737 745 Hehdrick RL, Pregitzer KS (1993) Ptterns of fine root mortlity in two sugr mple forest. Nture 361:59 61 Iversen CM, O Brien SL (2010) Reports from the 2009 ESA nnul meeting, missing links in the root soil orgnic mtter continuum. Ecology 91:54 64 Joslin JD, Wolfe MH, Hnson PJ (2000) Effects of ltered wter regimes on forest root systems. New Phytol 147:117 129 Leppälmmi-Kujnsuu J, Slem M, Klej DB, Linder S, Helmisri HS (2014) Fine root turnover nd litter production of Norwy spruce in long-term temperture nd nutrient mnipultion experiment. Plnt Soil 374:73 88 Li CL, Ho XY, Zho ML, Hn GD, Willms WD (2008) Influence of historic sheep grzing on vegettion nd soil properties of Desert Steppe in Inner Mongoli. Agric Ecosyst Environ 128:109 116 López B, Ste S, Grci CA (2001) Fine-root longevity of Quercus ilex. New Phytol 151:437 441 Loy WM, Johnson LC, Ndelhoffer KJ (2004) Sesonl dynmics of lef- nd root-derived C in rctic tundr mesocosms. Soil Biol Biochem 36:655 666 Mjdi H, Ohrvik J (2004) Interctive effects of soil wrming nd fertiliztion on root production, mortlity, nd longevity in Norwy spruce stnd in northern Sweden. Glo Chnge Biol 10:182 188 Mjdi H, Pregitzer K, Moren AS, Nylund JE, Agren GI (2005) Mesuring fine root turnover in forest ecosystems. Plnt Soil 276:1 8 McCormck ML, Guo DL (2014) Impcts of environmentl fctors on fine root lifespn. Front Plnt Sci 5:1 11 McCormck ML, Adms TS, Smithwick EAH, Eissenst DM (2012) Predicting fine root lifespn from plnt functionl trits in temperte trees. New Phytol 195:823 8312 McCormck ML, Eissenst DM, Prsd AM, Smithwick EAH (2013) Regionl scle ptterns of fine root lifespn nd turnover under current nd future climte. Glo Chnge Biol 19:1697 1708 Mommer L, Weemstr M (2012) The role of roots in the resource economics spectrum. New Phytol 195:725 727 Nory RJ, Ledford J, Reilly CD, Miller NE, O Neill EG (2004) Fine-root production domintes response of deciduous forest to tmospheric CO 2 enrichment. Proc Ntl Acd Sci USA 101:9689 9693 Peek MS (2007) Explining vrition in fine root life spn. Prog Bot 68:382 398 Pregitzer KS, DeForest JA, Burton AJ, Allen MF, Ruess RW, Hendrick RL (2002) Fine root rchitecture of nine North Americn trees. Ecol Monogr 72:293 309 Rnwl AP, Miller WB (2008) Blckwell Pulishing Ltd Anlysis of nonstructurl crohydrtes in storge orgns of 30 ornmentl geophytes y high-performnce nionexchnge chromtogrphy with pulsed mperometric detection. New Phytol :421 433 Reich PB, Luo YJ, Brdford JB, Poorter H, Perry CH, Oleksyn J (2014) Temperte drives glol ptterns in forest iomss distriution in leves, stems, nd root. Pro Ntl Acd Sci USA 111:13721 13726 Strnd AE, Pritchrd SG, McCormck ML, Dvis MA, Oren R (2008) Irreconcille differences: fine-root life spns nd soil cron persistence. Science 310:456 458 Tylor HM, Upchurch DR, McMichel BL (1990) Applictions nd limittions of rhizotrons nd minirhizotrons for root studies. Plnt Soil 129:29 35 Vn der Krift TAJ, Berendse F (2002) Root life spns of four grss species from hitts differing in nutrient vilility. Funct Ecol 16:198 203 Vogt KA, Grier CC, Vogt DJ (1986) Production, turnover, nd nutrient dynmics of ove- nd elowground detritus of world forests. Adv Ecol Res 15:303 377 Wrren JM, Hnson PJ, Iversen CM, Kumr J, Wlker AP, Wullschleger SD (2015) Root structurl nd functionl dynmics in terrestril iosphere models evlution nd recommendtions. New Phytol 205:59 78 Weemstr M, Mommer L, Visser EJW, vn Ruijven J, Kuyper TW, Mohren GMJ, Sterck FJ (2016) Towrds multidimensionl root trit frmework: tree root review. New Phytol. doi:10.1111/nph.03 West JB, Espelet JF, Donovn LA (2003) Root longevity nd phenology differences etween two co-occurring svnn unchgrsses with different lef hits. Funct Ecol 17:20 28 Withington JM, Reich PB, Oleksyn J, Eissenstt DM (2006) Comprisons of structure nd life spn in roots nd leves mong temperte trees. Ecol Monogr 76:381 397 Yng HJ, Wu MY, Liu WX, Zhng Z, Zhng NL, Wn SQ (2011) Community structure nd composition in response to climte chnge in temperte steppe. Glo Chnge Biol 17:452 465