The manipulation of organic residues affects tree growth and heterotrophic CO2 efflux in a tropical Eucalyptus plantation

Forest Ecology and Management 301 (2013) 79–88

Contents lists available at SciVerse ScienceDirect

Forest Ecology and Management

journal homepage: www.elsevier .com/locate / foreco

The manipulation of organic residues affects tree growth and heterotrophicCO2 efflux in a tropical Eucalyptus plantation

Antoine Versini a,b,c, Yann Nouvellon a,d, Jean-Paul Laclau a,e, Antoine Kinana b, Louis Mareschal a,b,Bernd Zeller c, Jacques Ranger c, Daniel Epron a,b,f,g,⇑a CIRAD, UMR111, Ecologie Fonctionnelle & Biogéochimie des Sols & Agro-écosystèmes, F34060 Montpellier, Franceb Centre de Recherche sur la Durabilité et la Productivité des Plantations Industrielles, BP 1291 Pointe Noire, Congoc INRA, UR1138, Biogéochimie des Ecosystèmes forestiers, Champenoux, Franced Universidade de São Paulo, Departmento de Ciências Atmosféricas, 05508-090 São Paulo, Brazile Universidade de São Paulo, Departamento de Ecologia, 05508-900 São Paulo, Brazilf Université de Lorraine, UMR1137, Ecologie et Ecophysiologie Forestières, Faculté des Sciences, F-54500 Vandoeuvre-les-Nancy, Franceg INRA, UMR1137, Ecologie et Ecophysiologie Forestières, Centre de Nancy, F-54280 Champenoux, France

a r t i c l e i n f o

Article history:Available online 25 August 2012

Keywords:Tropical forest plantationHarvest organic residueSoil respirationEucalyptus growthSoil organic matter mineralization

0378-1127/$ - see front matter � 2012 Elsevier B.V. Ahttp://dx.doi.org/10.1016/j.foreco.2012.07.045

⇑ Corresponding author at: Université de LorraEcophysiologie Forestières, Faculté des Sciences, F-5France. Tel.: +33 (0) 383684249; fax: +33 (0) 383684

E-mail address: [email protected] (D. E

a b s t r a c t

Fast-growing plantations are increasingly being established on tropical soils, where fertility is largelysupported by soil organic matter (SOM) and where different management options of harvest organic res-idues is thought to impact the long-term sustainability of these plantations. The objectives of this studywere: (1) to quantify the effect of contrasting methods of organic residue management on tree growthand soil CO2 effluxes in the first 2 years after planting and (2) to evaluate the impact of organic residuemanipulations on the mineralization of soil organic matter over the length of the experiment. Three treat-ments were setup in 0.125 ha plots and replicated in three blocks at the harvesting of a Congolese Euca-lyptus stand, resulting in an aboveground organic residue mass ranging from 0 to 6.3 kg m�2. Themineralization of SOM was deduced in each treatment by partitioning sources of soil CO2 effluxes usingdecomposition experiments and by upscaling specific root respiration. Soil CO2 effluxes were greatlyaffected by seasons and organic residue manipulation, although there were no significant changes in top-soil water content and topsoil temperature over most of the study period. Aboveground organic residuewas the first contributor to soil CO2 efflux in the two treatments with a litter layer. Organic residue man-agement did not significantly influence the mineralization of SOM in our study, probably due to the lowquality of Eucalyptus litter, or to the hypothetical lack of dissolved organic carbon transfers from litter tosoil. A strong relationship was found between cumulative heterotrophic CO2 efflux and tree growth, sup-porting the hypothesis that the early growth of Eucalyptus trees in a sandy tropical soil is largely depen-dent on the nutrients released by the decomposition of organic residues.

� 2012 Elsevier B.V. All rights reserved.

1. Introduction

Forest plantations expanded at a rate of almost 5 million hect-ares per year from 2005 to 2010, mainly in tropical regions (FAO,2010). A large share of these plantations consists of exotic speciesplanted on highly weathered soils and managed in short rotations(FAO, 2006). The sustainability of such short-rotation plantations isof concern, since large amounts of nutrients contained in the bolesare frequently exported off site at harvesting (Corbeels et al., 2005;Nambiar, 2008). Soil organic matter (SOM) remains a majorcomponent of fertility as a support for cations and as a source of

ll rights reserved.

ine, UMR1137, Ecologie et4500 Vandoeuvre-les-Nancy,240.pron).

nutrients through mineralization (Feller and Beare, 1997; Lal,2004). In plantation forests, the decomposition of organic residues(OR) left on site at the harvesting release large amounts of nutri-ents, which is likely to increase soil carbon pools through directincorporation of C into the upper soil horizons (Mendham et al.,2002; Paul et al., 2003). Appropriate OR management is thereforecrucial for the sustainability of these plantations (Jones et al.,1999; Laclau et al., 2010a, 2010b; Powers et al., 2005).

OR manipulation has often been shown to have a large impacton soil CO2 efflux through changes in soil water content, soil tem-perature and substrate availability (Sayer, 2006). It is expected togreatly influence soil CO2 efflux in tropical climates because of highdecomposition rates (Aerts, 1997; Couteaux et al., 1995). The effectof OR management on SOM contents is more controversial(Johnson and Curtis, 2001). Whilst OR manipulations had no effecton SOM stocks in Eucalyptus plantations (Carneiro et al., 2009;

80 A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88

Mendham et al., 2003) or in different forests for 26 American studysites (Powers et al., 2005), OR removal led to a reduction in SOMstocks in Pinus radiata plantations (Huang et al., 2011), and ORaddition enhanced SOM stocks in a pine plantation of subtropicalAustralia (Chen and Xu, 2005) or in Eucalyptus globulus plantationsin southwestern Australia (Mathers et al., 2003). Such contrastingresults may be attributed to the quantity and the quality of organicinputs from different forests, and the great influence of soil typesand environmental conditions. Some in situ studies have shownthat increasing the amount of OR at the surface of forest soilsincreases soil CO2 effluxes from the mineral soil (ChemidlinPrévost-Bouré et al., 2010; Crow et al., 2009; Sulzman et al.,2005), and laboratory studies have provided evidence that freshorganic carbon inputs may increase soil C mineralization via‘‘priming effect’’ mechanisms (Fontaine et al., 2007; Kuzyakovet al., 2000; Nottingham et al., 2009). Priming was also demon-strated in tropical forests where aboveground litter amounts wereexperimentally increased (Sayer et al., 2011).

The effect of OR management practices on tree growth andnutrient cycling was studied over a full rotation of fast-growingplantations in a common experimental design replicated in eighttropical countries (Nambiar, 2008). The highest tree response tomanipulation was observed in Eucalyptus plantations establishedon Ferralic Arenosol soils in the Congo (Saint-André et al., 2004).The mass of organic residue deposited on the soil surface at har-vesting was an excellent predictor of the yield of the next rotation,suggesting that tree growth was highly dependent on the release ofnutrients contained in the organic residue (Laclau et al., 2010a,2010b).

We thus aimed to estimate the influence of OR managementpractices on the main components of soil CO2 effluxes of a Eucalyp-tus plantation in the Congo. We put forward the hypothesis that (i)contrasting amounts of organic residue deposited on the soil sur-face at harvesting largely influence soil CO2 effluxes, (ii) organicresidue manipulation significantly modifies SOM mineralization,and (iii) tree response to organic residue management is positivelycorrelated to heterotrophic soil CO2 effluxes that will reflect nutri-ent release from organic residues.

2. Material and methods

2.1. Site description

The study site of Kondi is located on the coastal plains of theCongo (4�340S, 11�540E, 100 m above sea level). The climate issub-equatorial with a rainy season from October to May and adry season from June to September. Mean annual rainfall is about1350 mm with high inter-annual variations, and the mean annualtemperature is 25 �C with limited seasonal variations of around5 �C. Eucalyptus plantations in this region have been establishedon Ferralic Arenosols (FAO classification), characterized by a uni-form sandy texture down to more than 10 m, a moderately acidicsoil pH, and very low amounts of exchangeable base cations andorganic matter. The soil mineralogy is dominated by quartz andto a lesser extent by kaolinite. Nutrient-bearing minerals are veryscarce. A thorough description of the soil at the study site can befound in Mareschal et al. (2011).

The native herbaceous savannah was destroyed by glyphosateapplications and afforested in 1992 with a natural Eucalyptushybrid (PF1-41) planted at a stocking density of 532 trees per ha(Laclau et al., 2003). This stand was harvested in 2001, the stumpswere killed by glyphosate application, and the plot was replantedwith clone 18–52 of the hybrid Eucalyptus urophylla (STBlake) � Eucalyptus grandis (Hill ex Maid.) at a stocking density of800 trees per ha (3.3 m � 3.7 m spacing). That stand was harvested

in March 2009 and the same clone (18–52) was replanted in June2009 at the same stocking density, after glyphosate applicationto kill the stumps and the understory.

2.2. Experimental design

A complete randomized design was set up in April 2009 withthree blocks of three treatments (nine plots):

R (removed): all aboveground organic residues were removedfrom the plot (litter from the previous rotation as well as harvestslash). The boundary between the ‘‘organic’’ and ‘‘mineral’’ soil lay-ers is easily recognizable at this site where soils are characterizedby a mull humus, i.e. with a rapid rate of litter decomposition andthe absence of a true humified soil layer (Oa horizon). Moreover, adense root mat is located just between the mineral soil and thefragmented litter (Oe horizon), thus clearly materializing thisboundary (Laclau et al., 2004).

SWH (stem wood harvest): only debarked commercial-sizedboles (top-end over-bark diameter exceeding 2 cm) were removed.This treatment left on the soil surface 2.1 kg m�2 of litter from theprevious rotation and 2.1 kg m�2 of harvest slash. Organic residueswere uniformly distributed on the ground.

DS (double slash): all the trees were logged as in the SWH treat-ment and the slash of the R plots was added. This treatment left onthe soil surface 2.1 kg m�2 of litter from the previous rotation and4.3 kg m�2 of slash.

Each plot covered an area of 1250 m2 (100 trees) with an innerplot of 450 m2 (36 trees) and two buffer rows. In each inner plot,nine PVC collars were installed at different distances from ninetrees to sample spatial variability related to the planting scheme(Epron et al., 2012). The collars were located in the interrow at adistance of 0.42–2.09 m from the nearest tree. We used PVC collarsmeasuring 20 cm in diameter and 11.5 cm in height (6581-044, Li-Cor Inc., Lincoln, NE, USA). Soil collars were inserted down to adepth of 5 cm in the soil 15 days before the first measurements.This depth ensured collar stability and was considered sufficientto minimize potential underestimations of soil respiration due tolateral diffusion of CO2 (Hutchinson and Livingston, 2001). Similaramounts of slash were added in all the collars of the same treat-ment, relatively to the values given above for each treatment.

2.3. Soil CO2 efflux and soil water content

Soil CO2 efflux (RS) was measured every 14 days for each collarfrom January 2009 (3 months before harvesting the previousstand) to June 2011 (2 years after planting). In all, 81 measure-ments were taken on each date (nine collars � three treat-ments � three blocks). We used a 20 cm respiration chamber(Li8100-103, LiCor Inc., Lincoln, NE, USA), in which the increasein CO2 concentration was recorded at 1 s intervals for 2 min usingan infrared gas analyser (Li8100). Volumetric soil water contentwithin the 0–6 cm soil layer (WS) was measured simultaneouslywith soil CO2 effluxes at about 10 cm from the collars, using a The-ta Probe (Type ML2X, Delta-T Devices Ltd., Cambridge, UK). Soiltemperature (TS) was also measured with a copper/constantanthermocouple penetration probe (Li8100-102 TC, LiCor Inc.) in-serted in the soil to a depth of 10 cm. Cumulative soil CO2 effluxwas estimated for each treatment using linear interpolations ofRS between measurement dates.

2.4. Above- and belowground organic residue decomposition rates

Belowground biomass of the previous rotation was sampled inMarch 2009, on 12 trees covering the distribution of cross-sectionareas of trunks at 1.3 m height, just before harvesting the stand.Fine root biomass (diameter < 2 mm) was quantified using a root

A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88 81

auger (8 cm inner diameter) down to a depth of 3 m, at 13 posi-tions close to each sampled tree (12 � 13 samples). Medium-sized(2–10 mm in diameter) and coarse roots (diameter > 10 mm) wereexcavated in six trenches measuring 3.3 m � 3.7 m down to adepth of 2 m (Levillain et al., 2011; Saint-André et al., 2005).Decomposition rates for fine, medium-sized and coarse roots mea-sured in the same stand from April 2001 to April 2003 (Kazotti andDeleporte, unpublished data) were used to assess the contributionof root decomposition to total soil CO2 effluxes using an exponen-tial decay model. Meteorological conditions were similar for thetwo studies (cumulative rainfall over 2 years of 2350 mm in theirstudy versus 2550 mm in our study).

The decomposition rate for the aboveground litter from the pre-vious rotation was studied in a 175 m2 area in the same stand,where slash had not been added and where new litterfall was reg-ularly removed. The previous rotation litter was collected in 10randomly-chosen quadrats with a 0.5 m2 metallic square framejust before harvesting (March 2009) and at the end of the studyperiod (June 2011). Living roots were removed carefully by handfrom the litter layer. Samples were oven-dried at 65 �C to constantweight and weighed. Ash contents were determined after combus-tion of subsamples in a muffle furnace at 450 �C for 6 h, and used tocorrect the dry mass for mineral soil contamination.

The decomposition rate for slash was monitored over the first2 years in the DS treatment of block 1 with 9 bags (internal dimen-sions 22 � 30 cm by 5 cm in height), using a 1 mm nylon meshcontaining 285 g of dry matter. The mesh size was chosen to pre-vent the loss of matter from the bags during fragmentation. Theamount of slash in the bags was similar to that deposited on thesoil surface in the DS treatment and in the PVC collars (4.3 kg m�2).Small slits (length about 1.5 cm) were made at the top of the bagsin order to avoid macrofauna exclusion resulting from the meshsize. The nine bags were recovered at the end of the study period,2 years after planting. Sample preparation was similar to that indi-cated above for the litter of the previous rotation.

The contributions of each decomposing compartment (root, lit-ter and slash) to cumulative soil CO2 efflux at the end of the exper-iment were estimated by multiplying the initial mass of eachcomponent by their relative mass loss in the decomposition exper-imentations. The carbon contents of roots, litter and slash were notdetermined in this study and it was thus set at 0.5 kg C kg�1 of drymatter in agreement with many other studies (Giardina and Ryan,2002; Keith et al., 1997). According to the negative exponentialrelationship between initial C/N litter and microbial C use effi-ciency (MCUE) from 2600 litterbag samplings of 21 decompositiondatasets spanning arctic to tropical ecosystems (Manzoni et al.,2010), a single MCUE of 0.23 was assumed to estimate the C con-tribution of each decomposing compartment to soil CO2 efflux(mean C/N of 70). Briefly, the carbon contribution of each compart-ment was calculated as:

RX ¼ AX0 � ð1� RMRXÞ � ð1�MCUEÞ � 0:5 ð1Þ

where RX is the C contribution of compartment X to cumulative soilCO2 efflux (kg m�2), AX0 the initial amount of dry matter of the com-partment (kg m�2), RMRX the relative mass of the compartmentremaining at the end of the study period (kg kg�1).

Tree leaves began to fall in the experiment from 1 year afterplanting onwards. Litterfall was collected every 4 weeks in four lit-ter traps (0.72 � 0.72 m) in each plot, at various distances fromfour trees with the same basal area as the mean of the plot. Litter-bags (1 mm mesh size) containing 10 g of leaf litter from the cur-rent rotation were installed on the forest floor in each plot (eightper plot � three treatments � three blocks = 72 litter bags) 1 yearafter planting. The methodology described above for bags contain-ing slash was used for bags containing leaf litter. In each plot, twobags were sampled 3, 6, 9 and 12 months after installation (15, 18,

21 and 24 months after planting, respectively) in order to assessleaf litter decomposition rates and to estimate the contributionmade by the decomposition of leaf litter from the current rotationto the heterotrophic respiration (RH) in each treatment.

The contribution of root litter from the current rotation to soilCO2 effluxes was estimated considering a fine root turnover of1.8 y�1 (Jourdan et al., 2008). The contribution of medium-sizedroots was not considered, assuming a lower turnover for this com-partment (Joslin et al., 2006).

2.5. Above- and belowground biomass accumulation

Tree height (H) and diameter at breast height (DBH) were mea-sured six times from November 2009 to June 2011, and basal areas(m2 ha�1) were calculated on each date for each treatment. Below-ground biomass was measured 1 and 2 years after planting on ninetrees (1 tree of mean ‘‘DBH2 � H’’ per treatment and per block),using the method described above for root biomass estimations be-fore harvesting the previous rotation. At each age, fine root bio-mass was quantified down to a depth of 3 m (13 auger samplesat different distances from each selected tree) and medium-sizedroots were excavated in three trenches down to a depth of 3 min each treatment. Aboveground biomass was also measureddestructively 12 and 24 months after planting on 8 trees per treat-ment covering the range of basal areas at each age.

Allometric equations were fitted between biomass and treeparameters (H � DBH2) using the NLP procedure of SAS software(Saint-André et al., 2005). The objective was to find the best modelfor each compartment. For aboveground biomass of the first andthe second year, the equations were fitted by treatment (‘‘treat-ment model’’) or all together (‘‘global model’’) and the best modelwas chosen using the Akaike Information Criterion. For fine andmedium-sized root biomass, global models including the two datesand the three treatments were chosen. These allometric relation-ships exhibited good R2 values comprised between 0.77 and 0.96respectively for aboveground biomass at 1 and 2 years and for fineand medium-sized root biomass over the study period.

2.6. Autotrophic respiration

Autotrophic respiration (RA) of young Eucalyptus trees was esti-mated according to Nouvellon et al. (2008) using root biomass andspecific root respiration rates measured in situ between April andJune 2005 on excised fine roots and the intact medium-sized rootson trees of the same clone at the same study site (Marsden et al.,2008):

RA ¼ ½BfrRfr30 þ BmrRmr30�Q ððTS�30Þ=10Þ10 ð2Þ

where RA is the autotrophic respiration of the stand (lmol m�2 s�1),Bfr and Bmr, the fine and medium-sized root biomass (g m�2), Rfr30

and Rmr30, the mean specific root respiration rates normalized at areference temperature of 30 �C. Rfr30 and Rmr30 were respectively10.32 ± 2.54 nmol g�1 s�1 for fine roots (n = 95) and 4.24 ±1.97 nmol g�1 s�1 for medium-sized roots (n = 12; Marsden et al.,2008). Eq. (2) was applied on a daily time-step, using stand root bio-mass values (Bfr and Bmr) linearly interpolated between each standinventory. Daily root respiration was corrected for daily changesin soil temperature using a Q10 value of 2.2 reported for Eucalyptusroots (Thongo M’Bou et al., 2010). Daily soil temperature was com-puted using a linear regression between soil temperatures mea-sured bi-weekly and air temperatures recorded simultaneously ata nearby weather station. RA was then cumulated to estimate theannual contribution of the autotrophic component of soil CO2 efflux.

0

1

2

3

4

5

6RSWHDS

Dai

ly p

reci

pita

tion

(mm

day

-1)

Stan

d ba

sala

rea

(m2

ha-1

)

RSWHDS

210 330 450 570 690 810900

20

40

60

80

100

120

Day from March 15, 2009

RSWHDS

RSWHDS

Fig. 1. Time-course of stand basal area for the three treatments and dailyprecipitation over the first 2 years after planting (June 15, 2009) in treatments R(all residues removed), SWH (stemwood harvest) and DS (double slash). Theprevious stand was harvested on March 15, 2009 and planted again 90 days after.The dotted black line indicates the date of the first year of the plantation. Standarderrors among blocks (n = 3) are indicated by vertical bars.

82 A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88

2.7. Impact of organic residue management on SOM decomposition

The heterotrophic component of soil CO2 efflux (RH) was calcu-lated deducting RA from RS in each treatment:

RH ¼ RS � RA ð3Þ

Soil CO2 effluxes resulting from the mineralization of SOM(RSOM) were calculated by subtracting the contribution of eachcompartment to the heterotrophic component of soil CO2 effluxfor each treatment:

RSOM ¼ RH � RRp � RLp � Rsl � RRc � RLc ð4Þ

where RSOM is the cumulative CO2 efflux from the mineralization ofSOM for each treatment, RH is the cumulative heterotrophic soil CO2

efflux estimated from Eq. (3) for each treatment, RRp, RLp, Rsl, RRc andRLc are the cumulative CO2 efflux calculated from Eq. (1) resultingfrom the decomposition of roots, litter and slash from the previousrotation, and roots and leaf litter from the current rotation. In the Rtreatment, RLp and Rsl were null. Rsl was double in DS in comparisonto the SWH treatment.

2.8. Statistical analysis

A one-way analysis of variance with repeated measures (RMANOVA) was used to check for differences in TS and WS amongtreatments. RM ANOVA were performed on five periods: beforeharvesting in order to characterize the initial variability of thesevariables and for the first and the second year after planting, bothfor the dry and wet seasons. The limits between dry and rainy sea-sons were graphically determined from consistent rainfall events.

Two-way RM ANOVAs were used to check the effect of treat-ments, blocks and interactions of these variables on RS. These anal-yses were carried out separately on each period defined for WS andTS. The normality of the residues and homoscedasticity werechecked prior to performing ANOVAs using Shapiro and Bartletttests, respectively. When significant differences among treatmentlevels were detected in a one-way RM ANOVA, the Tukey testwas used to compare treatment means. For the two-way RM ANO-VA, the Bonferroni multiple range test was used to compare treat-ment or block means.

Pearson correlation coefficients were computed between the TS

of each treatment and the air temperature measured at a nearby(200 m away) weather station and between soil CO2 efflux andthe amounts of organic matter on the forest floor at planting orduring tree growth. All the data were processed using the SAS9.2 software package (SAS Inc., Cary, NC, USA). The probability levelused to determine significance was P < 0.05.

3. Results

3.1. Tree growth

Tree growth exhibited high seasonal variability (Fig. 1). In thesecond year after planting, the basal area increment amounted to0.04–0.09 m2 ha�1 month�1 for the dry season and 0.33–0.51 m2 ha�1 month�1 for the wet season, depending on the treat-ment. The manipulation of organic residues significantly influ-enced tree growth over the first 2 years after planting (Fig. 1).Aboveground biomass was 38% and 33% lower in R than in SWHat age 1 and 2 years while DS values were only slightly but not sig-nificantly higher than those of SWH (Table 1). Belowground bio-mass was also significantly lower in R than in SWH and DS 1 and2 years after planting (Table 1).

3.2. Volumetric soil water content and temperature in the top soil

Over the six measurement dates before planting, WS rangedfrom 10.8% to 16.0% and differences among plots where the treat-ments were established were not significant (Table 2). WS in thetopsoil exhibited strong seasonal variations, irrespective of thetreatment (Fig. 2). Organic residue removal significantly reducedthe WS value by about 12% on average in the first year after plant-ing relative to the SWH treatment while slash addition had no ef-fect on WS (Table 2). In the second year after planting WS wassimilar in the three treatments during the dry season and signifi-cantly higher in the R treatment than in the SWH and the DS treat-ments during the wet season (Table 2).

Soil temperature (TS) at a depth of 10 cm was not significantlydifferent among treatments before and after planting (Table 2). TS

values showed similar seasonal trends to WS with higher valuesduring wet seasons than during dry seasons.

3.3. Total soil CO2 efflux

The differences in soil CO2 efflux among plots were low beforethe establishment of the treatments. RS was not significantly differ-ent between SWH and R plots, but was slightly lower in DS plotsrelative to SWH plots (Table 2).

During the dry seasons, RS was not significantly different be-tween the SWH and the DS treatments, while mean values wereabout 33% lower in the R treatment (Table 2). Sharp differencesin RS were observed among all the treatments during the wet sea-sons (Table 2). Seasonal variations in RS were substantial, rangingfrom 1.6, 2.3 and 2.5 lmol m�2 s�1 on average for the two dry sea-sons, to 2.9, 4.9 and 6.1 lmol m�2 s�1 for the two wet seasons inthe R, SWH and DS plots respectively. An increase in RS was ob-served from the first to the second year after planting but the mag-nitude was dependent on the seasons and treatments. RS increasedby 13%, 15% and 17% between the two dry seasons and by 21%, 17%and 13% between the two wet seasons, for the R, SWH and DStreatments, respectively (Table 2). RS was slightly lower in block1 than in block 2 and exhibited intermediate values in block 3,but differences among blocks were only significant in the first yearafter planting (Table 2).

3.4. Components of soil CO2 efflux

Autotrophic respiration (RA) was about threefold higher in thesecond year after planting than in the first year, regardless of treat-ment (Table 3). On average over the 2 years, RA released0.33 ± 0.04, 0.42 ± 0.07 and 0.43 ± 0.01 kg m�2 yr�1 of C in the R,

Table 1Aboveground and belowground biomass (kg m�2) in each treatment 1 and 2 years after planting. Different letters in the same line indicate significant differences amongtreatments (P < 0.05). The uncertainties are standard errors among blocks (n = 3).

First year Second year

R SWH DS R SWH DS

Aboveground biomass 0.31 ± 0.02a 0.50 ± 0.7b 0.55 ± 0.03b 1.12 ± 0.09a 1.45 ± 0.08b 1.65 ± 0.06bFine roots (0–2 mm) 0.09 ± 0.00a 0.10 ± 0.00b 0.10 ± 0.00b 0.16 ± 0.01a 0.18 ± 0.00b 0.19 ± 0.00bMedium-sized roots (2–10 mm) 0.03 ± 0.00a 0.07 ± 0.01b 0.07 ± 0.00b 0.12 ± 0.00a 0.13 ± 0.00b 0.14 ± 0.00bBelowground biomass 0.12 ± 0.01a 0.17 ± 0.02b 0.17 ± 0.01b 0.28 ± 0.02a 0.31 ± 0.02b 0.33 ± 0.02b

Table 2Mean soil CO2 efflux (RS) values for each block and treatment, volumetric soil water content (WS) and soil temperature (TS) for each treatment before harvesting, and for the firstand the second years after planting (for dry and wet seasons). Different letters in the same column indicate significant differences among treatments (P < 0.05). Treatments R (allresidues removed), SWH (stemwood harvest) and DS (double slash) were studied. The four measurement dates among treatment establishment and planting are included in thefirst year data set.

Before exploitation First year Second year

Dry season Wet season Dry season Wet season

RS (lmol m�2 s�1)Block 1 5.6a 1.7a 3.8a 2.2a 4.9aBlock 2 5.8a 2.2b 4.5b 2.4a 5.2aBlock 3 5.9a 2.0ab 4.2b 2.3a 5.0a

R 5.8ab 1.5a 2.4a 1.7a 3.4aSWH 6.0b 2.1b 4.3b 2.5b 5.4bDS 5.5a 2.2b 5.8c 2.7b 6.4c

WS (%)R 12.9a 6.3a 10.3a 6.5a 11.1bSWH 13.0a 7.6b 11.0b 6.4a 10.0aDS 12.7a 7.7b 11.0b 6.5a 10.2a

TS (�C)R 29.0a 26.9a 28.9a 26.5a 28.8aSWH 29.2a 27.3a 29.5a 26.6a 29.0aDS 28.8a 27.3a 29.4a 26.5a 28.8a

0 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840-30-60-90

RSWHDS

A

Year 1 Year 2

Day from March 15, 2009

WS

(%)

0

20

40

60

80

100

120

Dai

ly p

reci

pita

tion

(mm

day

-1)So

il C

O2

effl

ux (

µmol

m-2

s-1)

B

RSWHDS

0 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660 690 720 750 780 810 840-30-60-90

RSWHDS

A

Year 1 Year 2

0

20

40

60

80

100

120

Dai

ly p

reci

pita

tion

(mm

day

-1)

02468

1012141618

8

76

5

4

0B

32

1

RSWHDS

Fig. 2. Time-courses of (A) total soil CO2 efflux, (B) daily precipitation and mean volumetric soil water content (WS) in the top 6 cm of soil, from harvesting on March 15, 2009.Treatments R (all residues removed), SWH (stemwood harvest) and DS (double slash) were studied. Dotted and full black lines indicate clearcutting and replanting dates,respectively. Years and seasons are shown at the top of the figures and red lines indicate seasonal limits (SDS and SWS for short dry and short wet seasons, respectively).Standard errors among the collars of the three blocks are indicated on each date by vertical bars (n = 27).

A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88 83

SWH and DS treatments, respectively. Over the first 2 years afterplanting, RA contributed to 36%, 28% and 24% of total soil CO2 ef-fluxes in the R, SWH and DS treatments, respectively (Fig. 3).

The amounts of C released by the decomposition of above-ground organic matter in the first 2 years were greatly influencedby the amount of slash left on the soil surface at harvesting

Table 3Cumulative soil carbon fluxes (kg m�2) for the first and the first 2 years after planting in treatments R (all residues removed), SWH (stem wood harvest) and DS (double slash).Mean values and standard errors are calculated from the three blocks.

First year First 2 years

R SWH DS R SWH DS

Autotrophic respiration RAa 0.17 ± 0.01 0.21 ± 0.02 0.21 ± 0.01 0.67 ± 0.04 0.84 ± 0.07 0.86 ± 0.01

Heterotrophic respiration RHb 0.63 ± 0.08 1.11 ± 0.05 1.44 ± 0.06 1.19 ± 0.12 2.14 ± 0.10 2.71 ± 0.06

Total soil CO2 effluxes RS 0.80 ± 0.09 1.32 ± 0.07 1.65 ± 0.06 1.86 ± 0.14 2.98 ± 0.14 3.57 ± 0.06

a Calculated from Eq. (2).b Calculated from Eq. (3).

0.480.51

0.951.45

0.67 a 0.84 b 0.86 b

0.68 a 0.68 a 0.75 a

0.51

0

1

2

3

4

R SWH DS

Treatments

Autotrophic respiration SOM

Bellowground OM Aboveground OM

Cum

ulat

ive

soil

CO

2 ef

flux

(kg

m-2

of

C)

Fig. 3. Cumulative contributions of aboveground OM (previous and current rotationlitter + slash), belowground OM (decomposition of roots from the previous rotationand from the current rotation), soil organic matter and autotrophic respiration tototal soil CO2 efflux over the first 2 years after planting. Different letters in the samecompartment indicate significant differences among treatments (P < 0.05).

84 A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88

(Table 4). Over the study period, the decomposition of roots fromthe previous rotation released 0.15 ± 0.07, 0.13 ± 0.01 and0.09 ± 0.00 kg m�2 of C for fine, medium-sized and coarse roots,respectively. The contributions of fine root litter from the currentrotation were 0.12 ± 0.01, 0.14 ± 0.01 and 0.15 ± 0.01 kg m�2 of Cfor the R, SWH and DS treatments, respectively. All together, thedecomposition of roots amounted to 26%, 17% and 14% of total soilCO2 effluxes in the R, SWH and DS treatments, respectively (Fig. 3).The aboveground litter of the previous rotation had lost 51 ± 12% of

Table 4Dry matter, relative mass remaining (RMR) and estimated carbon losses as CO2 at the end ofand belowground 2 years after planting. Uncertainties were calculated from replicationsblock variability for leaf litter and fine root turnover.

Dry matter (kg m�2) RMR (kg kg�1)

R SWH DS R

Slash 2.15 4.30Previous rot. litter 2.14Current rotation leaf litter 0.24 ± 0.04

Total aboveground 4.29 6.44Coarse roots (>10 mm) 0.71 0.46 ± 0.20Medium roots (2–10 mm) 0.38 0.14 ± 0.06Fine roots (<2 mm) 0.25 0.05 ± 0.03Current rotation fine root litter

Total belowground 1.33

a Calculated from Eq. (1).

its initial dry mass 2 years after planting and contributed 14% and12% to total soil CO2 efflux in the SWH and DS treatments, respec-tively (Table 4). Slash had lost 59 ± 7% of its initial dry mass 2 yearsafter treatment establishment and its contribution to total soil CO2

effluxes was 16% and 27% in SWH and DS treatments, respectively.The decomposition rates for the leaf litter from the current rotationwere not significantly different among the three treatments (datanot shown). Cumulative litterfall during the current rotationamounted to 0.11, 0.13 and 0.13 kg m�2 of C in the R, SWH andDS treatments, respectively. The amount of C released by thedecomposition of that litter over the study period was estimatedat 0.04, 0.05 and 0.05 kg m�2 of C, in the R, SWH and DS treatments,respectively. Together, the decomposition of all aboveground litterand slash contributed to 2%, 32% and 41% of total soil CO2 effluxesin R, SWH and DS, respectively, over the first 2 years after planting(Fig. 3). SOM mineralization (RSOM) calculated from Eq. (4) over thesame period released 0.30 ± 0.09, 0.30 ± 0.08 and0.33 ± 0.05 kg m�2 y�1 of C in the R, SWH and DS treatmentsrespectively. The differences among treatments were not signifi-cant. RSOM contributed to 36%, 23% and 21% of total soil CO2 effluxesin R, SWH and DS, respectively (Fig. 3).

3.5. Relationship between soil CO2 effluxes and tree growth

The amount of organic residue left on the mineral soil at plant-ing (aboveground litter + slash from the previous rotation) and RS

or RH showed strong positive linear relationships (R > 0.95,P < 0.001, n = 9, Fig. 4).

Over the two first years after the treatments were established,RS was also highly correlated to the aboveground biomass for thenine plots of the experiment (R = 0.97, P < 0.001, n = 9). Annualaboveground biomass increments were also highly correlated toRH in the nine plots (Fig. 5). The linear regression slopes betweencumulative annual RH (dependent variable) and annual above-ground biomass increments (independent variable) remained

the study period for slash, previous rotation litter, current rotation litter abovegroundin each decomposition experiment for slash, litter and root decomposition and from

Carbon loss (kg m�2)

SWH DS R SWH DS

0.41 ± 0.07 0.49 ± 0.03a 0.98 ± 0.07a

0.49 ± 0.12 0.42 ± 0.05a

0.26 ± 0.02 0.29 ± 0.01 0.04 ± 0.01 0.04 ± 0.01 0.05 ± 0.00

0.04 ± 0.01 0.95 ± 0.09 1.45 ± 0.120.15 ± 0.07a

0.13 ± 0.01a

0.09 ± 0.00a

0.12 ± 0.01 0.14 ± 0.01 0.15 ± 0.01

0.48 ± 0.09 0.51 ± 0.09 0.51 ± 0.09

2 yearsy = 0.27 x + 1.85

R2 = 1.00

1.5 yearsy = 0.20 x + 1.19

R2 = 1.00

1 yeary = 0.13 x + 0.79

R2 = 1.00

0.5 yeary = 0.08 x + 0.34

R2 = 0.990 2 4 6 8

Organic residues at planting (kg m-2)

2 yearsy = 0.27 x + 1.85

R2 = 1.00

1.5 yearsy = 0.20 x + 1.19

R2 = 1.00

1 yeary = 0.13 x + 0.79

R2 = 1.00

0.5 yeary = 0.08 x + 0.34

R2 = 0.990

1

2

3

4C

umul

ativ

e so

ilC

O2

effl

ux (

kg m

-2of

C)

Fig. 4. Relationships among amounts of aboveground organic residues (previousrotation litter + slash) at planting and cumulative soil CO2 efflux (RS) at 0.5, 1, 1.5and 2 years in treatments R (circles), SWH (triangles) and DS (squares).

2nd yeary = 0.39 x + 0.58

R2 = 0.69

1st yeary = 0.31 x + 0.12

R2 = 0.820.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0.0 0.5 1.0 1.5 2.0

R H (kg m-2 year-1 of C)

Abo

vegr

. bio

mas

s (k

g m

-2 y

ear-1

)

Fig. 5. Relationships between annual aboveground biomass increment in each plotand annual cumulative heterotrophic CO2 effluxes (RH) estimated by our partition-ing method at 1 year (white symbols) and 2 years (black symbols) after planting intreatments R (circles), SWH (triangles) and DS (squares).

A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88 85

similar at 1 and 2 years of age. Nevertheless, the annual biomassincrement was about twice as high in the second year as in the firstyear after planting despite very similar annual heterotrophic CO2

effluxes between years. The amount of C released from the mineralsoil (RSOM) over the first 2 years after harvesting in the nine plotswas not significantly correlated with the aboveground biomass(R = 0.38, P = 0.32, n = 9, data not shown).

4. Discussion

4.1. Effects of organic residue manipulation on soil CO2 effluxes

Annual soil CO2 effluxes (0.82, 1.32 and 1.59 kg m�2 yr�1 of C inthe R, SWH and DS treatments, respectively) are within the rangeof values for tropical moist forests (0.89–1.45 kg m�2 yr�1 of C,Raich and Schlesinger, 1992). Annual soil CO2 effluxes for the com-mercial SWH treatment were similar to those of a 3-year-old Euca-lyptus stand in its third rotation (1.25–1.30 kg m�2 yr�1 of C, Epronet al., 2004) and tallied with several Brazilian studies in Eucalyptusplantations (1.05–1.61 kg m�2 yr�1 of C, Campoe et al., 2012; 1.16–1.38 kg m�2 yr�1 of C, Epron et al., 2012; 1.13 kg m�2 yr�1 of C,Nouvellon et al., 2012; 0.67–1.34 kg m�2 yr�1 of C, Ryan et al.,2010). Annual soil CO2 efflux in the R treatment was close to values

reported for first afforestation with Eucalyptus where slash and lit-ter from a previous rotation do not exist (0.66–0.71 kg m�2 yr�1 ofC, Keith et al., 1997; 0.66 kg m�2 yr�1 of C, Nouvellon et al., 2008;0.58 kg m�2 yr�1 of C, Nouvellon et al., in press). The manipulationof organic residue had a strong effect on annual soil CO2 effluxes,since the values were twofold higher in the DS treatment than inthe R treatment. Forest floor removal reduced soil CO2 effluxesby 38% in our study, which was more than the 20% decrease inan old-growth moist lowland tropical forest of Panama (Sayeret al., 2007) and the 28% drop recorded in an Amazonian regrowthforest (Vasconcelos et al., 2004).

Soil CO2 effluxes were strongly affected by seasons and the ef-fect of organic residue manipulation on RS was mostly observedduring the wet seasons as a result of WS control over soil CO2 effluxin these tropical plantations (Epron et al., 2004; Marsden et al.,2008; Nouvellon et al., in press). During the first dry season, WS

was lower in the topsoil of the R treatment where litter removalprobably increased soil desiccation due to evaporation (Sayer,2006). These differences in WS among treatments may also accountfor the lower soil CO2 efflux observed during this period in the Rtreatment than in the other treatments. However, differences inWS among treatments disappeared after several months whenthe interception of solar radiation by the tree crowns reducedevaporation at the soil surface and tree growth increased rootwater uptake. A low buffering effect of litter still remained withfaster rises in WS after rainfall events and faster decreases duringdry periods in the R treatment than in the other treatments. Addi-tional input of slash did not influence WS in the DS treatment com-pared to SWH treatment, as observed in a temperate deciduousforest (Chemidlin Prévost-Bouré et al., 2010) or in an old-growthmoist lowland tropical forest of Panama (Sayer et al., 2007). TheRS differences among treatments probably mainly resulted fromthe change in substrate availability since TS and WS were not signif-icantly modified among treatments over most of the study period.

4.2. Effect of OR manipulation on SOM mineralization through soil CO2

efflux partitioning

Autotrophic respiration (RA) was estimated in order to deduceheterotrophic respiration from total soil CO2 effluxes. Our estima-tion of RA was derived from specific root respiration rates mea-sured at the same site (Marsden et al., 2008), and up-scaled tothe plot from root biomass data. The uncertainty involved in thistechnique is mainly due to the temperature sensitivity of root res-piration and disturbance of the root soil interface, especially for thefine root biomass, which contributed up to 80% to RA in our study.The seasonal variations in soil temperature did not exceed 6 �C at adepth of 10 cm and RA was corrected using soil temperatureextrapolated from air temperature. Root biomass estimations (fi-ne + medium-sized roots: 0–10 mm in diameter) fell within the95% confidence interval of predicted root biomass values frominventories using the allometric equations calibrated by Saint-And-ré et al. (2005) for a slightly less productive clone (1–41) in thesame area on 18 trees aged from 11 to 135 months.

Quantifications of RA in young tropical plantations are scarcebut Nouvellon et al. (2008) estimated very similar values the firstyear after planting in a Eucalyptus stand on a native savannah inthe Congo (0.16–0.20 kg m�2 yr�1 versus 0.17–0.21 kg m�2 yr�1 ofC for the first year in our study). Over the first 2 years after plant-ing, our results (0.33–0.43 kg m�2 yr�1 of C) were also consistentwith the values estimated for the youngest Eucalyptus stands of achronosequence in the Congo (0.29 kg m�2 yr�1 of C; Nouvellonet al., in press). However, at these early growth stages, and becauseof the large contribution made by the forest floor, the RA to RS ratioremained low for the SWH and DS treatments (28% and 24% overthe first 2 years after planting, respectively) compared to the R

86 A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88

treatment (36%). The latter value is in agreement with the contri-bution of RA to RS that was estimated at 31% in an adjacent 6-month-old plantation (Epron et al., 2009), at 39% in another 1-year-old plantation (Nouvellon et al., 2008), at 48% in a nearby 3-year-old Eucalyptus stand (Marsden et al., 2008) and at 59% in a9-year-old stand (Epron et al., 2006).

The partitioning of soil CO2 effluxes in our study showed thatforest floor decomposition (litter + slash) was the main contributorto total soil CO2 effluxes in the SWH and DS treatments (32% and41%, respectively). These values are comparable to those found ina subtropical native forest and fir plantations in China (27–29%;Yang et al., 2007), a subtropical coniferous plantation in southernChina (33%; Wang et al., 2009), a tropical cloud forest of the Peru-vian Andes (37%; Zimmermann et al., 2009) and an Amazonianrainforest (37%; Chambers et al., 2004).

Organic residue management did not significantly influence themineralization of SOM in our study (0.30 ± 0.09, 0.30 ± 0.08 and0.33 ± 0.05 kg m�2 yr�1 of C for the R, SWH and DS treatments,respectively) although many studies manipulating litter, mainlyin temperate forests, have reported 10–30% increases in SOM min-eralization (Chemidlin Prévost-Bouré et al., 2010; Crow et al.,2009; Sayer et al., 2011; Subke et al., 2006; Sulzman et al., 2005).The limited impact of management practices on the decompositionof SOM tallied with a lack of soil tillage effect on soil CO2 effluxesobserved in a nearby Congolese sandy soil (Nouvellon et al., 2008).

Nutrient-poor soils have nevertheless been considered moresensitive to priming effects than nutrient-rich soils (Fontaineet al., 2003; Hamer and Marschner, 2005; Kuzyakov et al., 2000).Isotopic studies close to the study site showed that SOM wasmainly derived from an ancient forest (8000 BP) and was com-posed of old and stable organic matter with a mean residence timeof several millennia (Schwartz et al., 1992; Schwartz and Namri,2002). However, only 5% of this SOM at our study site was chemi-cally recalcitrant (Poirier et al., 2002) and the stability of subsoilOM certainly resulted from the unavailability of substrates formicrobial activity (Hamer and Marschner, 2005). Carbon inputsalong the soil profile might thus stimulate the mineralization ofancient buried carbon (Fontaine et al., 2007). Our result howeversuggested that the management of organic matter have limited im-pact on SOM mineralization but the absence of priming cannotfully be claimed since our calculation of SOM mineralization wasprone to uncertainties resulting from the partitioning approachand from the number of blocks. The poor quality of Eucalyptus littermight be involved in the lack of any significant priming effect, eventhought it has been shown that additions of recalcitrant C can alsoresult in priming, especially when SOM is already recalcitrant (Ras-mussen et al., 2008). More accurate quantification of SOM decom-position in laboratory experiments involving isotopic tools todiscriminate SOM and added substrate would be necessary tostudy SOM priming effects in our nutrient-poor tropical soil.

One major source of uncertainty in our computations of respira-tion related to the decomposition of organic residue came from theassumption that 23% of carbon was not lost as CO2 during thedecomposition process (i.e. MCUE). The chosen value was in therange estimated by Ågren et al. (2001) for various forest ecosystems(0.18–0.26) or Ngao et al. (2005) for leaf decomposition in a temper-ate beech stand (0.20). The MCUE includes microbial assimilation,DOC and particulate OM transferred from the floor organic layerto the mineral soil. On this biannual time scale, some of the C trans-ferred to soil might however be lost as CO2 and contribute to RSOM.

4.3. Consequences of OR management for the sustainability ofEucalyptus plantations

Organic residue management is likely to play a major role inplantation productivity as shown in this study by the differences

in tree growth among treatments. The large differences observedamong treatments are unlikely to have resulted from a largeenhancement of N availability due to a priming effect on SOM, orto changes in the soil microclimate, as discussed above. Treegrowth was largely related to the heterotrophic efflux at 1 and2 years after planting, which suggested a strong dependence onthe amounts of nutrients released through the decomposition oforganic residues. Our study, based on C fluxes, reinforces the re-sults of Laclau et al. (2010a) who found a concordance betweennutrient stocks contained in organic residues at planting and instand biomass at the end of the rotation. The higher biomass incre-ment in the second year after planting than in the first year, despitesimilar amounts of OM being mineralized, might be explained bythe improvement of nutrient foraging with larger root biomassthe second year, the rise in production of nutrient-poor tissues(in particular wood) in the second year after planting (Laclauet al., 2000) and large retranslocations of nutrients taken up fromthe soil in the first year of growth (Laclau et al., 2001, 2010b).The biological cycle of nutrients is very conservative in CongoleseEucalyptus plantations (Laclau et al., 2003) and the amounts ofnutrients accumulated in tree biomass at early growth stages aredecisive throughout the whole rotation (Du Toit et al., 2010; Laclauet al., 2010a, 2010b; Mendham et al., 2003).

Soil organic matter is a major component of the fertility of trop-ical soils (Feller and Beare, 1997) and the sustainability of tropicalforest plantations is strongly dependent on SOM stocks (Zech et al.,1997). It has been shown in the Congo that afforestation with Euca-lyptus trees on native savannahs leads to a slight decrease in soilcarbon stocks 1 year after planting (Nouvellon et al., 2008) result-ing from the loss of labile savannah-derived carbon which is coun-terbalanced by accretion of Eucalyptus-derived soil carbonthroughout the first rotation (Epron et al., 2009). Removal of floororganic matter led to C depletion in the mineral soil estimated atabout 0.52 kg m�2 yr�1 of C over the 2 years after planting, and ac-counted for about 19% of the total soil C stock within the 0–1 m soillayer (soil C stocks were measured at the same site by Mareschalet al., 2011). In the SWH and DS treatments, C depletions in themineral soil were lower (about 0.24 and 0.16 kg m�2 yr�1 of C,respectively) as a result of C incorporation from organic residues(based on the hypothesis of a MCUE of 0.23). However, the totalsoil C balance remained positive for the SWH and DS treatmentsover the first 2 years after planting with additional C amountingto 0.69 and 1.19 kg m�2 of C, respectively, as a consequence oflarge amounts of C in litter and slash (above- and belowground)from the previous rotation still remaining 2 years after planting.All the organic residues will be decomposed in the third year afterplanting and about 0.27 kg m�2 yr�1 of C for SWH and0.37 kg m�2 yr�1 of C for DS will thus be transferred into theSOM compartment, counterbalancing the reduction in soil C esti-mated 2 years after planting. These results are consistent with aprevious experiment in the Congo which showed that contrastingamounts of organic residues left on the soil surface at harvestingdo not modify C stocks in the mineral soil at the end of the follow-ing rotation (Laclau et al., 2010a).

5. Conclusions

The management of organic harvest residues plays a crucial roleto ensure the sustainability of these poor tropical soils. The short-term productivity of Eucalyptus stands with low fertilizer additionis largely dependent on the nutrients released throughout thedecomposition of harvest residues as shown by the strong relation-ship between cumulative heterotrophic CO2 efflux and tree growth.Furthermore, harvest residues addition has probably limitedinteraction with old SOM but contributes to maintain soil organic

A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88 87

matter stocks through successive rotations, ensuring the long-termproductivity of these plantations. Therefore, the retention of organ-ic residues at harvest should be optimized, wild fires must be pre-vented, slash burning must be prohibited and organic amendmentmight be recommended.

Acknowledgements

We acknowledge CRDPI, the Republic of Congo and EFC. Wethank J.C. Mazoumbou, T. Matsoumbou, S. Ngoyi and A. Diamessofor field measurements. We thank Peter Biggins for the revisionof the English.

References

Aerts, R., 1997. Climate, leaf litter chemistry and leaf litter decomposition interrestrial ecosystems: a triangular relationship. Oikos 79, 439–449.

Ågren, G., Bosatta, E., Magill, A., 2001. Combining theory and experiment tounderstand effects of inorganic nitrogen on litter decomposition. Oecologia 128,94–98.

Carneiro, M., Serrão, V., Fabião, A., Madeira, M., Balsemão, I., Hilário, L., 2009. Doesharvest residue management influence biomass and nutrient accumulation inunderstory vegetation of Eucalyptus globulus Labill. plantations in aMediterranean environment? Forest Ecol. Manage. 257, 527–535.

Chambers, J.Q., Tribuzy, E.S., Toledo, L.C., Crispim, B.F., Higuchi, N., Dos Santos, J.,Araujo, A., Kruijt, B., Nobre, A.D., Trumbore, S.E., 2004. Respiration from atropical forest ecosystem: partitioning of sources and low carbon efficiency.Ecol. Appl. 14, 72–88.

Chemidlin Prévost-Bouré, N., Soudani, K., Damesin, C., Berveiller, D., Lata, J.-C.,Dufrêne, E., 2010. Increase in aboveground fresh litter quantity over-stimulatessoil respiration in a temperate deciduous forest. Appl. Soil Ecol. 46, 26–34.

Chen, C.R., Xu, Z.H., 2005. Soil carbon and nitrogen pools and microbial properties ina 6-year-old slash pine plantation of subtropical Australia: impacts of harvestresidue management. Forest Ecol. Manage. 206, 237–247.

Campoe, O.C., Stape, J.L., Laclau, J.P., Marsden, C., Nouvellon, Y., 2012. Stand-levelpatterns of carbon fluxes and partitioning in a Eucalyptus grandis plantationacross a gradient of productivity, in São Paulo State, Brazil. Tree Physiol. 32,696–706.

Corbeels, M., McMurtrie, R.E., Pepper, D.A., Mendham, D.S., Grove, T.S., O’Connell,A.M., 2005. Long-term changes in productivity of eucalypt plantations underdifferent harvest residue and nitrogen management practices: a modellinganalysis. Forest Ecol. Manage. 217, 1–18.

Couteaux, M., Bottner, P., Berg, B., 1995. Litter decomposition, climate and litterquality. Trends Ecol. Evol. 10, 63–66.

Crow, S.E., Lajtha, K., Bowden, R.D., Yano, Y., Brant, J.B., Caldwell, B.A., Sulzman, E.W.,2009. Increased coniferous needle inputs accelerate decomposition of soilcarbon in an old-growth forest. Forest Ecol. Manage. 258, 2224–2232.

Du Toit, B., Smith, C.W., Little, K.M., Boreham, G., Pallett, R.N., 2010. Intensive, site-specific silviculture: manipulating resource availability at establishment forimproved stand productivity. A review of South African research. Forest Ecol.Manage. 259, 1836–1845.

Epron, D., Laclau, J.-P., Almeida, J.C.R., Gonçalves, J., Ponton, S., Sette, C.R., Delgado-Rojas, J.S., Bouillet, J.-P., Nouvellon, Y., 2012. Do changes in carbon allocationaccount for the growth response to potassium and sodium applications intropical Eucalyptus plantations? Tree Physiol. 31, 1–13.

Epron, D., Marsden, C., Thongo M’Bou, A., Saint-André, L., D’Annunzio, R., Nouvellon,Y., 2009. Soil carbon dynamics following afforestation of a tropical savannahwith Eucalyptus in Congo. Plant Soil 323, 309–322.

Epron, D., Nouvellon, Y., Deleporte, P., Ifo, S., Kazotti, G., Thongo M’Bou, A.,Mouvondy, W., Andre, L.S., Roupsard, O., Jourdan, C., Hamel, O., 2006. Soilcarbon balance in a clonal Eucalyptus plantation in Congo: effects of logging oncarbon inputs and soil CO2 efflux. Global Change Biol. 12, 1021–1031.

Epron, D., Nouvellon, Y., Roupsard, O., Mouvondy, W., Mabiala, A., Saint-André, L.,Joffre, R., Jourdan, C., Bonnefond, J.-M., Berbigier, P., Hamel, O., 2004. Spatial andtemporal variations of soil respiration in a Eucalyptus plantation in Congo.Forest Ecol. Manage. 202, 149–160.

FAO, 2006. Global Planted Forests Thematic Study. Results and Analysis. PlantedForests and Trees. Working Paper FP38E, Rome, pp. 178.

FAO, 2010. Global Forest Resources Assessment 2010. Main Report. FAO ForestryPaper 163, Rome, pp. 378.

Feller, C., Beare, M.H., 1997. Physical control of soil organic matter dynamics in thetropics. Geoderma 79, 69–116.

Fontaine, S., Barot, S., Barre, P., Bdioui, N., Mary, B., Rumpel, C., 2007. Stability oforganic carbon in deep soil layers controlled by fresh carbon supply. Nature450, 277–280.

Fontaine, S., Mariotti, A., Abbadie, L., 2003. The priming effect of organic matter: aquestion of microbial competition? Soil Biol. Biochem. 35, 837–843.

Giardina, C.P., Ryan, M.G., 2002. Total belowground carbon allocation in a fast-growing Eucalyptus plantation estimated using a carbon balance approach.Ecosystems 5, 487–499.

Hamer, U., Marschner, B., 2005. Priming effects in different soil types induced byfructose, alanine, oxalic acid and catechol additions. Soil Biol. Biochem. 37, 445–454.

Huang, Z., Clinton, P.W., Davis, M.R., 2011. Post-harvest residue management effectson recalcitrant carbon pools and plant biomarkers within the soil heavy fractionin Pinus radiata plantations. Soil Biol. Biochem. 43, 404–412.

Hutchinson, G.L., Livingston, G.P., 2001. Vents and seals in non-steady-statechambers used for measuring gas exchange between soil and the atmosphere.Eur. J. Soil Sci. 52, 675–682.

Johnson, D.W., Curtis, P.S., 2001. Effects of forest management on soil C and Nstorage: meta analysis. Forest Ecol. Manage. 140, 227–238.

Jones, H.E., Madeira, M., Herraez, L., Dighton, J., Fabiâo, A., González-Rio, F.,Fernandez Marcos, M., Gomez, C., Tomé, M., Feith, H., Magalhâes, M.C., Howson,G., 1999. The effect of organic matter management on the productivity ofEucalyptus globulus stands in Spain and Portugal: tree growth and harvestresidue decomposition in relation to site and treatment. Forest Ecol. Manage.122, 73–86.

Joslin, J.D., Gaudinski, J.B., Torn, M.S., Riley, W.J., Hanson, P.J., 2006. Fine-rootturnover patterns and their relationship to root diameter and soil depth in a14C-labeled hardwood forest. New Phytol. 172, 523–535.

Jourdan, C., Silva, E.V., Gonçalves, J.L.M., Ranger, J., Moreira, R.M., Laclau, J.P.,2008. Fine root production and turnover in Brazilian Eucalyptus plantationsunder contrasting nitrogen fertilization regimes. Forest Ecol. Manage. 256,396–404.

Keith, H., Raison, R.J., Jacobsen, K.L., 1997. Allocation of carbon in a mature eucalyptforest and some effects of soil phosphorus availability. Plant Soil 196, 81–99.

Kuzyakov, Y., Friedel, J.K., Stahr, K., 2000. Review of mechanisms and quantificationof priming effects. Soil Biol. Biochem. 32, 1485–1498.

Laclau, J.-P., Bouillet, J.-P., Ranger, J., 2000. Dynamics of biomass and nutrientaccumulation in a clonal plantation of Eucalyptus in Congo. Forest Ecol. Manage.128, 181–196.

Laclau, J.-P., Bouillet, J.-P., Ranger, J., Joffre, R., Gouma, R., Saya, A., 2001. Dynamics ofnutrient translocation in stemwood across an age series of a Eucalyptus hybrid.Ann. Bot. 88, 1079–1092.

Laclau, J.-P., Ranger, J., Bouillet, J.-P., de Dieu Nzila, J., Deleporte, P., 2003. Nutrientcycling in a clonal stand of Eucalyptus and an adjacent savanna ecosystem inCongo: 1. Chemical composition of rainfall, throughfall and stemflow solutions.Forest Ecol. Manage. 176, 105–119.

Laclau, J.P., Toutain, F., Thongo M’Bou, A., Arnaud, M., Joffre, R., Ranger, J., 2004. Thefunction of the superficial root mat in the biogeochemical cycles of nutrients incongolese eucalyptus plantations. Ann. Bot. 93, 249–261.

Laclau, J.-P., Levillain, J., Deleporte, P., Nzila, J.d.D., Bouillet, J.-P., Saint André, L.,Versini, A., Mareschal, L., Nouvellon, Y., Thongo M’Bou, A., Ranger, J., 2010a.Organic residue mass at planting is an excellent predictor of tree growth inEucalyptus plantations established on a sandy tropical soil. Forest Ecol. Manage.260, 2148–2159.

Laclau, J.-P., Ranger, J., De Moraes Gonçalves, J.L., Maquère, V., Krusche, A.V., ThongoM’Bou, A., Nouvellon, Y., Saint-André, L., Bouillet, J.-P., De Cassia Piccolo, M.,Deleporte, P., 2010b. Biogeochemical cycles of nutrients in tropical Eucalyptusplantations: main features shown by intensive monitoring in Congo and Brazil.Forest Ecol. Manage. 259, 1771–1785.

Lal, R., 2004. Soil carbon sequestration impacts on global Climate change and foodsecurity. Science 304, 1623–1627.

Levillain, J., Thongo M’Bou, A., Deleporte, P., Saint-André, L., Jourdan, C., 2011. Is thesimple auger coring method reliable for below-ground standing biomassestimation in Eucalyptus forest plantations? Ann. Bot. 108, 221–230.

Manzoni, S., Trofymow, J.A., Jackson, R.B., Porporato, A., 2010. Stoichiometriccontrols on carbon, nitrogen, and phosphorus dynamics in decomposing litter.Ecol. Monogr. 80, 89–106.

Mareschal, L., Nzila, J.D.D., Turpault, M.P., Thongo M’Bou, A., Mazoumbou, J.C.,Bouillet, J.P., Ranger, J., Laclau, J.P., 2011. Mineralogical and physico-chemicalproperties of Ferralic Arenosols derived from unconsolidated Plio-Pleistocenicdeposits in the coastal plains of Congo. Geoderma 162, 159–170.

Marsden, C., Nouvellon, Y., M’Bou, A.T., Saint-André, L., Jourdan, C., Kinana, A., Epron,D., 2008. Two independent estimations of stand-level root respiration on clonalEucalyptus stands in Congo: up scaling of direct measurements on roots versusthe trenched-plot technique. New Phytol. 177, 676–687.

Mathers, N.J., Mendham, D.S., O’Connell, A.M., Grove, T.S., Xu, Z., Saffigna, P.G., 2003.How does residue management impact soil organic matter composition andquality under Eucalyptus globulus plantations in southwestern Australia? ForestEcol. Manage. 179, 253–267.

Mendham, D.S., O’Connell, A.M., Grove, T.S., Rance, S.J., 2003. Residue managementeffects on soil carbon and nutrient contents and growth of second rotationeucalypts. Forest Ecol. Manage. 181, 357–372.

Mendham, D.S., Sankaran, K.V., O’Connell, A.M., Grove, T.S., 2002. Eucalyptusglobulus harvest residue management effects on soil carbon and microbialbiomass at 1 and 5 years after plantation establishment. Soil Biol. Biochem. 34,1903–1912.

Nambiar, E.K.S., 2008. Introduction: sustained productivity of plantation forests inthe tropics: a decade of research partnership. In: Nambiar, E.K.S. (Ed.), SiteManagement and Productivity in Tropical Plantation Forests. Proceedings ofWorkshops in Piracicaba (Brazil), 22–26 November 2004 and Bogor (Indonesia),6–9 November 2006. CIFOR, Bogor, Indonesia, pp. 1–3.

Ngao, J., Epron, D., Brechet, C., Granier, A., 2005. Estimating the contribution of leaflitter decomposition to soil CO2 efflux in a beech forest using 13C-depletedlitter. Global Change Biol. 11, 1768–1776.

88 A. Versini et al. / Forest Ecology and Management 301 (2013) 79–88

Nottingham, A., Turner, B., Chamberlain, P., Stott, A., Tanner, E., 2009. Priming andmicrobial nutrient limitation in lowland tropical forest soils of contrastingfertility. Biogeochemistry, 1–19.

Nouvellon, Y., Epron, D., Kinana, A., Hamel, O., Mabiala, A., D’Annunzio, R.,Deleporte, P., Saint-André, L., Marsden, C., Roupsard, O., Bouillet, J.-P., Laclau,J.-P., 2008. Soil CO2 effluxes, soil carbon balance, and early tree growthfollowing savannah afforestation in Congo: comparison of two site preparationtreatments. Forest Ecol. Manage. 255, 1926–1936.

Nouvellon, Y., Epron, D., Marsden, C., Kinana, A., Le Maire, G., Deleporte, P., SaintAndré, L., Bouillet, J.-P., Laclau, J.-P., in press. Age-related changes in litter inputsexplain annual trends in soil CO2 effluxes over a full Eucalyptus rotation afterafforestation of a tropical savannah. Biogeochemistry. http://dx.doi.org/10.1007/s10533-011-9685-9.

Nouvellon, Y., Laclau, J.P., Epron, D., Le Maire, G., Bonnefond, J.M., Gonçalves, J.L.M.,Bouillet, J.P., 2012. Carbon allocations in monospecific and mixed-speciesplantations of Eucalyptus grandis and Acacia mangium in South-Eastern Brazil.Tree Physiol. 32, 680–695.

Paul, K.I., Polglase, P.J., Richards, G.P., 2003. Predicted change in soil carbonfollowing afforestation or reforestation, and analysis of controlling factors bylinking a C accounting model (CAMFor) to models of forest growth (3PG), litterdecomposition (GENDEC) and soil C turnover (RothC). Forest Ecol. Manage. 177,485–501.

Poirier, N., Derenne, S., Balesdent, J., Rouzaud, J.-N., Mariotti, A., Largeau, C., 2002.Abundance and composition of the refractory organic fraction of an ancient,tropical soil (Pointe Noire, Congo). Org. Geochem. 33, 383–391.

Powers, R.F., Andrew Scott, D., Sanchez, F.G., Voldseth, R.A., Page-Dumroese, D.,Elioff, J.D., Stone, D.M., 2005. The North American long-term soil productivityexperiment: findings from the first decade of research. Forest Ecol. Manage.220, 31–50.

Raich, J.W., Schlesinger, W.H., 1992. The global carbon dioxide flux in soilrespiration and its relationship to vegetation and climate. Tellus B 44, 81–99.

Rasmussen, C., Southard, R.J., Horwath, W.R., 2008. Litter type and soil mineralscontrol temperate forest soil carbon response to climate change. Global ChangeBiol. 14, 2064–2080.

Ryan, M.G., Stape, J.L., Binkley, D., Fonseca, S., Loos, R.A., Takahashi, E.N., Silva, C.R.,Silva, S.R., Hakamada, R.E., Ferreira, J.M., Lima, A.M.N., Gava, J.L., Leite, F.P.,Andrade, H.B., Alves, J.M., Silva, G.G.C., 2010. Factors controlling Eucalyptusproductivity: how water availability and stand structure alter production andcarbon allocation. Forest Ecol. Manage. 259, 1695–1703.

Saint-André, L., Laclau, J.-P., Deleporte, P., Gava, P., Gonçalves, J., Nzila, J.d.D., Smith,C., Du Toit, B., Xu, D., Sankaran, K., Marien, J., Nouvellon, Y., Bouillet, J.-P.,Ranger, J., 2008. Slash and litter management effects on Eucalyptus productivity:a synthesis using a growth and yield modelling approach. In: Nambiar, E.K.S.,Ranger, J., Tiarks, A., Toma, T. (Eds.), Site Management and Productivity inTropical Plantation Forests. Proceedings of Workshop in Piracicaba (Brazil), 22-

26 November 2004 and Bogor (Indonesia), 6–9 November 2006. CIFOR, Bogor,Indonesia, pp. 173–190.

Saint-André, L., M’Bou, A.T., Mabiala, A., Mouvondy, W., Jourdan, C., Roupsard, O.,Deleporte, P., Hamel, O., Nouvellon, Y., 2005. Age-related equations for above-and below-ground biomass of a Eucalyptus hybrid in Congo. Forest Ecol.Manage. 205, 199–214.

Sayer, E.J., 2006. Using experimental manipulations to access the roles of leaf litterin the functioning of forest ecosystems. Biol. Rev. 81, 1–31.

Sayer, E.J., Heard, M.S., Grant, H.K., Marthews, T.R., Tanner, E.V.J., 2011. Soil carbonrelease enhanced by increased tropical forest litterfall. Nat. Clim. Change 1,304–307.

Sayer, E.J., Powers, J.S., Tanner, E.V.J., 2007. Increased litterfall in tropical forestsboosts the transfer of soil CO2 to the atmosphere. PloS One 2, e1299.

Schwartz, D., Mariotti, A., Trouvé, C., Van Den Borg, K., Guillet, B., 1992. Etude desprofils isotopiques 13C et 14C d’un sol ferralitique sableux du littoral congolais.Implications sur la dynamique de la matière organique et l’histoire de lavégétation. Compte-rendus l’Acad. Sci. 315, 1411–1417.

Schwartz, D., Namri, M., 2002. Mapping the total organic carbon in the soils of theCongo. Global Planet Change 33, 77–93.

Subke, J.-A., Inglima, I., Cotrufo, F.M., 2006. Trends and methodological impacts insoil CO2 efflux partitioning: a metaanalytical review. Global Change Biol. 12,921–943.

Sulzman, E.W., Brant, J.B., Bowden, R.D., Lajtha, K., 2005. Contribution ofaboveground litter, belowground litter, and rhizosphere respiration to totalsoil CO2 efflux in an old growth coniferous forest. Biogeochemistry 73, 231–256.

Thongo M’Bou, A., Saint-André, L., De Grandcourt, A., Nouvellon, Y., Jourdan, C.,Mialoundama, F., Epron, D., 2010. Growth and maintenance respiration of rootsof clonal Eucalyptus cuttings: scaling to stand-level. Plant Soil 332, 41–53.

Vasconcelos, S., Zarin, D.J., Capanu, M., Litell, R., Davidson, E.A., Ishida, F.Y., Santos,E.B., Araujo, M.M., Aragão, D.V., Rangel-Vasconcelos, L.G.T., De Assis Oliveira, F.,McDowell, W.H., De Carvalho, C.J.R., 2004. Moisture and substrate availabilityconstrain soil trace gas fluxes in an eastern Amazonian regrowth forest. Glob.Biogeochem. Cycl. 18, GB2009. http://dx.doi.org/10.1029/2003GB002210.

Wang, Y., Wang, H., Ma, Z., Wen, X., Li, Q., Liu, Y., Sun, X., Yu, G., 2009. Contributionof aboveground litter decomposition to soil respiration in a subtropicalconiferous plantation in southern China. Asiapacific J. Atmos. Sci. 45, 137–147.

Yang, Y.-S., Chen, G.-S., Guo, J.-F., Xie, J.-S., Wang, X.-G., 2007. Soil respiration andcarbon balance in a subtropical native forest and two managed plantations.Plant Ecol. 193, 71–84.

Zech, W., Senesi, N., Guggenberger, G., Kaiser, K., Lehmann, J., Miano, T.M., Miltner,A., Schroth, G., 1997. Factors controlling humification and mineralization of soilorganic matter in the tropics. Geoderma 79, 117–161.

Zimmermann, M., Meir, P., Bird, M., Malhi, Y., Cahuana, A., 2009. Litter contributionto diurnal and annual soil respiration in a tropical montane cloud forest. SoilBiol. Biochem. 41, 1338–1340.