long-term natural succession improves nitrogen storage capacity of soil on the loess plateau, china

Long-term natural succession improves nitrogen storagecapacity of soil on the Loess Plateau China

Lei DengA Kai-bo WangB and Zhou-ping ShangguanAC

AState Key Laboratory of Soil Erosion and Dryland Farming on the Loess PlateauNorthwest AampF University Yangling Shaanxi 712100 PR China

BState Key Laboratory of Loess and Quaternary Geology Institute of Earth EnvironmentChinese Academy of Sciences Xirsquoan 710075 PR China

CCorresponding author Email shangguanmsiswcaccn

Abstract Land-use change resulting from natural succession enhances the nitrogen (N) accumulation capacity ofterrestrial ecosystems To explore those factors that foster changes in soil N storage under evolving conditions of vegetationsuccession a study on N storage at differing stages along a 150-year chronosequence was conducted in the Ziwuling ForestRegion in the central part of the Loess Plateau China A principal finding was the rapid increase in N storage in the 0ndash60 cmsoil layer which achieves a stable value after the shrub community stage (~50ndash60 years) leading to the overall long-term(~150 years) accumulation of soil stored N in the post-abandonment secondary forest Soil N accumulated mainly in thepioneer stage and showed a significant increase before the shrub community stage (P lt 005) The N storage in the 0ndash60 cmsoil layer changed from 58 to 84Mg handash1 during the transition from abandoned farmland (~3ndash5 years) to climaxcommunity (Quercus liaotungensis Koidz forest) (~150 years) The N storage values were higher in the upper (lt20 cm)than the deeper soil layers (gt20 cm) In the topsoil (0ndash20 cm) N storage values showed a markedly positive correlation withsoil organic carbon (SOC) total soil N and fine roots In the deeper soil layers (20ndash40 and 40ndash60 cm) there was a correlationonly with TN Soil bulk density soil water content and soil pH were not the determining factors behind N storage values inthe topsoil (0ndash20 cm) although they did show negative positive and negative correlations respectively In addition theyshowed no consistent correlations in the lower soil layer (lt20 cm) The results suggest that changes to N storage valueswere the result of the accumulation of SOC total N and primary productivity during the process of forest succession andthis capacity is positively related to post-abandonment forest succession on the Loess Plateau China

Additional keywords China natural succession nitrogen storage Ziwuling Forest Region

Received 23 December 2012 accepted 11 December 2013 published online 31 March 2014

Introduction

Nitrogen (N) is a key nutrient for all living organisms and is themost common limiting element for plant production in theterrestrial biosphere (Galloway et al 2004) Therefore itplays an important role in regulating both the structure(Gundale et al 2006) and function (Cleveland et al 2006) ofterrestrial ecosystems Identifying a terrestrial N sink where theelement accumulates and is held for a long period has long beenone of the most important issues in the field of biogeochemistryStudies have shown that soil has the potential to function assuch a long-term N sink (Burke et al 2002 Chapuis-Lardy et al2007)

Soil is the largest N reservoir in the terrestrial biospherecontaining more than vegetation (Yang et al 2007)Consequently the dynamics of soil N storage coulddetermine whether terrestrial ecosystems function as N

sources or sinks (Tian et al 2006) The role N plays in thebiogeochemical cycle of carbon (C) is crucial because CndashNinteractions are essential in determining whether terrestrial Csinks can be sustained over the long term (Luo et al 20042006) Carbon dioxide enrichment experiments indicate thatN limitation directly influences terrestrial C sequestration(Hu et al 2001 Schlesinger and Lichter 2001) a processfundamental to either mitigating or deferring global warmingThus there is a need to identify and describe biogeochemicalcycles not only globally but especially those processes andsystem variables relevant at the regional spatio-temporal scalebecause it is at the regional scale that landscape policies areimplemented

Luo et al (2004) reported that N dynamics is a key parameterin the regulation of long-term terrestrial C sequestration Forexample increased N deposition could have the effect of

Abbreviations Afs Abandoned farmland stage Hs herbaceous community stage Ss shrub community stage Pas pioneeringarbour community stage Mfs mixed forest community stage Cs climax community stage

Journal compilation CSIRO 2014 wwwpublishcsiroaujournalssr

CSIRO PUBLISHING

Soil Research 2014 52 262ndash270httpdxdoiorg101071SR12377

attenuating atmospheric CO2 by stimulating the accumulation offorest biomass (Giardina et al 2003) Therefore elucidating theN dynamics in soils has important implications both forthe sustainable management of land resources regionally andpredictions of future C and N cycling globally

The Loess Plateau China is an area suffering severe soiland water losses (Liu et al 2007) Many studies report thatsecondary succession can recover the properties of degradedsoil and maintain soil fertility (Feldpausch et al 2004 Jia et al2005 Wang et al 2011a Deng et al 2013a) For this reasonunderstanding secondary forest succession processes in thecentral region of the Loess Plateau is becoming increasinglyimportant (Jia et al 2005 Deng et al 2013a) To improve thefrail natural ecosystems of the Loess Plateau and alleviatedegradation the Chinese government launched a series ofconservation programs aimed at restoring agricultural land toforest and grassland this is the case for the study area wherefarmland had already been abandoned and the process ofnatural secondary succession was under way With regard tothese ecosystem functions the retention of C and N in soil iscrucial (Prietzel and Bachmann 2012) Consequentlysuccession in the Ziwuling secondary forest has receivedsignificant attention from Chinese scientists (Zou et al 2002Jia et al 2005 Wang et al 2010 Deng et al 2013a) Severalstudies have focused on changes to aboveground vegetationduring secondary forest succession in the centre of the LoessPlateau (Zou et al 2002 Wang et al 2010) Deng et al (2013a)studied soil C-storage responses to forest succession in theZiwuling Forest Region of the Loess Plateau However thereis little available information on soil N storage dynamics atdifferent stages of natural regeneration in the area

We hypothesised that natural regeneration significantlyaffects soil N conditions in the central region of the LoessPlateau To test our hypothesis and evaluate N dynamicsduring succession from managed to natural communities insemi-arid regions the objectives of this study were (i) toestimate soil N storage along a long-term natural regenerationgradient and (ii) to discuss the contributing factors andmechanisms controlling soil N storage

Materials and methods

The Ziwuling Forest Region (338500ndash368500N 1078300ndash1098400E 1100ndash1756m asl) is in the hinterland of theLoess Plateau The study area has landforms typical of hilly-gully loess topography and a mid-temperate continentalmonsoon climate The study was conducted on theLianjiabian Forest Farm of the Heshui General Forest Farmof Gansu (358030ndash368370N 1088100ndash1098180E 1211ndash1453masl) covering a total area of 23 000 km2 The arearsquos meanannual temperature is 108C (1960ndash2010) and the mean annualrainfall 587mm (1960ndash2010) (Deng et al 2013a) The regionrsquossoils are largely loessial having developed from primitive orsecondary loess parent materials (Cheng et al 2012) which areevenly distributed 50ndash130m above red earth consisting ofcalcareous cinnamon soil (Jia et al 2005) Soil pH rangesfrom 792 to 831 (Deng et al 2013a) The area is covered inspecies-rich uniform forests with a forest canopy density of80ndash95 (Deng et al 2013a) and the distribution of ground

litter was uniform The regionrsquos natural biomes are deciduousbroadleaf forests of which the climax vegetation is the Quercusliaotungensis Koidz forest Throughout the region Populusdavidiana Dode and Betula platyphylla Suk communitiesdominate the pioneer forests Sophora davidii (Franch)Skeels Hippophae rhamnoides (Linn) Rosa xanthina Lindland Spiraea pubescens Turcz are the main shrub species andBothriochloa ischaemun (Linn) Keng Carex lanceolata BoottPotentilla chinensis (Ser) Lespedeza dahurica (Laxm) Schindland Stipa bungeana Trin are the main herbaceous species

The post-abandonment secondary forests regeneratednaturally on abandoned farmland after many local inhabitantsemigrated from the Ziwuling Forest Region during a nationalconflict in 1842ndash1866 Chen (1958) who investigated thevegetation recovery of the Ziwuling Forest Region inthe 1950s determined that P davidiana made up 70 of thevegetation cover after about a century Zou et al (2002)investigated the vegetation succession three times (19621982 and 2000) They found that Q liaotungensis forestshad replaced P davidiana forests within 50 years Thusthe recovery period for Q liaotungensis forests was~150 years (Wang et al 2010) In the 20 or more yearsranging from 1940 to the 1960s famine war and disastersled some people to immigrate and reclaim land in the regionThe process of human emigration and immigration in whicharable land has been abandoned more than once in differentareas has yielded different successional stages on the degradedsloping land Based on previous research we chose sixcommunities that were at different stages of succession in theZiwuling Forest Region Two methods were used to determinethe ages of the communities For the shrub and herbaceouscommunities with lt60 years of succession we verified thelength of time both by visiting local elders and by referringto land contracts between farmers and the government Forthe forest community which had gt60 years of recovery wedetermined the length of time by boring tree rings and checkingrelated written sources (Wang et al 2010)

Experiment design and sampling

A field survey was undertaken between 15 July and 5 August2005 Five sampling areas were randomly chosen from eachof the six vegetation communities (Table 1) representing 3ndash520ndash25 50ndash60 90ndash100 120ndash130 and 150 years of naturalsuccession The size of the plots varied with thecommunities being 20m by 20m in the forest communities5m by 5m in the shrub communities and 2m by 2m in theherbaceous communities The plots were not more than 5 kmapart and the largest relative elevation difference between twoplots was lt120m Most of the plots were facing north with aslope gradient of lt208

Soil samples were taken at five points ie the four cornersand centre of the soil sampling sites in the six successionalstages as described above Sampling using a soil-drillingsampler was done in three soil layers 0ndash20 20ndash40 and40ndash60 cm In each plot the ground litter was first removedbut not measured the soil samples were then taken at the fivesampling points and mixed together to effectively form one soilsample for each soil layer After removing the stones leaves

Natural succession improves soil N storage Soil Research 263

roots and large pieces of debris the soil samples were sievedthrough a 2-mm screen air-dried and stored at room temperaturefor determination of physical and chemical properties The bulkdensity (BD) of the different soil layers was measured using asteel soil bulk sampler 50 cm in diameter and 50 cm high (threereplicates) at points adjacent to the soil sampling plots Thusone pit was dug to 60 cm depth in the centre of the plot and threesoil BD samples were taken The original volume of each soilcore and its dry mass after oven-drying at 1058C for 48 h weremeasured

To measure roots soil sampling was repeated three times inthree soil layers (0ndash20 20ndash40 and 40ndash60 cm) in the centre ofeach plot using a 9-cm-diameter root auger The majority of theroots found in the soil samples were isolated using a 2-mm sieveThe remaining fine roots taken from the soil samples wereisolated by spreading the samples in shallow trays overfillingthe trays with water and allowing the outflow from the trays topass through a 05-mm mesh sieve No attempts were made todistinguish between living and dead roots All of the roots thusisolated were oven-dried at 658C for 48 h The fine roots (FRdiameter lt2mm) were then isolated using a vernier caliper andweighed to within 001 g

Physical and chemical analysis

Because the occurrence of coarse particles in the loessial soils ofthe study region is rare they were considered negligible (Liuet al 2011) so soil BD was calculated depending on the innerdiameter of the core sampler sampling depth and the oven-driedweight of the composite soil samples (Jia et al 2005) Soilorganic C (SOC) was assayed by dichromate oxidation(Kalembasa and Jenkinson 1973) and soil total N (TN)concentration was assayed using the Kjeldahl method(Jackson 1973) Soil water content was measuredgravimetrically and expressed as a percentage of soil water todry soil weight Soil pH was measured in distilled water mixed5 1 (by mass) with dry soil using a Delta 320 pHmeter (Mettler-Toledo Instruments (Shanghai) Ltd China) equipped with acalibrated combined glass electrode

Table 1 Geographical and vegetation characteristics at different succession stages in the Ziwuling Forest Region of the Loess PlateauAfs Abandoned farmland stage Hs herbaceous community stage Ss shrub community stage Pas pioneering arbour community stage Mfs mixed forest

community stage Cs climax community stage H S F Herb shrub forest respectively

Successionstage (year)

Quadrat size(number)

Latitude(N)

Longitude(E)

Cover()

Altitude(m)

Aspect Slope(8)

Dominant species

Afs (3ndash5) 2m by 2m(5)

36805005400 108831057400 62 1357 Half-northfacing

3ndash6 Lespedeza bicolor

Hs (20ndash25) 2m by 2m(5)


14ndash16 Bothriochloa ischaemum Lespedeza dahurica

Ss (50ndash60) 5m by 5(5)

36804057900 108831053100 85 1346 Northfacing

12ndash22 S Sophora davidii H Carex lanceolataor S Hippophae rhamnoides H Carex lanceolata

Pas (90ndash100) 20m by 20m(5)


12ndash16 S Spiraea schneideriana H Carex lanceolata

Mfs (120ndash130) 20m by 20m(5)


12ndash15 F Populus davidiana +Quercus liaotungensisS Spiraea schneideriana H Carex lanceolata

Cs (150) 20m by 20m(5)

36802056100 108832013600 95 1427 Northfacing

20ndash22 F Quercus liaotungensis S Rosa hugonisH Carex lanceolata

30(a)

(b)

0ndash20 cm

20ndash40 cm

40ndash60 cm

c

cc

b

b

b

b

a

b

b

a a

a ab

b b

a

c

b

bcc

c

b

b

b

abab

ab

bc

a

a

a

a

a

aa

25

20

15

10SO

C (

g kg

ndash1)

TN

(g

kgndash1

)

Successional stage

5

30

25

15

10

05

00Afs Hs Ss Pas Mfs Cs

20

0

Fig 1 Changes in (a) soil organic carbon (SOC) and (b) total nitrogenconcentration (TN) in three soil layers with the vegetation successiongradient at 0ndash60 cm soil depth Bars are mean and capped lines are sesample size n= 5 Letters above the bars are for comparison in the same soillayer at the different restoration stages bars with the same letter are notsignificantly different at P= 005 Afs Abandoned farmland stage Hsherbaceous community stage Ss shrub community stage Pas pioneeringarbour community stage Mfs mixed forest community stage Cs climaxcommunity stage

264 Soil Research L Deng et al

Calculation of soil N storage

Because our soil samples did not contain any coarse fraction(gt2mm) the study used the following equation to calculate soilN storage (Rytter 2012)

Ns frac14 BD TN D=10

where Ns is soil N storage (Mg handash1) BD is soil bulk density(g cmndash3) TN is soil total N concentration (g kgndash1) and D is soilthickness (cm)

Statistical analyses

One-way ANOVAwas used to analyse the mean of the same soillayers taken from across the different successional stagesDifferences were evaluated at P = 005 When significancewas observed (P lt 005) a least significant difference (lsd)post-hoc test was used to carry out multiple comparisons GLM(generalised linear model) was adopted to determine whetherthere were significant correlations between soil N storage and thesoil properties measured in the study All statistical analyseswere performed using the software program SPSS ver 170(SPSS Inc Chicago IL USA)

Results

Soil organic carbon and total nitrogen

Along the vegetation succession gradient the SOC increasedquickly and tended to be stable after the pioneering arbour

a

15

14

13

12

11

1018

16

14

12

10

86

84

82

80

78

76

8

a

a

a

a

a a a

a

a

aa

a

b

b

bbc

bcbc

bc

cdd

c cc c

cc

a aa

a

a

a

a aab

ab

ab

a

c

c

bb

b

b

b

b

b

b

b

bb

b

0ndash20 cm

20ndash40 cm

40ndash60 cm

Afs Hs Ss Pas Mfs Cs

Successional stage

pHS

WC

(

)B

D (

g cm

ndash3)

(a)

(b)

(c)

Fig 2 Changes in (a) soil bulk density (BD) (b) soil water content (SWC)and (c) soil pH in three soil layers with the vegetation succession gradientat 0ndash60 cm soil depth Bars are mean and capped lines are se samplesize n= 5 Letters above the bars are for comparison in the same soil layerat the different restoration stages bars with the same letter are notsignificantly different at P= 005 Afs Abandoned farmland stage Hsherbaceous community stage Ss shrub community stage Pas pioneeringarbour community stage Mfs mixed forest community stage Cs climaxcommunity stage

Successional stage

Rat

io (

)

Soi

l N s

tora

ge (

Mg

handash1

)

Successional stageAfs Hs Ss Pas Mfs Cs


0ndash20 cm20ndash40 cm40ndash60 cm


0ndash60 cm

10

8

6

4

2

0

100

80

60

40

20

0

Fig 3 Changes in (a) soil N storage and (b) the ratio of N storage in eachsoil layer changes with the vegetation succession gradient at 0ndash60 cm soildepth Bars are mean and capped lines are se sample size n= 5 Lettersabove the bars are for comparison in the same soil layer at the differentrestoration stages bars with the same letter are not significantly differentat P= 005 Afs Abandoned farmland stage Hs herbaceous communitystage Ss shrub community stage Pas pioneering arbour community stageMfs mixed forest community stage Cs climax community stage


community stage (Pas gt90ndash100 years) (Fig 1a) Before Pas(lt50ndash60 years) the SOC in the 0ndash20 20ndash40 and 40ndash60 cmsoils increased significantly along with vegetation succession(P lt 005) (Fig 1a) The highest and the lowest SOC contentswere in the climax community stage (Cs ~150 years) and theabandoned farmland stage (Afs 3ndash5 years) A similar trend wasobserved for TN The TN in the 0ndash20 cm soil increased withvegetation succession becoming nearly constant (Fig 1b)following Pas (gt90ndash100 years) The TN in the 20ndash40 cm soilincreased slowly and tended to be stable after the mixed forestcommunity stage (Mfs gt120ndash130 years) (Fig 1b) the TN in the40ndash60 cm soil remained almost unchanged except in the shrubcommunity stage (Ss 50ndash60 years) (Fig 1b) which may dueto the occurrence of the N2-fixing species S davidii andH rhamnoides (Table 1)

Soil bulk density

With vegetation succession BD varied mainly in the topsoillayer (0ndash20 cm) (Fig 2a) where it decreased significantly withvegetation succession (Plt 005) (Fig 2a) In the early stages ofvegetation succession (Afs and Hs) BD did not significantly

vary (P gt 005) and tended to be stable after Ss (50ndash60 years)(Fig 2a) The BD in the 20ndash40 layers showed no significantdifference in the six restoration stages (P gt 005) (Fig 2a) TheBD in the 40ndash60 cm layers at Hs was significantly higher thanother stages

Soil N storage

In the long term (~150 years) N storage in the 0ndash60 cm soil layerincreased rapidly and tended to be stable after Ss (50ndash60 years)and achieved its highest value at roughly Mfs (120ndash130 years)(Fig 3a) The soil N accumulated mainly in the early successionstages and significantly increased before Ss (P lt 005) FromAfs to Cs the N storage values in the 0ndash60 cm soil were 58 7277 82 85 and 84Mg handash1 respectively The N storage wasgreater in the upper than the lower soil layers (Fig 3b) The Nstorage in the 0ndash20 cm soil increased with vegetation successionand that in the 20ndash40 cm soil was higher in the latter stages thanthe early stage The N storage in the 40ndash60 cm soil layer showeda similar trend to TN in the 40ndash60 cm soil layer which wasnearly unchanged except at Ss (50ndash60 years) (Fig 3a)

Ns = 02452SOC ndash 02126R2 = 08783P lt 001

Ns = 00489SOC + 14246R2 = 05429P gt 05 ns

Ns = ndash00085SOC + 14596R2 = 00040P gt 005 ns

Ns = 24803TN ndash 00160R2 = 09750P lt 001

Ns = 25035TN ndash 00800R2 = 09733P lt 001

Ns = 19973TN + 05871R2 = 09749P lt 001

0ndash20 cm

20ndash40 cm

40ndash60 cm40ndash60 cm

20ndash40 cm

0ndash20 cm

(a)

(b)

Soi

l N s

tora

ge (

Mg

handash1

)

(c)

(d )

(e)

(f )

SOC (g kgndash1) TN (g kgndash1)

Fig 4 Relationship between soil N storage (Ns) and (andashc) soil organic carbon (SOC) and (dndashf) total Nconcentration (TN) in three soil layers with the vegetation succession gradient at 0ndash60 cm soil depthCapped horizontal and vertical lines are the SE


DiscussionSoil organic C is an essential contributor to soil physical andchemical properties (Schoenholtz et al 2000) It is also a keyelement in the process of trapping atmospheric CO2 in terrestrialecosystems through primary production (Post and Kwon 2000)Changes in land use caused by vegetation restoration probablyenhance the C sequestration capacity of terrestrial ecosystems onthe Loess Plateau (Deng et al 2013a) and soil C and N showsignificant positive correlations in the process of the vegetationrestoration (Deng et al 2013b)

In our study the N storage in the 0ndash60 cm soil layersignificantly increased (Plt 005) especially in the 0ndash20 cmsoil (Fig 3) indicating the accumulation of N through naturalregeneration These results agree with those of Wang et al(2011b) who studied changes to the physicochemical propertiesof topsoil during natural succession on abandoned farmlandThe accumulation of nutrients and organic matter in surfacesoils results from complex interactions between biotic processesmoderated by plants and soil biota and abiotic processes drivenby environmental processes (Hooper et al 2000) Storage ofN was largely determined by organic C in the soil There was a

marked correlation between the N storage value and SOC inthe 0ndash20 cm soil layer (Fig 4a) but the correlation was notsignificant in the deeper layers (Fig 4b c) The altered sizecomposition and dynamics of microbial communities and alteredvegetation biomass along the vegetation succession gradientmay influence the biogeochemical cycles of soil C and othernutrients (Cleveland et al 2004 Wang et al 2011a) The varieddistribution of N storage in the different soil layers may be due todifferences in the SOC in those soil layers

Aboveground vegetation plays an important role inregulating the biogeochemistry of ecosystems through C-fixing and use of nutrients and by preventing the loss ofnutrients through disturbance (Bormann and Sidle 1990 Liet al 2007) It is also clear from these results that vegetationsuccession has a great impact on N storage In our study therewas a markedly positive correlation between N storage and theTN in the soil (Plt 0001) (Fig 4dndashf) indicating that TN is thedetermining factor in Ns Deng et al (2013a) found that soil Cstorage tended to be stable at the latter stage of forest succession(gt50 years) in the Ziwuling Forest Region indicating a strongcorrelation between soil C and N storage Deng et al (2013b)

Ns = ndash88931BD + 14637R2 = 04208P gt 005 ns

Ns = 02760SW + 07767R2 = 04150P gt 005 ns

Ns = ndash42967pH + 382608R2 = 05875P gt 005 ns


Ns = 00654SW + 11555R2 = 02926P gt 005 ns

Ns = 36332BD ndash 23527R2 = 01631P gt 005 ns

(a) (b) (c)

(d ) (e) (f )

(g) (h) (i )

Ns = 06325pHndash37544R2 = 00379P gt 005 ns

Ns = ndash004628SW = 19790R2 = 00991P gt 005 ns


Soi

l N s

tora

ge (

Mg

handash1

)

pH

BD (g cmndash3)

SWC ()

0ndash20 cm

0ndash20 cm

0ndash20 cm

0ndash20 cm 40ndash60 cm

40ndash60 cm

40ndash60 cm

Fig 5 Relationship between soil N storage (Ns) and (andashc) soil bulk density (BD) (dndashf) soil water content (SWC) and (gndashi) soil pH in three soil layerswith the vegetation succession gradient at 0ndash60 cm soil depth Capped horizontal and vertical lines are the SE


also found that soil C storage and N storage were significantlyand positively correlated In addition Luo et al (2004) reportedthat long-term terrestrial C sequestration is related to soiltherefore we should focus on long-term soil N dynamics topredict terrestrial C sequestration on the Loess Plateau Inaddition in deeper soil layers (40ndash60 cm) SOC increased butTN was almost unchanged following natural successionindicating that soil organic matter (SOM) input more thandecomposition and the input and uptake of N in the SOMare balanced in the process of vegetation succession

Soil BD soil water content (SWC) and pH were not thedetermining factors behind the N storage values although theyhad non-significant positive positive and negative correlationsrespectively (Fig 5) Generally soil depth is fixed meaning thatN storage values are determined by TN and soil BD Withvegetation succession 0ndash20 cm soil BD did not significantlyvary (Pgt 005) in the early stages of vegetation succession (Afsand Hs) but BD significantly decreased until Ss (P lt 005) andthen tended to be stable after Ss (50ndash60 years) (Fig 2a) meaningthat our findings agree with Wang et al (2011a) The N storagevalue was negatively correlated with BD (Fig 5a) leaving TNitself to act as the key factor affecting N storage during theprocess of restoration (Figs 1 and 4)

On the Loess Plateau porosity is a key attribute of soilstructure affecting the soil reservoir under conditions of naturalvegetation recovery (Zhao et al 2010) Soil moisture is anindicator of SOM accumulation (Deng et al 2013a) andSOM determines SOC and TN in the soil Therefore amarkedly positive correlation between N storage SOC andTN strongly suggests that soil moisture is a reflection of theN storage value However the study did not indicate significantrelationships between SWC and N storage along the vegetationsuccession gradient (Fig 5dndashf) although the SWC was higher

in the latter stage (gt50ndash60 years) than earlier (lt50ndash60 years)(Fig 2b) This demonstrated instead that SWC in the soil was notthe key factor influencing N storage in the Ziwuling ForestRegion Rather as Deng et al (2013a) point out SWC is morelikely related to the annual rainfall which is close to 600mmin the Ziwuling Forest Region in the south of the LoessPlateau where the climate is relatively humid compared withthe other drier regions of the Loess Plateau Wang et al (2011b)reported that soil pH decreased after farmland was convertedto forest and our findings were similar (Fig 2c) in that the Nstorage value was non-significantly negatively correlatedwith pH (Fig 5g h) The BD SWC and pH are not thedetermining factors behind the N storage values However inthe topsoil (0ndash20 cm) they did show stronger negative positiveand negative correlations respectively which indicates thatvegetation restoration can increase N storage and improve thephysicochemical properties of topsoil There was littlecorrelation in the 40ndash60 cm soil (Fig 5c f i) indicating thatthese mechanisms did not function in the lower soils (Figs 2 5)

600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage

Fig 6 Changes in fine roots in three soil layers with the vegetationsuccession gradient at 0ndash60 cm soil depth Bars mean and capped linesare se sample size n= 5 Letters above the bars are for comparison thesame soil layer at the different restoration stages bars with the same letter arenot significantly different at P = 005) Afs Abandoned farmland stageHs herbaceous community stage Ss shrub community stage Paspioneering arbour community stage Mfs mixed forest community stageCs climax community stage

(a)

(b)

(c)

Fine root (g mndash2)

Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns

Ns = ndash00035FR + 15360R2 = 00591P gt 005 ns

0ndash20 cm

20ndash40 cm

40ndash60 cm

Fig 7 Relationship between soil N storage (Ns) and fine roots in three soillayers with the vegetation succession gradient at 0ndash60 cm soil depth Cappedhorizontal and vertical lines are the SE


There may be a range of potential mechanisms throughwhich N storage values in the topsoil are increased throughnatural regeneration A prime candidate is the return of soil Nfrom increased aboveground biomass and litter As soil N inputis mainly derived from the decomposition of litter (Wang et al2011b) primary productivity is the main driver of soil Naccumulation and results primarily in an increase in Nstorage in the topsoil (Wang et al 2011b) We also found amarked increase in fine roots in the topsoil (0ndash20 cm) (Fig 6)with which N storage showed a significant positive correlation(P lt 005) (Fig 7a) In addition an increased N input byrhizodeposition (dead roots mycorrhizae and exudates) isan important element of soil N accumulation (Langley andHungate 2003 Prietzel and Bachmann 2012) Fine roots areeasy to turnover and their dynamics controls a dominant fluxof C from vegetation into soils (McCormack et al 2013)which in turn leads into N dynamics because of the strongpositive correlative relationship between C and N (Luo et al2004 2006) Because fine roots increased with vegetationsuccession (Fig 6) their turnover (dead roots) increased Inaddition changes in the vegetation composition and dominantplant functional group affect N sequestration in the soil(De Deyn et al 2008 Mareschal et al 2010) In agreementwith Deng et al (2013b) natural regeneration had a direct effecton the dominant vegetation species vegetation cover heightand above- and belowground biomass and thus has a positiveeffect on soil N Increased atmospheric deposition andorsubsoil mining also increased soil N storage (Feldpauschet al 2004)

Conclusions

The impact of long-term vegetation succession on the soil Npool was significantly different from grassland to forest The Nstorage value increased with the length of vegetation successionOver the long term (~150 years) N storage in secondaryforest succession in the soil at 0ndash60 cm increased rapidly andtended to be stable after the shrub communities stage (Ss)(~50ndash60 years) Soil N accumulated mainly in the earlysuccession stages it had significantly increased before SsThe results suggest that changes in N storage values were theresult of the accumulation of SOC and TN and primaryproductivity during forest succession This capacity has beenshown positively correlated with forest succession on the LoessPlateau China In light of this important finding it is clearthat restored soil can play an extensive role in N-sink effortsFurther study is now required to understand what drives changesin N storage over the long-term the time-frame in which theprocess of sequestration occurs The combined mechanismsof increased primary productivity litter decompositionrhizodeposition changes in vegetation composition andmicrobial communities increased atmospheric deposition andsubsoil mining were likely responsible for increasing the storageof soil N along this vegetation gradient

Acknowledgements

The study was funded by the Major Program of the National Natural ScienceFoundation of China (41390463 and 41301610) and the Key ResearchProgram of the Chinese Academy of Sciences (KZZD-EW-04)

References

Bormann BT Sidle RC (1990) Changes in productivity and distribution ofnutrients in a chronosequence at Glacier Bay National Park AlaskaJournal of Ecology 78 561ndash578 doi1023072260884

Burke IC Lauenroth WK Cunfer G Barrett JE Mosier A Lowe P (2002)Nitrogen in the central grasslands region of the United States Bioscience52 813ndash823 doi1016410006-3568(2002)052[0813NITCGR]20CO2

Chapuis-Lardy L Wrage N Metay A Chotte J Bernoux M (2007) Soils asink for N2O A review Global Change Biology 13 1ndash17 doi101111j1365-2486200601280x

Chen CD (1958) The vegetation and its roles in soil and water conservationin the secondary forest area in the boundary of Shaanxi and Gansuprovinces Acta Phytoecologica et Geobotanica Sinica 2 152ndash223[in Chinese]

Cheng JM Cheng J Shao HB Zhao LP Yang XM (2012) Soil seed banksand forest succession direction reflect soil quality in Ziwuling MountainLoess Plateau China Clean ndash Soil Air Water 40 140ndash147 doi101002clen201000377

Cleveland CC Townsend AR Constance BC Ley RE Schmidt SK (2004)Soil microbial dynamics in Costa Rica seasonal and biogeochemicalconstraints Biotropica 36 184ndash195

Cleveland CC Reed SC Townsend AR (2006) Nutrient regulation oforganic matter decomposition in a tropical rain forest Ecology 87492ndash503 doi10189005-0525

De Deyn GB Cornelissen JHC Bardgett RD (2008) Plant functional traitsand soil carbon sequestration in contrasting biomes Ecology Letters 11516ndash531 doi101111j1461-0248200801164x

Deng L Wang KB Chen ML Shangguan ZP Sweeney S (2013a) Soilorganic carbon storage capacity positively related to forest successionon the Loess Plateau China Catena 110 1ndash7 doi101016jcatena201306016

Deng L Shangguan ZP Sweeney S (2013b) Changes in soil carbon andnitrogen following land abandonment of farmland on the Loess PlateauChina PLoS ONE 8 e71923 doi101371journalpone0071923

Feldpausch TR Rondon MA Fernandes ECM Riha SJ Wandelli E (2004)Carbon and nutrient accumulation in secondary forests regenerating onpastures in central Amazonia Ecological Applications 14 S164ndashS176doi10189001-6015

Galloway JN Dentener FJ Capone DG Boyer EW Howarth RWSeitzinger SP Asner GP Cleveland CC Green PA Holland EAKarl DM Michaels AF Porter JH Townsend AR Voumlosmarty CJ(2004) Nitrogen cycles Past present and future Biogeochemistry 70153ndash226 doi101007s10533-004-0370-0

Giardina CP Ryan MG Binkley D Fownes JH (2003) Primaryproduction and carbon allocation in relation to nutrient supply in atropical experimental forest Global Change Biology 9 1438ndash1450doi101046j1365-2486200300558x

Gundale MJ Metlen KL Fiedler CE DeLuca TH (2006) Nitrogenspatial heterogeneity influences diversity following restoration in aponderosa pine forest Montana Ecological Applications 16479ndash489 doi1018901051-0761(2006)016[0479NSHIDF]20CO2

Hooper DU Bignell DE Brown VK (2000) Interactions betweenaboveground and belowground biodiversity in terrestrial ecosystemspatterns mechanisms and feedbacks Bioscience 50 1049ndash1061doi1016410006-3568(2000)050[1049IBAABB]20CO2

Hu S Chapin FS III FirestoneMK Field CB Chiariello NR (2001) Nitrogenlimitation of microbial decomposition in a grassland under elevated CO2Nature 409 188ndash191 doi10103835051576

Jackson ML (1973) lsquoSoil chemical analysisrsquo (Prentice-Hall EnglewoodCliffs NJ)

Jia GM Cao J Wang CY Wang G (2005) Microbial biomass and nutrientsin soil at the different stages of secondary forest succession in Ziwulingnorthwest China Forest Ecology and Management 217 117ndash125doi101016jforeco200505055


Kalembasa SJ Jenkinson DS (1973) A comparative study of titrimetric andgravimetric methods for the determination of organic carbon in soilJournal of the Science of Food and Agriculture 24 1085ndash1090doi101002jsfa2740240910

Langley JA Hungate B (2003) Mycorrhizal controls on belowground litterquality Ecology 84 2302ndash2312 doi10189002-0282

Li XR Kong DS Tan HJ Wang XP (2007) Changes in soil and vegetationfollowing stabilisation of dunes in the southeastern fringe of the TenggerDesert China Plant and Soil 300 221ndash231 doi101007s11104-007-9407-1

Liu SL Guo XD Fu BJ Lian G Wang J (2007) The effect of environmentalvariables on soil characteristics at different scales in the transition zoneof the Loess Plateau in China Soil Use and Management 23 92ndash99doi101111j1475-2743200600064x

Liu ZP Shao MA Wang YQ (2011) Effect of environmental factors onregional soil organic carbon stocks across the Loess Plateau regionChina Agriculture Ecosystems amp Environment 142 184ndash194doi101016jagee201105002

Luo YQ Su B Currie WS Dukes JS Finzi A Hartwig U Hungate BMcmurtrie RE Oren R Parton WJ Pataki DE Shaw MR Zak DRField CB (2004) Progressive nitrogen limitation of ecosystemresponses to rising atmospheric carbon dioxide Bioscience 54731ndash739 doi1016410006-3568(2004)054[0731PNLOER]20CO2

Luo YQ Field CB Jackson RB (2006) Does nitrogen constrain carboncycling or does carbon input stimulate nitrogen cycling Ecology 873ndash4 doi10189005-0923

Mareschal L Bonnaud P Turpault MP Ranger J (2010) Impact of commonEuropean tree species on the chemical and physicochemical propertiesof fine earth an unusual pattern European Journal of Soil Science 6114ndash23 doi101111j1365-2389200901206x

McCormack ML Eissenstat DM Prasad AM Smithwick EAH (2013)Regional scale patterns of fine root lifespan and turnover undercurrent and future climate Global Change Biology 19 1697ndash1708doi101111gcb12163

Prietzel J Bachmann S (2012) Changes in soil organic C and N stocks afterforest transformation from Norway spruce and Scots pine into Douglasfir Douglas firspruce or European beech stands at different sites inSouthern Germany Forest Ecology and Management 269 134ndash148doi101016jforeco201112034

Post WM Kwon KC (2000) Soil carbon sequestration and land-use changeprocesses and potential Global Change Biology 6 317ndash327doi101046j1365-2486200000308x

Rytter RM (2012) Stone and gravel contents of arable soils influenceestimates of C and N stocks Catena 95 153ndash159 doi101016jcatena201202015

Schlesinger WH Lichter J (2001) Limited carbon storage in soil and litter ofexperimental forest plots under increased atmospheric CO2 Nature 411466ndash469 doi10103835078060

Schoenholtz SH van Miegroet H Burger JA (2000) A review of chemicaland physical properties as indicators of forest soil quality challengesand opportunities Forest Ecology and Management 138 335ndash356doi101016S0378-1127(00)00423-0

Tian HQ Wang SQ Liu JY Pan SF Chen H Zhang C Shi XZ (2006)Patterns of soil nitrogen storage in ChinaGlobal Biogeochemical Cycles20 GB1001 doi1010292005GB002464

Wang KB Shao RX Shangguan ZP (2010) Changes in species richness andcommunity productivity during succession on the Loess Plateau (China)Polish Journal of Ecology 58 549ndash558

Wang B Liu GB Xue S Zhu BB (2011a) Changes in soil physico-chemicaland microbiological properties during natural succession on abandonedfarmland in the Loess Plateau Environmental Earth Sciences 62915ndash925 doi101007s12665-010-0577-4

WangWJ Qiu L Zu YG Su DX An J Wang HY Zheng GY SunW ChenXQ (2011b) Changes in soil organic carbon nitrogen pH and bulkdensity with the development of larch (Larix gmelinii) plantations inChina Global Change Biology 17 2657ndash2676 doi101111j1365-2486201102447x

Yang YH Ma WH Mohammat A Fang JY (2007) Storage patterns andcontrols of soil nitrogen in ChinaPedosphere 17 776ndash785 doi101016S1002-0160(07)60093-9

Zhao SW Zhao YG Wu JS (2010) Quantitative analysis of soil pores undernatural vegetation successions on the Loess Plateau Science China EarthSciences 53 617ndash625

Zou HY Liu GB Wang HS (2002) The vegetation development in NorthZiwuling forest region in last fifty years Acta Botanica Boreali-Occidentalia Sinica 22 1ndash8 [in Chinese]


wwwpublishcsiroaujournalssr

attenuating atmospheric CO2 by stimulating the accumulation offorest biomass (Giardina et al 2003) Therefore elucidating theN dynamics in soils has important implications both forthe sustainable management of land resources regionally andpredictions of future C and N cycling globally

The Loess Plateau China is an area suffering severe soiland water losses (Liu et al 2007) Many studies report thatsecondary succession can recover the properties of degradedsoil and maintain soil fertility (Feldpausch et al 2004 Jia et al2005 Wang et al 2011a Deng et al 2013a) For this reasonunderstanding secondary forest succession processes in thecentral region of the Loess Plateau is becoming increasinglyimportant (Jia et al 2005 Deng et al 2013a) To improve thefrail natural ecosystems of the Loess Plateau and alleviatedegradation the Chinese government launched a series ofconservation programs aimed at restoring agricultural land toforest and grassland this is the case for the study area wherefarmland had already been abandoned and the process ofnatural secondary succession was under way With regard tothese ecosystem functions the retention of C and N in soil iscrucial (Prietzel and Bachmann 2012) Consequentlysuccession in the Ziwuling secondary forest has receivedsignificant attention from Chinese scientists (Zou et al 2002Jia et al 2005 Wang et al 2010 Deng et al 2013a) Severalstudies have focused on changes to aboveground vegetationduring secondary forest succession in the centre of the LoessPlateau (Zou et al 2002 Wang et al 2010) Deng et al (2013a)studied soil C-storage responses to forest succession in theZiwuling Forest Region of the Loess Plateau However thereis little available information on soil N storage dynamics atdifferent stages of natural regeneration in the area

We hypothesised that natural regeneration significantlyaffects soil N conditions in the central region of the LoessPlateau To test our hypothesis and evaluate N dynamicsduring succession from managed to natural communities insemi-arid regions the objectives of this study were (i) toestimate soil N storage along a long-term natural regenerationgradient and (ii) to discuss the contributing factors andmechanisms controlling soil N storage

Materials and methods

The Ziwuling Forest Region (338500ndash368500N 1078300ndash1098400E 1100ndash1756m asl) is in the hinterland of theLoess Plateau The study area has landforms typical of hilly-gully loess topography and a mid-temperate continentalmonsoon climate The study was conducted on theLianjiabian Forest Farm of the Heshui General Forest Farmof Gansu (358030ndash368370N 1088100ndash1098180E 1211ndash1453masl) covering a total area of 23 000 km2 The arearsquos meanannual temperature is 108C (1960ndash2010) and the mean annualrainfall 587mm (1960ndash2010) (Deng et al 2013a) The regionrsquossoils are largely loessial having developed from primitive orsecondary loess parent materials (Cheng et al 2012) which areevenly distributed 50ndash130m above red earth consisting ofcalcareous cinnamon soil (Jia et al 2005) Soil pH rangesfrom 792 to 831 (Deng et al 2013a) The area is covered inspecies-rich uniform forests with a forest canopy density of80ndash95 (Deng et al 2013a) and the distribution of ground

litter was uniform The regionrsquos natural biomes are deciduousbroadleaf forests of which the climax vegetation is the Quercusliaotungensis Koidz forest Throughout the region Populusdavidiana Dode and Betula platyphylla Suk communitiesdominate the pioneer forests Sophora davidii (Franch)Skeels Hippophae rhamnoides (Linn) Rosa xanthina Lindland Spiraea pubescens Turcz are the main shrub species andBothriochloa ischaemun (Linn) Keng Carex lanceolata BoottPotentilla chinensis (Ser) Lespedeza dahurica (Laxm) Schindland Stipa bungeana Trin are the main herbaceous species

The post-abandonment secondary forests regeneratednaturally on abandoned farmland after many local inhabitantsemigrated from the Ziwuling Forest Region during a nationalconflict in 1842ndash1866 Chen (1958) who investigated thevegetation recovery of the Ziwuling Forest Region inthe 1950s determined that P davidiana made up 70 of thevegetation cover after about a century Zou et al (2002)investigated the vegetation succession three times (19621982 and 2000) They found that Q liaotungensis forestshad replaced P davidiana forests within 50 years Thusthe recovery period for Q liaotungensis forests was~150 years (Wang et al 2010) In the 20 or more yearsranging from 1940 to the 1960s famine war and disastersled some people to immigrate and reclaim land in the regionThe process of human emigration and immigration in whicharable land has been abandoned more than once in differentareas has yielded different successional stages on the degradedsloping land Based on previous research we chose sixcommunities that were at different stages of succession in theZiwuling Forest Region Two methods were used to determinethe ages of the communities For the shrub and herbaceouscommunities with lt60 years of succession we verified thelength of time both by visiting local elders and by referringto land contracts between farmers and the government Forthe forest community which had gt60 years of recovery wedetermined the length of time by boring tree rings and checkingrelated written sources (Wang et al 2010)

Experiment design and sampling

A field survey was undertaken between 15 July and 5 August2005 Five sampling areas were randomly chosen from eachof the six vegetation communities (Table 1) representing 3ndash520ndash25 50ndash60 90ndash100 120ndash130 and 150 years of naturalsuccession The size of the plots varied with thecommunities being 20m by 20m in the forest communities5m by 5m in the shrub communities and 2m by 2m in theherbaceous communities The plots were not more than 5 kmapart and the largest relative elevation difference between twoplots was lt120m Most of the plots were facing north with aslope gradient of lt208

Soil samples were taken at five points ie the four cornersand centre of the soil sampling sites in the six successionalstages as described above Sampling using a soil-drillingsampler was done in three soil layers 0ndash20 20ndash40 and40ndash60 cm In each plot the ground litter was first removedbut not measured the soil samples were then taken at the fivesampling points and mixed together to effectively form one soilsample for each soil layer After removing the stones leaves










Latitude(N)

Longitude(E)

Cover()

Altitude(m)

Aspect Slope(8)

Dominant species








36804057900 108831053100 85 1346 Northfacing





Mfs (120ndash130) 20m by 20m(5)



Cs (150) 20m by 20m(5)

36802056100 108832013600 95 1427 Northfacing


30(a)

(b)

0ndash20 cm

20ndash40 cm

40ndash60 cm

c

cc

b

b

b

b

a

b

b

a a

a ab

b b

a

c

b

bcc

c

b

b

b

abab

ab

bc

a

a

a

a

a

aa

25

20

15

10SO

C (

g kg

ndash1)

TN

(g

kgndash1

)

Successional stage

5

30

25

15

10

05


20

0









Results



a

15

14

13

12

11

1018

16

14

12

10

86

84

82

80

78

76

8

a

a

a

a

a a a

a

a

aa

a

b

b

bbc

bcbc

bc

cdd

c cc c

cc

a aa

a

a

a

a aab

ab

ab

a

c

c

bb

b

b

b

b

b

b

b

bb

b

0ndash20 cm

20ndash40 cm

40ndash60 cm


Successional stage

pHS

WC

(

)B

D (

g cm

ndash3)

(a)

(b)

(c)


Successional stage

Rat

io (

)

Soi

l N s

tora

ge (

Mg

handash1

)





0ndash60 cm

10

8

6

4

2

0

100

80

60

40

20

0




Soil bulk density



Soil N storage


Ns = 02452SOC ndash 02126R2 = 08783P lt 001

Ns = 00489SOC + 14246R2 = 05429P gt 05 ns


Ns = 24803TN ndash 00160R2 = 09750P lt 001

Ns = 25035TN ndash 00800R2 = 09733P lt 001

Ns = 19973TN + 05871R2 = 09749P lt 001

0ndash20 cm

20ndash40 cm


20ndash40 cm

0ndash20 cm

(a)

(b)

Soi

l N s

tora

ge (

Mg

handash1

)

(c)

(d )

(e)

(f )









Ns = 02760SW + 07767R2 = 04150P gt 005 ns



Ns = 00654SW + 11555R2 = 02926P gt 005 ns


(a) (b) (c)

(d ) (e) (f )

(g) (h) (i )




Soi

l N s

tora

ge (

Mg

handash1

)

pH

BD (g cmndash3)

SWC ()

0ndash20 cm

0ndash20 cm

0ndash20 cm


40ndash60 cm

40ndash60 cm







600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage


(a)

(b)

(c)


Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns


0ndash20 cm

20ndash40 cm

40ndash60 cm




Conclusions


Acknowledgements


References



















































Latitude(N)

Longitude(E)

Cover()

Altitude(m)

Aspect Slope(8)

Dominant species








36804057900 108831053100 85 1346 Northfacing





Mfs (120ndash130) 20m by 20m(5)



Cs (150) 20m by 20m(5)

36802056100 108832013600 95 1427 Northfacing


30(a)

(b)

0ndash20 cm

20ndash40 cm

40ndash60 cm

c

cc

b

b

b

b

a

b

b

a a

a ab

b b

a

c

b

bcc

c

b

b

b

abab

ab

bc

a

a

a

a

a

aa

25

20

15

10SO

C (

g kg

ndash1)

TN

(g

kgndash1

)

Successional stage

5

30

25

15

10

05


20

0









Results



a

15

14

13

12

11

1018

16

14

12

10

86

84

82

80

78

76

8

a

a

a

a

a a a

a

a

aa

a

b

b

bbc

bcbc

bc

cdd

c cc c

cc

a aa

a

a

a

a aab

ab

ab

a

c

c

bb

b

b

b

b

b

b

b

bb

b

0ndash20 cm

20ndash40 cm

40ndash60 cm


Successional stage

pHS

WC

(

)B

D (

g cm

ndash3)

(a)

(b)

(c)


Successional stage

Rat

io (

)

Soi

l N s

tora

ge (

Mg

handash1

)





0ndash60 cm

10

8

6

4

2

0

100

80

60

40

20

0




Soil bulk density



Soil N storage


Ns = 02452SOC ndash 02126R2 = 08783P lt 001

Ns = 00489SOC + 14246R2 = 05429P gt 05 ns


Ns = 24803TN ndash 00160R2 = 09750P lt 001

Ns = 25035TN ndash 00800R2 = 09733P lt 001

Ns = 19973TN + 05871R2 = 09749P lt 001

0ndash20 cm

20ndash40 cm


20ndash40 cm

0ndash20 cm

(a)

(b)

Soi

l N s

tora

ge (

Mg

handash1

)

(c)

(d )

(e)

(f )









Ns = 02760SW + 07767R2 = 04150P gt 005 ns



Ns = 00654SW + 11555R2 = 02926P gt 005 ns


(a) (b) (c)

(d ) (e) (f )

(g) (h) (i )




Soi

l N s

tora

ge (

Mg

handash1

)

pH

BD (g cmndash3)

SWC ()

0ndash20 cm

0ndash20 cm

0ndash20 cm


40ndash60 cm

40ndash60 cm







600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage


(a)

(b)

(c)


Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns


0ndash20 cm

20ndash40 cm

40ndash60 cm




Conclusions


Acknowledgements


References

















































Results



a

15

14

13

12

11

1018

16

14

12

10

86

84

82

80

78

76

8

a

a

a

a

a a a

a

a

aa

a

b

b

bbc

bcbc

bc

cdd

c cc c

cc

a aa

a

a

a

a aab

ab

ab

a

c

c

bb

b

b

b

b

b

b

b

bb

b

0ndash20 cm

20ndash40 cm

40ndash60 cm


Successional stage

pHS

WC

(

)B

D (

g cm

ndash3)

(a)

(b)

(c)


Successional stage

Rat

io (

)

Soi

l N s

tora

ge (

Mg

handash1

)





0ndash60 cm

10

8

6

4

2

0

100

80

60

40

20

0




Soil bulk density



Soil N storage


Ns = 02452SOC ndash 02126R2 = 08783P lt 001

Ns = 00489SOC + 14246R2 = 05429P gt 05 ns


Ns = 24803TN ndash 00160R2 = 09750P lt 001

Ns = 25035TN ndash 00800R2 = 09733P lt 001

Ns = 19973TN + 05871R2 = 09749P lt 001

0ndash20 cm

20ndash40 cm


20ndash40 cm

0ndash20 cm

(a)

(b)

Soi

l N s

tora

ge (

Mg

handash1

)

(c)

(d )

(e)

(f )









Ns = 02760SW + 07767R2 = 04150P gt 005 ns



Ns = 00654SW + 11555R2 = 02926P gt 005 ns


(a) (b) (c)

(d ) (e) (f )

(g) (h) (i )




Soi

l N s

tora

ge (

Mg

handash1

)

pH

BD (g cmndash3)

SWC ()

0ndash20 cm

0ndash20 cm

0ndash20 cm


40ndash60 cm

40ndash60 cm







600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage


(a)

(b)

(c)


Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns


0ndash20 cm

20ndash40 cm

40ndash60 cm




Conclusions


Acknowledgements


References












































Soil bulk density



Soil N storage


Ns = 02452SOC ndash 02126R2 = 08783P lt 001

Ns = 00489SOC + 14246R2 = 05429P gt 05 ns


Ns = 24803TN ndash 00160R2 = 09750P lt 001

Ns = 25035TN ndash 00800R2 = 09733P lt 001

Ns = 19973TN + 05871R2 = 09749P lt 001

0ndash20 cm

20ndash40 cm


20ndash40 cm

0ndash20 cm

(a)

(b)

Soi

l N s

tora

ge (

Mg

handash1

)

(c)

(d )

(e)

(f )









Ns = 02760SW + 07767R2 = 04150P gt 005 ns



Ns = 00654SW + 11555R2 = 02926P gt 005 ns


(a) (b) (c)

(d ) (e) (f )

(g) (h) (i )




Soi

l N s

tora

ge (

Mg

handash1

)

pH

BD (g cmndash3)

SWC ()

0ndash20 cm

0ndash20 cm

0ndash20 cm


40ndash60 cm

40ndash60 cm







600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage


(a)

(b)

(c)


Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns


0ndash20 cm

20ndash40 cm

40ndash60 cm




Conclusions


Acknowledgements


References
















































Ns = 02760SW + 07767R2 = 04150P gt 005 ns



Ns = 00654SW + 11555R2 = 02926P gt 005 ns


(a) (b) (c)

(d ) (e) (f )

(g) (h) (i )




Soi

l N s

tora

ge (

Mg

handash1

)

pH

BD (g cmndash3)

SWC ()

0ndash20 cm

0ndash20 cm

0ndash20 cm


40ndash60 cm

40ndash60 cm







600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage


(a)

(b)

(c)


Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns


0ndash20 cm

20ndash40 cm

40ndash60 cm




Conclusions


Acknowledgements


References















































600

500

400

300

200Fin

e ro

ots

(g m

ndash2)

100



Successional stage


(a)

(b)

(c)


Soi

l N s

tora

ge (

Mg

handash1

)

Ns = 00053FR + 25113R2 = 07960P lt 005

Ns = 00024FR + 17865R2 = 02643P gt 005 ns


0ndash20 cm

20ndash40 cm

40ndash60 cm




Conclusions


Acknowledgements


References












































Conclusions


Acknowledgements


References











































long-term natural succession improves nitrogen storage capacity of soil on the loess plateau, china

Documents