lable at ScienceDirect

Soil Biology & Biochemistry 68 (2014) 402e409

Contents lists avai

Soil Biology & Biochemistry

journal homepage: www.elsevier .com/locate/soi lbio

Nutrient availability limits carbon sequestration in arable soils

Clive A. Kirkby a,b,*, Alan E. Richardson a, Len J. Wade b, John B. Passioura a,Graeme D. Batten b, Chris Blanchard b, John A. Kirkegaard a

aCSIRO National Sustainable Agriculture Flagship, CSIRO Plant Industry, GPO Box 1600, Canberra ACT 2601, AustraliabCharles Sturt University, E H Graham Centre for Agricultural Innovation, Wagga Wagga, NSW 2678, Australia

a r t i c l e i n f o

Article history:Received 12 June 2013Received in revised form11 September 2013Accepted 30 September 2013Available online 14 October 2013

Keywords:Nutrient availabilityCarbon sequestrationHumification efficiencyPriming effectFine fraction soil organic matterr strategists

* Corresponding author. Tel.: þ61 02 6246 5102; faE-mail address: Clive.Kirkby@csiro.au (C.A. Kirkby

0038-0717/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.soilbio.2013.09.032

a b s t r a c t

Soils are the largest reservoir of global terrestrial carbon (C). Conversion from natural to agriculturalecosystems has generally resulted in a significant loss of soil organic-C (SOC, up to 50% or w30e40 t ha�1) and ‘restoring’ this lost C is a significant global challenge. The most stable component of soil organicmatter (SOM), hereafter referred to as fine fraction SOM (FF-SOM), contains not only C, hydrogen (H) andoxygen (O), but substantial amounts of nitrogen (N), phosphorus (P) and sulphur (S), in approximatelyconstant ratios. The availability of these associated nutrients is essential for the formation of FF-SOM.Here we show, in short term (56 day) incubation experiments with 13C labelled wheaten straw addedto four soils with differing clay content, that conversion of straw into “new” FF-SOM is increased by up tothree-fold by augmenting the residues with supplementary nutrients. We also show that the loss of “old”pre-existing FF-SOM increased with straw addition, compared to soils with no straw addition, but thatthis loss was ameliorated by nutrient addition in two of our soils. This finding may illuminate why thebuild-up of SOC in some productive agricultural soils is often much less than expected from the amountsof C-rich residues returned to them because optimum C sequestration requires additional nutrientsabove that required for crop production alone. Moreover, it provides greater understanding of short-termdynamics of C turnover in soil, and in the longer term, may have important implications for globalstrategies aimed at increasing soil C sequestration to restore fertility and help mitigate climate change.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Six elements, C, N, P and S, alongwith H and O, contribute 97% ormore of the fresh weight of a wide range of organisms (Morowitz,1968). As a consequence the biogeochemical cycles of C, N, P and Sare often tightly linked (Mackenzie and Lerman, 1993). Redfield(1958) showed that the C:N:P atomic ratios of marine particulateorganicmatter (OM), primarily phytoplankton, were approximatelyconstant and equal to the ratios of the dissolved nutrients in theocean. Thus, treating the phytoplankton as essentially one singlepool of living OM interacting with its oceanic environment hasproved extremely useful to understand the ocean’s biogeochem-istry (Cooper et al., 1996; Field et al., 1998). We suggest a similarapproach may prove useful in understanding the biogeochemistryof soil.

SOM affects many important soil properties. Although freshplant residues are the primary raw materials for SOM formation,we define SOM throughout this paper as the organic fraction of the

x: þ61 02 6246 5399.).

All rights reserved.

soil exclusive of the coarse fraction (CF) organic materialcomposed of un-decayed plant and animal residues (SSSA, 2008).Magid and Kjaergaard (2001) showed that CF material(CF > 0.4 mm) was synonymous with light fraction organic ma-terial (LF < 1.4 g cm�3) with regards to the amounts of C and N aswell as physical appearance. This CF is often recognised as highlylabile material with fast breakdown rates in soil (Golchin et al.,1994; Wander, 2004; Crow et al., 2007). In contrast, fine fractionSOM (FF-SOM <0.4 mm; synonymous with heavy fraction SOM>1.4 g cm�3) is considered to be a more stabilised and slowlydecomposing pool of SOM (Golchin et al., 1994; Magid andKjaergaard, 2001). After removal of the CF, it was the remainingFF-SOM that Kirkby et al. (2011) showed had near constant C:N:P:Sratios. In this study we use FF-C to estimate this SOM pool. The FF-SOM accounts for 70e80% of the organic material in soils(Stevenson, 1994; Kogel-Knaber, 2002) while most of theremainder is partially degraded, but still recognisable, plant resi-dues (CF). Repairing the world’s degraded soils requires theadoption of management practices that increase the size of the FF-SOM pool in soil, thereby improving fertility and helping mitigateclimate change by removing atmospheric carbon dioxide andstoring it as more stabilised SOM.

Table 1Mass of N, P and S per 10,000 units C for some common crop residues, bacteria, fungiand fine fraction-SOM (Kirkby, 2011).

C N P S

Wheat straw 10,000 152 23 37Maize stover 10,000 225 29 32Rice straw 10,000 158 19 18Bacteria 10,000 2504 494 264Fungi 10,000 1034 110 94

FF-SOM 10,000 893 187 143

Table 2Initial total organic-C and FF-C (mg kg�1 soil), atom% 13C of the FF-C and otherselected properties of the four soils used for the incubation experiment.

Soil Buntine (1) Harden (2) Hamilton (3) Leeton (4)

Soil group Tenosol Kandosol Chromosol VertosolTexture Sand Sandy loam Sandy clay loam Clay loamClay (%) 8 15 25 60pH 4.80 5.29 5.15 5.84Total-organic C 12,071 10,815 31,633 24,284FF-C 6903 10,620 30,790 12,86013C atom% of FF-C 1.082 1.083 1.081 1.084

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409 403

A common expectation in conservation agriculture, where soil isnot ploughed and crop residues are retained rather than burnt orremoved, is that SOC levels should substantially increase. However,several studies over long periods have shown that increased res-idue inputs have resulted in little, if any, increase in SOC (Campbellet al., 1991; Soon, 1998; Rumpel, 2008). In addition, it is now rec-ognised that the addition of fresh organic material can increase themineralisation of “old” OM already in the soil by a phenomenonknown as the “priming effect” (PE) (Bingeman et al., 1953), leadingto even greater SOC losses. Fontaine et al. (2004) showed that thisPE could be of sufficient magnitude that the loss of “old” SOMexceeded the formation of “new” SOM, through humification of thefreshly added organic material, leading to a net loss of SOC overall.It is now recognised that the PE can either retard or increase “old”SOM mineralisation. When the addition of fresh-C substrates in-creases the mineralisation of “old” SOM the phenomenon isreferred to as a “positive PE” but when it retards the mineralisationof the “old” SOM it is referred to as a “negative PE”. The activation ofmicroorganisms by easily available substrates is generally consid-ered to be the main reason for the occurrence of positive PE’s in soil(Kuzyakov et al., 2000). While Kuzyakov and Bol (2006) and Hamerand Marschner (2002) suggest that microorganisms having r-strategy (organisms with a short lifespan but able to reproducerapidly under favourable conditions, e.g. when food becomesavailable) were mainly responsible for the positive PE’s theyobserved, there is no general agreement on this.

Because crop residues, such as wheaten straw, are C-rich butnutrient (N, P, S) poor, we hypothesised that supplementing a soilestrawmix with these nutrients would result in a greater proportionof straw-C being transformed into new FF-C, i.e. the humificationefficiency (HE) would increase, than with straw alone. The nutrientratios in FF-SOM are similar to those of soil microorganisms andthis similarity supports the growing body of evidence that a largeproportion of FF-SOM originates from microbial detritus, ratherthan directly from recalcitrant plant material (e.g. Kramer et al.,2003; Sollins et al., 2006; Lehmann et al., 2007; Liang, 2008; Bolet al., 2009; Miltner et al., 2009; Liang and Balser, 2011). Theimpact of added nutrients on microbial growth and turnover istherefore the likely key to increasing the conversion of C-rich res-idues to FF-SOM. Further, we hypothesised that using 13C-enrichedstraw would allow us to track the transformation of added straw-Cinto “new” FF-C independent of any loss of pre-existing “old” FF-C.We did so because FF-SOM continually turns over, albeit slowly, andthe common observation that SOC failed to build up in soilsreceiving large C inputs (Campbell et al., 1991; Soon, 1998; Rumpel,2008) may be due to faster decomposition of “old” pre-existing FF-SOM (Kuzyakov et al., 2000) rather than no, or low, conversion ofstraw-C into “new” FF-C. In this paper we investigate the role ofinorganic nutrients (N, P and S) in short-term C dynamics in soil bymeasuring both the amount of “new” FF-SOM formed by humifi-cation and the amount of “old” FF-SOM lost by decompositionwhen wheaten straw was added to a range of arable soils.

2. Materials and methods

We incubated four diverse soils in laboratory microcosms withand without straw and supplementary inorganic N, P and S andfollowed the changes in size of the pre-existing and newly formedFF-SOM pools. The added nutrients, equivalent to 5 kg N, 2 kg P and1.4 kg S per tonne of straw, were designed to augment the nutrientswithin the straw to achieve the ratios of C:N:P:S found in FF-SOM(Table 1; Kirkby et al., 2013). In all experiments any partiallydecomposed plant residue (CF) was removed from the soils bothbefore the experiment began and after the incubation with strawusing a dry-sieving/winnowing procedure (Kirkby et al., 2011) such

that any C remaining in the soil was assumed to be FF-C. In this wayonly changes in themore stable FF-SOM pool were assessed (Kirkbyet al., 2011).

2.1. Characterisation of the initial soil

Surface soil samples (0e15 cm) were collected from fourdiverse agricultural soils (clay range 8e60%; starting FF-C range0.6e3%, Table 2) representative of different agro-ecological regionsof Australia. All CF organic material was removed as describedabove. The difference between the total-OC and FF-C values inTable 2, particularly for soils 1 and 4, indicates the level of partiallydegraded plant remains in some soils and indicates why it is soimportant to remove this material when studying FF-C dynamics.More detailed information regarding the soils can be found inKirkby et al. (2013). Note that soil 1 (Buntine) is a Tenosol (Table 2)which was previously reported as a Kandosol in Kirkby et al.(2013).

2.2. Incubation experiment

Uniformly-stable isotope-labelled mature wheat stem(internode) > 97 atom% 13C, was used as the enriched material(IsoLife Wageningen, Netherlands). This highly-enriched materialwas diluted (1:25; final atom% 13C ¼ 4.940) with non-labelledwheat stem (internode) material collected at harvest from Site2 which had been oven dried (70 �C) and stored. Only internodestem material was used to maximise the uniformity of the ma-terial added to each replicate (Kirkby et al., 2013). The straw wascut into pieces approximately 5 mm long prior to mixing with anysoil.

2.3. Treatments

There were three treatments in the experiment: (soil alone, thecontrol), (soil þ straw), (soil þ straw þ supplementary nutrients).All weights were to 0.1 mg accuracy and the initial atom% 13C wascalculated for each replicate based on the soil weight and theweights of the labelled and non-labelled straw added.

Table 3Selected properties of fresh and partially degraded straw after 56 days incubation.Data are means and SEM. Fresh straw n ¼ 5; degraded straw n ¼ 8, two from each ofthe four soils.

Fresh straw Degraded straw

Ash (%) 3.3 (0.1) 37.3 (2.1)% Of straw that is organic matter 100 66.0 (2.1)C (as % of OM) 45.5 (0.5) 45.4 (1.3)OM remaining (as % of original) 100 61.8 (3.9)C remaining (as % of original) 100 60.9 (4.4)

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409404

Treatment 1 (control soils)Three replicates for each soil type consisting of 40 g soil

(equivalent oven dry weight) brought up to 70% field capacity (fieldcapacity is defined as the gravimetric water content at �100 cmsuction, Kirkby et al., 2013) using distilled water.

Treatment 2 (soil þ straw)Ten replicates for each soil type consisting of 40 g soil (equiva-

lent oven dry weight) 480 mg unlabelled straw and 20 mg labelledstraw were prepared and brought up to 70% field capacity usingdistilled water. 500 mg straw per 40 g soil equates to an approxi-mate field rate of 12 t straw ha�1e7.5 cm soil depth.

Treatment 3 (soil þ straw þ supplementary nutrients)Ten replicates as per treatment 2 but with added nutrients. A

nutrient solution was made containing 1.285, 0.5 and 0.325 mg perml of N, P and S respectively (as ammonium nitrate, potassiumdihydrogen phosphate and ammonium sulphate). The pH of thesolutionwas adjusted to pH 7 using 10M sodium hydroxide. 2ml ofthis nutrient solutionwas added per 40 g dry soil and the soils werethen brought up to 70% field capacity using distilled water. Therates of nutrient addition were calculated to achieve 30% humifi-cation of the freshly added straw-C (as described in detail in Kirkbyet al., 2013), as 30% is among the highest value for humification offresh crop residues when added to soil (e.g., Himes, 1998).

2.4. Incubation conditions

Each replicate was placed in an open plastic vial measuring40 mm diameter and 60 mm high. Each vial was placed in a 1 L air-tight Mason type jar containing 5 ml distilled water in the bottomof the jar to reduce drying during the incubation. An open scintil-lation vial containing 10 ml 1 M sodium hydroxide (NaOH) was alsoplaced in the jar along with the soil to absorb carbon dioxide (CO2)respired during the subsequent incubation. Four jars containingonly the CO2 traps and water in the bottomwere incubated to act asblanks. The samples were incubated in a darkened constant tem-perature room at 30 �C for 56 days after which time the CO2 evo-lution rate became low and the evolution rate curve declined toapproach an asymptote (Paul and Clark, 1996). The CO2 traps werechanged on days 7, 14, 28, 43 and 56. After removal from the Masonjar 5 ml 0.05 M barium chloride was added to the NaOH (to pre-cipitate any carbonate) plus a few drops of phenolphthalein. TheNaOH was titrated to a pink colour using 0.05 M HCl. The CO2evolved in each period was calculated using Equation (1).

C or CO2ðmgÞ ¼ ½ðBv � SvÞ � NE� (1)

Where:

Bv ¼ vol acid to titrate blankSv ¼ vol acid to titrate sampleN ¼ normality of acidE ¼ Equivalent weightEquivalent weight for CO2eC ¼ 6Equivalent weight for CO2 ¼ 22

When changing the CO2 traps the soil from each vial was tippedout onto a plastic sheet and remixed, with distilled water added toadjust the moisture if necessary. The soils were remixed to enhancestraw breakdown and improve homogeneity by: (i) re-aerating thesoil as it is generally acknowledged that oxygen is a requirement forthe dominant heterotrophic soil organisms responsible for OMbreakdown (Franzluebbers, 2005) (ii) create soil disturbance, whichpreferentially increases the breakdown rate of coarse LF organic

material (Cambardella and Elliott, 1992) and (iii) redistributing thesoilestraw-nutrient mixture to avoid ‘nutrient poor’ zones thatmay arise around straw pieces during straw breakdown. In addi-tion, ambient air was pushed into the open jar several times using adry 100 ml syringe. This was done to ensure the air in each jar wasconsistent with ambient air conditions prior to placing the sampleback in the jar and resealing.

2.5. Sample analysis

At the completion of the incubation period the samples weredried at 45 �C for several days. Any partially degraded wheatenstraw remaining was removed by the dry-sieving winnowingmethod such that any C remaining in the samples at the comple-tion of the 56 day incubation period was again assumed to be FF-C.In addition, each replicate was examined under a low power dis-secting microscope to ensure no straw fragments remained in thesample. The whole of each replicate soil sample, with anyremaining straw removed, was pulverised using a ring and puckmill to pass through a 100 mesh sieve (150 mm opening) and thenthoroughly mixed to ensure that any sub-samples subsequentlyanalysed were representative of the whole sample. Total FF-C andatom % FF-13C were determined using a dry combustion analyserwith an attached isotope ratio mass spectrometer (Europa Scien-tific Model 20-20). The partially degraded straw from each samplewas thoroughly brushed free of any mineral material adhering tothe straw (using a small artist’s brush while looking through adissecting microscope) and weighed. Eight replicates, two fromeach soil type, were analysed for ash content and total-C to assesshow much C remained in the partially degraded straw and thedegree of mineral contamination. The original “clean” straw wasconsidered to be 100% organic material and the degree ofcontamination of the partially degraded straw remaining after theincubation was calculated as the amount of ash in the sample overand above that found in the original “clean” straw samples. Theamount of organic material remaining in the partially degradedstraw was calculated as the oven dry weight of the remainingpartially degraded straw minus the weight of the ash over andabove that found in the “clean” straw. The C content of theremaining straw was calculated as a percentage of the mass oforganic material remaining.

2.6. Statistical analysis and associated calculations

The results were analysed using one or two way ANOVA asappropriate using the SigmaStat statistical software program(Systat Software Inc. USA). If the ANOVA detected a statistical dif-ference an all pair-wise multiple comparison was conducted usingthe Holm-Sidak method to identify which treatments were signif-icantly different. An explanation of which data was measuredand which was derived, and how it was derived, are shown inSupplementary Table 1. An example of the data calculations, usingthe full soil 2 data set, is shown in Supplementary Table 2.

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409 405

3. Results

3.1. Mass and composition of remaining partially degraded straw

The remaining partially degraded straw had 34% mineralcontamination (Table 3) over and above the original straw and thuswas 66% organic material. At the completion of the 56-day incu-bation period there was no difference in the C content of the initialand partially degraded straw (P > 0.8) and the loss of organicmaterial mass mirrored the loss of C. Using the weight of thecleaned partially degraded straw remaining, and assuming that allremaining straw samples contained 66% organic material, an esti-mate of the straw organic material remaining, or lost, wasmade. Onaverage, nutrient addition increased the amount of organic mate-rial lost from the straw during the 56-day incubation by 85% (Fig.1).

3.2. Respiration

Although the CO2 efflux from soils without straw was slightlyelevated during the early (day 7) incubation period, overall it waslow and relatively constant over the 56-day incubation period(Fig. 2). The total CO2 efflux from soils with added straw was 4.5times that from soils without straw (range 3e6.5 fold). For soilswith straw and straw plus nutrients, 74% and 80% respectively,occurred during the first half of the incubation period. The early(day 7) efflux from soils with added straw was 8.5 times that fromsoils without straw (range 4.9e15.4 fold). The early efflux from soilswith added nutrients was 20% greater (range 7e47%) than fromsoils with straw alone.

3.3. C mass balance

The C mass balance for soils without added straw is a straight-forward calculation as the only losses from the system are C lossesvia microbial respiration. By adding the total amount of CO2respired over the 56-day incubation period (integrating the areaunder the graphs in Fig. 2) to the final FF-C value (Table 4) between98% and 100% of the initial C could be accounted for.

Calculating a Cmass balance for soilswith added strawwasnot asstraightforward as these soils contained some partially degradedstraw that had to be carefully removed and cleaned. While theweight of the cleaned, partially-degraded strawwasmeasured for allreplicates (to estimate themass of organic material lost) an estimate

Fig. 1. Effect of nutrient treatment (addition or no addition) on straw organic matterlost at the completion of the 56 day incubation period. Data are means and SEM,n ¼ 10.

of the C balance using actual straw analyses was only done for tworeplicates from each soil where the remaining cleaned, partiallydegraded straw was analysed for ash and C content (Table 3). Usingthe weight of the remaining cleaned, partially degraded straw andthe appropriate C concentrations for this straw, the amount of Cremaining in the straw could therefore be estimated. Summing theamount of C in the remaining straw with the C lost via microbialrespiration and the C remaining in the soil, indicated that between97and 104% of the initial C could be accounted for (Table 5).

3.4. Changes in soil 13C

Soils incubated without added straw lost 13C via microbialrespiration as expected (Fig. 3). The atom% 13C of these soils did notchange (Table 4), indicating that 12C and 13C were mineralised atthe same rate. All soils incubated with added straw showed a netincrease in 13C (Fig. 3, Table 4). The net increase in 13C for soils withadded nutrients was 270% greater (range 50e480%) than for soilswith straw but without nutrients.

3.5. Changes in soil FF-C

While the net changes in C are reported as mg kg�1 soil (Figs. 3and 4) they can also be reported as a percentage of the straw Cadded to the system. Reported in this way, the net percentagechange in C is a measure of a system’s ability to sequester and retainC, which we define as the net humification efficiency (NHE).Although no straw was added to the control soils the net change inC can be expressed similarly by using the average C added to thesoils with added straw for the calculation. All soils incubatedwithout added straw lost FF-C with an effective NHE of �8.8%(range �7.0 to �11.2%, Fig. 4).

The net change in FF-C (NHE), for soils incubated with straw butwithout added nutrients, varied in response (Fig. 4). Two soils (2and 3) showed net FF-C increases (þ4.1 and þ10.4% NHE respec-tively), one (4) showed small but insignificant changes (zero NHE),while one (1) showed a net FF-C decrease (�16.9% NHE). Whennutrients were added with the straw, all soils showed net FF-Cincreases of between three- to ten-fold compared to soils incu-bated with straw alone (NHE range þ14.7 to þ40.4%).

Similar to the NHE we define the absolute mass of straw-Ctransformed into “new” FF-C, as a percentage of the straw Cadded to the system, the gross humification efficiency (GHE) andthis is a measure of a system’s ability to form new FF-C. The dif-ference between the GHE and the amount of “old” pre-existing FF-Clost by microbial respiration is the NHE. Because no straw wasadded to the control soils, the C lost must have been from the “old”pre-existing FF-C pool and the GHE and NHE are equivalent (Figs. 4and 5b). Adding straw alone significantly increased the amount ofold FF-C lost (Fig. 5b) indicating that the decomposition of old FF-SOM increased following the addition of the straw. Adding sup-plementary nutrients with the straw further increased the loss ofold FF-C in soils 3 and 4, but decreased the loss in soils 1 and 2(Fig. 5b). With straw alone the proportion of straw-C transformedinto new FF-C across the soils (the GHE) ranged from 13 to 33%. Forsoils with added nutrients the GHE increased to between 37 and74%, an increase of between 1.4 and 3.3 times compared to soilswith straw alone. The GHE for all soils with added straw waspositively and significantly correlated with the early (day 7) CO2efflux (Fig. 6, Pearson correlation coefficient ¼ 0.92, P < 0.01).

4. Discussion

Although microbial biomass was not directly measured in thisstudy, the 8-fold increase in early (day 7) CO2 efflux from soils with

Fig. 2. Effect of straw and nutrient treatment (addition or no addition) on the CO2eC efflux from four soils (soils 1 to 4 ¼ a, b, c, d respectively) over 56 days. Data are means andSEM. Treatments with no straw addition n ¼ 3, treatments with straw addition n ¼ 10. Where error bars not shown symbols are larger than error bars.

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409406

added straw is indicative of a large increase in microbial activity(and thus presumably microbial biomass) following straw addition(Wu et al., 1993; Rochette et al., 1999). Nutrient addition furtherincreased the CO2 efflux, which is consistent with the previousincubation study reported by Kirkby et al. (2013), where the mi-crobial biomass-C increased by 1.6e2.6-fold in soils 7 days after theaddition of straw with added nutrients.

The increase in atom% 13C in all soils with added straw indicatethat straw-Cwas transformed into new FF-C in all of these soils. Theincreased loss of pre-existing FF-C (12C þ 13C) in soils with strawalone, compared to soils without straw, is indicative of a positivepriming effect (Kuzyakov et al., 2000). In particular, while soil 1with added straw alone gained 13C, it also lost more pre-existing FF-C than the amount of new FF-C (12Cþ 13C) formed and thus, overall,it showed a net FF-C decrease. The net loss of FF-C, following fresh

Table 4Mean final FF-C (mg kg�1 soil) and atom% 13C levels of four soils incubated with/without wheaten straw and supplementary nutrients at the completion of the 56day incubation period. For FF-C values all partially degraded straw fragments wereremoved from the soils before C analysis.

Soil: 1 2 3 4

Soil only, FF-C 6367 (13) 10,177 (15) 30,147(55)

12,460 (85)

FF-C change from initial �536 �443 �643 �400Atom% 13C 1.082 1.083 1.081 1.084

Soil þ straw, FF-C 5943 (169) 10,860 (61) 31,390(216)

13,071 (123)a

FF-C change from initial �960 þ240 þ600 þ211Atom% 13C 1.486 1.729 1.206 1.417

Soil þ straw þ nutrients,FF-C

7727 (135) 12,410 (146) 32,346(139)

15,460 (73)

FF-C change from initial þ824 þ1790 þ1556 þ2600Atom% 13C 1.993 1.774 1.432 2.269

For all soils, except (soil 4 þ straw), the final FF-C values were significantly differentfrom initial FF-C values, P < 0.001.For all soils þ straw n ¼ 10, for soils only n ¼ 3. Values in parentheses are SEM.

a Soil 4 þ straw final FF-C values were not significantly different from initial FF-C:n ¼ 10, P ¼ 0.197.

OM addition in nutrient-poor soils, such as soil 1, without supple-mentary nutrient addition, is consistent with other published re-sults (Fontaine et al., 2004) even with repeated straw additions(Kirkby et al., 2013). Similarly, while soil 4 gained 13C there was nosignificant difference in the amount of new FF-C formed and pre-existing FF-C lost and so showed no net change in FF-C overall. Incontrast, with straw alone the amount of new FF-C formed in soils 2and 3 was greater than the amount of pre-existing FF-C lost and sototal FF-C increased.

Despite nutrient addition increasing the amount of pre-existingFF-C lost from soils 3 and 4, overall, for all soils with nutrientaddition the amount of new FF-C formed was always greater thanthe amount of pre-existing FF-C lost and thus all soils with nutrientaddition showed a net FF-C increase.

Co-metabolism is one mechanism that could explain the posi-tive PE following straw addition and which, in two of the soils,increased with nutrient addition. In co-metabolism, organismswhich specialise in the decomposition of fresh organic material(FOM) inputs release enzymes that are very efficient at mineralisingthese fresh inputs. However, these enzymes can also mineralise

Table 5C mass balance (mg kg�1 soil) for two replicates from each of four soils at thecompletion of the 56 day incubation period with added straw.

Soil 1a 1b 2a 2b

Initial total-C values 12,541 12,537 16,487 16,452Final FF-C values 6701 6534 10,730 10,730C in remaining straw 3508 4455 3193 3020C lost via respiration 1930 2058 2184 2218% C accounted for 96.8 104.1 97.7 97.1

Soil 3a 3b 4a 4b

Initial total-C values 36,510 36,586 18,603 18,795Final FF-C values 32,440 30,810 13,480 13,790C in remaining straw 2223 3519 3793 4188C lost via respiration 1903 1831 1572 1464% C accounted for 100.2 98.8 101.3 103.4

Initial total-C values include initial soil FF-C þ added straw-C.Final FF-C values are for soils with any partially degraded straw removed.

Fig. 3. Effect of straw and nutrient treatment (addition or no addition) on the changes in FF-13C of four soils at the completion of the 56 day incubation period. Data are means andSEM. Treatments with no straw addition n ¼ 3, treatments with straw addition n ¼ 10. Values above/below bars show changes in FF-13C expressed as a % of the straw-13C added. Fortreatment without straw addition mean straw-13C added to plus straw treatments used for the calculation.

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409 407

pre-existing SOM, although not as efficiently, leading to anenhanced mineralisation of pre-existing FF-C or a positive PE(Fontaine et al., 2003; Blagodatskaya and Kuzyakov, 2008). Wesuggest that: (i) straw and nutrient addition leads to an increase inthe size of the microbial biomass, as indicated by the increased CO2efflux (Fig. 2) (ii) there is an increase in amount of enzymes releasedthat are especially efficient at mineralising the FOM inputs, andwith nutrient addition, leading to even greater FOM mineralisation(iii) these enzymes, although not as efficiently, can also mineralisepre-existing SOM, leading to an enhanced mineralisation of pre-existing FF-C (Fig. 5) or an increased positive PE.

The decreased loss of pre-existing FF-C following nutrientaddition for soils 1 and 2 could be explained by assuming thatmicrobial nutrient mining was a major factor inducing the positivePE following straw alone addition. Following the addition of C-rich(but nutrient-poor) straw (Table 1), the growth of the nutrient-rich(but C-poor) microbial population could only reach its optimal levelif N, P and S could be obtained from a source other than the addedstraw. Without adequate freely-available inorganic nutrients, mi-croorganisms can only get the N, P and S they need to optimise theirgrowth by mineralising pre-existing FF-SOM, which represents analmost inexhaustible supply of these nutrients (Fontaine et al.,2011). The soil microbes are able to use the high energy-C of the

Fig. 4. Effect of straw and nutrient treatment (addition or no addition) on the net change in FSEM. Treatments with no straw addition n ¼ 3, treatments with straw addition n ¼ 10. Valuetreatment without straw addition mean straw-C added to plus straw treatments used for t

fresh inputs to offset the energy required to produce the enzymesnecessary to mineralise the more stabilised pre-existing FF-SOM,which in turn releases nutrients (particularly N, P and S) which areneeded to sustain microbial growth, but also leading to anenhanced mineralisation of the pre-existing FF-C (Blagodatskayaand Kuzyakov, 2008; Guenet et al., 2010a, 2010b). Following theaddition of inorganic N, P and S (along with the straw) microor-ganisms are able to obtain extra N, P and S from the soil solution tooptimise their use of the C-rich (but nutrient poor) straw and,therefore, do not have to expend energy producing enzymescapable of mineralising the relatively stable pre-existing FF-SOM.Thus, the mineralisation of the pre-existing FF-SOM is reduced (i.e.the positive PE is reduced). It is likely that multiple mechanismsoccur simultaneously and that the balance between them de-termines whether the decomposition rate of old FF-SOM increasesor decreases following the addition of fresh residues and nutrients(Guenet et al., 2010a).

The great majority of OM mineralisation in soils with addedstraw and straw plus nutrients, as indicated by CO2 efflux, occurredduring the first half of the incubation period (mean 74% and 80%respectively). The large increase in straw-C lost with nutrientaddition (85%), compared to soils with straw alone, but only a 9%increase in CO2 efflux indicates that much of the “lost” straw-C was

F-C for four soils at the completion of the 56 day incubation period. Data are means ands above/below bars show net changes in FF-C expressed as a % of the straw-C added. Forhe calculation.

Fig. 5. Effect of straw and nutrient treatment (addition or no addition) on (a) new FF-C formed and (b) old FF-C mineralised at the completion of the 56 day incubation period. Dataare means and SEM. Treatments with no straw addition n ¼ 3, treatments with straw addition n ¼ 10. Values above/below bars show changes in FF-C expressed as a % of the straw-Cadded. For treatment without straw addition mean straw-C added to plus straw treatments used for the calculation.

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409408

transformed into microbial-C rather than respired as CO2eC sug-gesting that most microbial assimilation of straw-C probablyoccurred during the first half of the incubation period as well. Thisis also supported by the strong correlation between early (day 7)CO2 efflux and GHE as microbial assimilation of straw-C is a pre-requisite for “new” FF-C formation if the generally held hypothesisthat microbial detritus, rather than recalcitrant plant material,contributes most to the FF-C pool is valid (e.g. Kramer et al., 2003;Sollins et al., 2006; Lehmann et al., 2007; Bol et al., 2009; Miltneret al., 2009). The greater proportion of microbial activity thatoccurred during the first half of the incubation period (i.e., threetimes more CO2 emitted in the first half of the incubation)) couldalso mean that it is fast-growing microorganisms, or r strategists,that are responsible for most of the OMmineralisation and straw-C

Fig. 6. CO2eC efflux from four soils after seven days of the 56 day incubation withadded straw (with and without supplementary nutrient addition) versus new FF-Cformed at the end of the 56 day incubation period. Data are means and SEM,n ¼ 10. Open symbols are for treatments without nutrient addition, closed symbols arefor treatments with nutrient addition. Where error bars not shown symbols are largerthan error bars.

assimilation. While Kuzyakov and Bol (2006) and Hamer andMarschner (2002) both suggest that microorganisms having rstrategy were mainly responsible for the PE’s they observed, theevidence presented here suggests they may also be responsible forthe majority of straw-C that is transformed into new FF-C as well.We hypothesise that, especially when nutrients were added alongwith the straw, these fast-growing specialist FOM-degrading mi-croorganisms could fully express their potential for rapid growth,as suggested by Fontaine et al. (2004). However, when the majorityof the labile material was exhausted or, when nutrients becamelimiting (as in the soils with straw alone) these organisms died,with a reduction in CO2 efflux rates and with their detritus even-tually contributing to the FF-SOM pool. This hypothesis warrantsfurther investigation.

Our findings clearly indicate that the addition of adequate nu-trients to incorporated energy-rich crop residues can increase, byup to three-fold, the amount of new FF-C formed from those resi-dues. Moreover, the addition of these nutrients may reduce theamount of pre-existing FF-C lost due to the PE resulting in a netincrease, by three- to ten-fold, in soil FF-C levels. While we recog-nise that this was a short-term experiment involving a single in-cubation period of 56 days, the results are consistent with ourprevious observations (Kirkby et al., 2013) that showed a similarand cumulative build-up of newly formed FF-SOM across 7consecutive incubations over 21months. Further work to assess thestability of the FF-SOM formed under these conditions is warrantedalong with investigation of its contribution to longer-term C sta-bilisation in soil.

These findings have several important implications. Firstly, theyprovide a possible explanation, namely an inadequate supply of N,P, and S, for the low levels of C-sequestration often observed inlong-term experiments in conservation agriculture, where cropresidues are retained in situ rather than burnt or removed. If newresidue-C is to be sequestered into the FF-SOM pool then new N, Pand S will need to be sequestered with it (Kirkby et al., 2011). This isespecially relevant in modern farming systems which aim to usenutrients more efficiently by supplying only the amounts neededfor optimum economic return for crop production and a low risk of

C.A. Kirkby et al. / Soil Biology & Biochemistry 68 (2014) 402e409 409

environmental damage when “surplus” nutrients are lost from thesystem. This conservative use of fertilisers, including the precisetargeting of nutrient supply to a crop’s requirement, may inadver-tently limit the supply of nutrients for microbial growth and FF-Cformation.

Secondly, it is apparent that understanding why large amountsof old pre-existing FF-C are often lost following the retention ofcrop residues is just as important as understanding hownew FF-C isformed. While such losses have been known for some time, theunderlyingmechanisms, and strategies to reduce the losses, remainunclear and warrant further investigation if optimum increases inFF-SOM levels are to be realised following crop residue retention.

And thirdly, while our results show that adding supplementarynutrients can greatly increase the proportion of straw-C trans-formed into new FF-C, they also imply that the nutrients will haveto remain sequestered with the C in the FF-SOM for as long as itexists. Thus the potential value of these nutrients is a hidden costgenerally overlooked when considering C sequestration in soil. Thisis an important consideration for C-trading systems currently beingdeveloped globally.

Acknowledgements

We thank two reviewers whose insightful comments and sug-gestions greatly improved the quality of the manuscript.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.soilbio.2013.09.032.

References

Bingeman, C., Varner, J., Martin, W., 1953. The effect of the addition of organicmaterials on the decomposition of an organic soil. Soil Sci. Soc. Am. Proc. 17(34), 38.

Blagodatskaya, E., Kuzyakov, Y., 2008. Mechanisms of real and apparent primingeffects and their dependence on soil microbial biomass and communitystructure: critical review. Biol. Fertil. Soils 45, 115e131.

Bol, R., Poirer, N., Balesdent, J., Gleixner, G., 2009. Molecular turnover time of soilorganic matter in particle-size fractions of an arable soil. Rapid Commun. MassSpectrom. 23, 2551e2558.

Cambardella, C.A., Elliott, E.T., 1992. Particulate soil organic matter changes across agrassland cultivation sequence. Soil Sci. Soc. Am. J. 56, 777e783.

Campbell, C.A., Bowren, K.E., Schnitzer, M., Zenter, R.P., Townley-Smith, L., 1991.Effect of crop rotations and fertilization on soil organic matter and somebiochemical properties of a thick black chernozem. Can. J. Soil Sci. 71, 377e387.

Cooper, D.J., Watson, A.J., Nightingale, P.D., 1996. Large increase in ocean-surfacefugacity in response to in-situ iron fertilization. Nature 383, 511e513.

Crow, S.E., Swanston, C.W., Lajtha, K., Brooks, J.R., Keirstead, H., 2007. Densityfractionation of forest soils: methodological questions and interpretation ofincubation results and turnover time in an ecosystem context. Biogeochemistry85, 69e90.

Field, C.B., Behrenfeld, M.J., Randerson, J.T., Falkowski, P., 1998. Primary productionof the biosphere: integrating terrestrial and oceanic components. Science 281,237e240.

Fontaine, S., Mariotti, A., Abbadie, L., 2003. The priming effect of organic matter: aquestion of microbial competition? Soil Biol. Biochem. 35, 837e843.

Fontaine, S., Bardoux, G., Abbadie, L., Mariotti, A., 2004. Carbon input may decreasesoil carbon content. Ecol. Lett. 7, 314e320.

Fontaine, S., Henault, C., Aamor, A., et al., 2011. Fungi mediate long term seques-tration of carbon and nitrogen in soil through their priming effect. Soil Biol.Biochem. 43, 86e96.

Franzluebbers, A.J., 2005. Organic residues, decomposition. In: Hillel, D. (Ed.),Encyclopedia of Soils in the Environment. Elsevier Ltd, Oxford, U.K.,pp. 112e118.

Golchin, A., Oades, J.M., Skjemstad, J.O., Clarke, P., 1994. Study of free and occludedparticulate organic matter in soils by solid state 13C CP/MAS NMR spectroscopyand electron scanning electron microscopy. Aust. J. Soil Res. 32, 285e309.

Guenet, B., Neill, C., Bardoux, G., Abbadie, L., 2010a. Is there a linear relationshipbetween priming effect intensity and the amount of organic matter input? Appl.Soil Ecol. 46, 436e442.

Guenet, B., Leloup, J., Raynaud, X., Bardoux, G., Abbadie, L., 2010b. Negative primingeffect on mineralization in a soil free of vegetation for 80 years. Eur. J. Soil Sci.61, 384e391.

Hamer, U., Marschner, B., 2002. Priming effects of sugars, amino acids, organic acidsand catechol on the mineralization of lignin and peat. J. Plant Nutr. Soil Sci. 165,261e268.

Himes, F.L., 1998. Nitrogen, sulfur, and phosphorus and the sequestering of carbon,p. 315e315. In: Lal, R., et al. (Eds.), Soil Processes and the Carbon Cycle. CRCPress, Boca Raton, FL.

Kirkby, C.A., 2011. Influence of Nutrient Availability on Changes in Stable SoilOrganic Matter Levels Following Crop Stubble Retention (PhD thesis). CharlesSturt University, Wagga Wagga, Australia, p. 320.

Kirkby, C.A., Kirkegaard, J.A., Richardson, A.E., Wade, L.J., Blanchard, C., Batten, G.,2011. Stable soil organic matter: a comparison of C: N:P: S ratios in Australianand other world soils. Geoderma 163, 197e208.

Kirkby, C.A., Richardson, A.E., Wade, L.J., Blanchard, C., Batten, G., Kirkegaard, J.A.,2013. Carbon-Nutrient stoichiometry to increase soil-carbon sequestration. SoilBiol. Biochem. 60, 77e86.

Kogel-Knaber, I., 2002. The macromolecular organic composition of plant and mi-crobial residues as inputs to soil organic matter. Soil Biol. Biochem. 34, 139e162.

Kramer, M.G., Sollins, P., Sletten, R.S., Swart, P.K., 2003. N Isotope fractionation andmeasures of organic matter alteration during decomposition. Ecology 84,2021e2025.

Kuzyakov, Y., Friedel, J., Stahr, K., 2000. Review of mechanisms and quantification ofpriming effects. Soil Biol. Biochem. 32, 1485e1498.

Kuzyakov, Y., Bol, R., 2006. Sources and mechanisms of priming effect induced intwo grassland soils amended with slurry and sugar. Soil Biol. Biochem. 38,747e758.

Lehmann, J., Kinyangi, J., Solomon, D., 2007. Organic matter stabilization in soilmicroaggregates: implications from spatial heterogeneity of organic carboncontents and carbon forms. Biogeochemistry 85, 45e57.

Liang, C., 2008. Evaluating Microbial Role in Terrestrial Carbon Storage (PhD thesis).University of Wisconsisn-Madison, Wisconsin, USA, p. 207.

Liang, C., Balser, T.C., 2011. Microbial production of recalcitrant organic matter inglobal soils: implications for productivity and climate policy. Nat. Rev. Micro-biol. 9, 75.

Mackenzie, F., Lerman, A., 1993. C, N, P, S global biogeochemical cycles andmodelling of global change. In: Wollast, R., Mackenzie, F., Chou, L. (Eds.), In-teractions of C, N, P and S Biogeochemical Cycles and Global Change. PublisherSpringer-Verlag, Berlin.

Magid, J., Kjaergaard, C., 2001. Recovering decomposing plant residues from theparticulate soil organic matter fraction: size versus density separation. Biol.Fertil. Soils 33, 252e257.

Miltner, A., Kindler, R., Knicker, H., Richnow, H., Kastner, M., 2009. Fate of microbialbiomass-derived amino acids in soil and their contribution to soil organicmatter. Org. Geochem. 40, 978e985.

Morowitz, H.J., 1968. Energy Flow in Biology: Biological Organization as a Problemin Thermal Physics. Academic Press, New York.

Paul, E.A., Clark, F.E., 1996. Microbiology and Biochemistry, second ed. AcademicPress, San Diego.

Redfield, A.C., 1958. The biological control of chemical factors in the environment.Am. Sci. 46, 205e221.

Rochette, P., Angers, D.A., Flanagan, L.B., 1999. Maize residue decomposition mea-surement using soil surface carbon dioxide fluxes and natural abundance car-bon-13. Soil Sci. Soc. Am. J. 63, 1385e1396.

Rumpel, C., 2008. Does burning of harvesting residues increase soil carbon storage?J. Soil Sci. Plant Nutr. 8, 44e51.

Sollins, P., Swanston, C., Kleber, C., et al., 2006. Organic C and N stabilization in aforest soil: evidence from sequential density fractionation. Soil Biol. Biochem.38, 3313e3324.

Soon, Y.K., 1998. Crop residue and fertilizer management effects on some biologicaland chemical properties of a dark grey solid. Can. J. Soil Sci. 78, 707e713.

SSSA Soil Science Society of America, 2008. Glossary of Soil Science Terms. Amer-ican Society of Agronomy, Madison, WI.

Stevenson, F.J., 1994. Humus Chemistry: Genesis, Composition, Reactions, seconded. John Wiley and Sons, New York.

Wander, M., 2004. Soil organic matter fractions and their relevance to soil function.In: Magdoff, F., Weil, R.R. (Eds.), Soil Organic Matter in Sustainable Agriculture.CRC Press, Boca Raton, Fl, pp. 67e102.

Wu, J., O’Donnell, A.G., Syers, J.K., 1993. Microbial growth and sulphur immobili-zation following the incorporation of plant residues into soil. Soil Biol. Biochem.25, 1567e1573.