Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

SBB6179_proof ■ 11 May 2015 ■ 1/12

Soil Biology & Biochemistry xxx (2015) 1e12

55 56 57 58 59 60 61 journal homepage: www.elsevier.com/locate/soilbio 62 63 64 65 66 67 68 69 70 71 a a, * a, b a N. Garcia-Franco , M. Martínez-Mena , M. Goberna , J. Albaladejo 72 a 73 Soil and Water Conservation Department, CEBAS-CSIC (Spanish Research Council), Campus de Espinardo, P.O. Box 164, 30100 Murcia, Spain b
quera km. 4.5, 46113, Valencia, Spain
n (CIDE-CSIC), Carretera Moncada-Na Centro de Investigaciones sobre Desertificacio 74 75 76 a r t i c l e i n f o a b s t r a c t 77 78 Article history: Changes in plant cover after afforestation induce variations in litter inputs and soil microbial community 79 Received 27 October 2014 structure and activity, which may promote the accrual and physical-chemical protection of soil organic 80 Received in revised form carbon (SOC) within soil aggregates. In a long-term experiment (20 years) we have studied the effects, on 81 27 March 2015 soil aggregation and SOC stabilization, of two afforestation techniques: a) amended terraces with organic 82 Accepted 24 April 2015 refuse (AT), and b) terraces without organic amendment (T). We used the adjacent shrubland (S) as Available online xxx 83 control. Twenty years after stand establishment, aggregate distribution (including microaggregates 84 within larger aggregates), sensitive and slow organic carbon (OC) fractions, basal respiration in macroKeywords: 85 aggregates, and microbial community structure were measured. The main changes occurred in the top Microaggregates within macroaggregates 86 layer (0e5 cm), where: i) both the sensitive and slow OC fractions were increased in AT compared to S Soil C pools 87 and T, ii) the percentage and OC content of microaggregates within macroaggregates (Mm) were higher Organic amendments in AT than in S and T, iii) basal respiration in macroaggregates was also higher in AT, and iv) significant 88 Microbial activity Basal respiration changes in the fungal (rather than bacterial) community structure were observed in the afforested soils 89 Priming effect (AT and T) e compared to the shrubland soil. These results suggest that the increase in OC pools linked to 90 the changes in microbial activity and fungal community structure, after afforestation, promoted the 91 formation of macroaggregates e which acted as the nucleus for the formation and stabilization of OC92 enriched microaggregates. 93 © 2015 Published by Elsevier Ltd. 94 95 96 97 98 99 afforestation (to introduce trees for the first time in the area) or 1. Introduction 100 reforestation (to re-plant a formerly wooded area). A better un101 derstanding is needed of the mechanisms and factors controlling Among the ecosystem services provided by soils, climate change 102 the accrual and stabilization of SOC following afforestation. mitigation through C sequestration is of growing interest. This 103 The amount and quality of plant litter inputs is a key factor arises especially from the suggested limitations of emissions, based € gel-Knabner, 2002), while Q1 104 controlling the accumulation of SOC (Ko on a C credit trading system, in the Kyoto Protocol 105 (Intergovernmental Panel on Climate Change, 1997; Six et al., 2002). promoting the processes involved in soil aggregation (Abiven et al., 106 Soil organic carbon (SOC) sequestration may be achieved by means 2007). Physical soil properties such as soil structure or aggregation 107 of afforestation and other types of land-use conversion (De Gryze regulate many biological and chemical soil processes linked with C 108 et al., 2004). Despite the considerable SOC sequestration potential sequestration. Particularly, the formation of soil aggregates pro109 of afforestation the results reported by different studies are conmotes the protection of organic matter against decomposition and 110 re et al., tradictory (Wiesmeier et al., 2009; Cao et al., 2010; Laganie oxidation (Jastrow et al., 2007). According to the conceptual model 111 2010). This may be attributed to: (a) the environmental conditions, of Golchin et al. (1994), the fresh and labile pools of organic matter 112 mainly rainfall regimes, and (b) according to the case of cause a rapid stimulation of the soil microbiota, accompanied by a 113 significant increase in macroaggregates formation. Other authors 114 showed significant correlations between the labile C pools and soil 115 aggregation (Bhattacharyya et al., 2012). In addition, the O-alkyl * Corresponding author. Tel.: þ34 968 396263; fax: þ34 968 396213. 116 groups e such as those of carbohydrates e have been considered as E-mail address: [email protected] (M. Martínez-Mena). 117 118 http://dx.doi.org/10.1016/j.soilbio.2015.04.012 119 0038-0717/© 2015 Published by Elsevier Ltd. Contents lists available at ScienceDirect

Soil Biology & Biochemistry

Q5 Q4

Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 2/12

2

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

a major source of labile organic C for microbial activity, fostering the binding of clay and silt-size particles and the formation of microaggregates within macroaggregates, increasing the stability of soil aggregates (Jastrow, 1996; Six et al., 2000a). In addition to the substantial role of labile organic matter inputs, many studies have pointed out the important function of soil microorganisms in the formation and stabilization of soil aggregates (Díaz et al., 1994; Siddiky et al., 2012). The microorganisms act in two ways: a) fungal hyphae favor the mechanical union of soil particles and b) the exudation of byproducts promotes the coalescence of primary particles (De Gryze et al., 2005; Helfrich et al., 2008). Generally, fungi are thought to be more important in soil aggregate formation than bacteria (De Gryze et al., 2005). This has led to the suggestion that manipulations to enhance C sequestration should include shifting the soil microbial community towards an increased fungal component (Jastrow et al., 2007). In this sense, the change of microbial structure or the introduction of microorganisms into the soil, with lasting effects, is very difficult to tackle with current technologies (Jastrow et al., 2007). A possible option could be to cause changes in the vegetation cover through afforestation, since the vegetation type can influence the microbial community structure (Costa et al., 2006). Afforestation is a key land-use change across the world and is considered to be a dominant factor controlling ecosystems functioning and biodiversity; however, the response of soil microbial communities to this change is not well understood (Macdonald et al., 2009). Here, we intend to increase our knowledge of this response, by using next-generation sequencing techniques to provide a detailed analysis of the structure, diversity, and taxonomic composition of both the bacterial and fungal communities in natural shrubland and afforested soil under semiarid conditions. In a previous publication, from the same experimental area, Garcia-Franco et al. (2014) showed that, based on results obtained 20 years after the plantation of trees, the afforestation of semiarid shrublands may result in either sequestration or loss of organic C in the ecosystem depending on the site preparation technique used. So, after 20 years the afforestation with soil organic amendment led to an increase of 1.3 kg C m2 in the ecosystem, while without soil amendment a decrease of 0.6 kg C m2 occurred. Here, we investigate the mechanisms of the processes defining the accrual and stabilization of SOC in afforested semiarid soils. Based on the earlier results, we hypothesized that the plantation of Pinus halepensis would increase fresh litter inputs into the soil, leading to (1) changes in soil organic C fractions which can related to soil organic C pools with different turnover rates, (2) changes in soil aggregatesize distribution due to the formation of new macroaggregates and organic C enriched microaggregates within macroaggregates, (3) increase in basal respiration within the macroaggregates as an indicator of higher microbial activity, which might be related with organic C protection in microaggregates formed within macroaggregates, (4) changes in the microbial populations structure due to the increase in ectomycorrhizal fungi associated with P. halepensis, which can produce aggregate-stabilizing mycelia, and (5) a close correlation between these changes in soil aggregation and microbial structure and activity. In our hypothesis we are assuming: (a) the OC accrual in microaggregates within macroaggregates is considered as an indicator of soil C stabilization and long-term sequestration (Six et al., 2013), and (b) the separated soil organic C fractions, arising from the fractionation procedure used, correspond to the sensitive and slow pools of Roth C (Zimmermann et al., 2007). This long-term experiment was performed under environmental conditions typical of Mediterranean semiarid areas, so the results could be extrapolated to extensive areas of land around the world. The specific objectives of this study were to analyze the

effects of the afforestation of degraded shrublands on: 1) changes in soil aggregation, 2) changes in the soil microbial community structure, and 3) physicalechemical processes of SOC protection and stabilization. 2. Material and methods 2.1. Site description and experimental design The study area was located in the Sierra de Carrascoy (Murcia), Southeast Spain (37 530 N, 1 150 W, 180 m a.s.l). The climate is semiarid, with an average annual precipitation of 300 mm and a mean annual temperature of 18  C. The mean annual potential evapotranspiration is 900e1000 mm y1. The soils are classified as Haplic Calcaric Leptosol with inclusions of Haplic Calcisols and Leptic Calcisols (FAO, 2006). The lithology is constituted by hard and compact limestone rocks. The fertility of the soils after each treatment is showed in Table 1. The dominant vegetation is composed of species typical of Mediterranean shrublands, such as Rosmarinus officinalis L., Thymus vulgaris L., and Anthyllis cytisoides L. with scattered P. halepensis. Miller. The experiment site was established in October 1992 in an area of 1800 m2 and consisted of three 20 m  30 m plots located on an east-facing hillside (25% mean slope), to test the following afforestation techniques: a) mechanical terracing with a single application of 10 kg m2 of an organic amendment, which consisted of the organic waste of urban soil refuse (USR) (García et al., 1998), and P. halepensis plantation (plot AT), and b) mechanical terracing and P. halepensis plantation, without organic amendment (plot T) addition. To test these afforestation techniques, an adjacent Mediterranean shrubland was considered as the control plot (S). More details about these afforestation techniques are given in GarciaFranco et al. (2014). 2.2. Soil sampling design In April 2012, 20 years after afforestation, a randomized soil sampling trial was designed to assess the effects of the tested factors. Six (1 m  1 m) soil sampling sub-plots were selected at each plot (18 sampling sites). The separation between sampling sites was about 10 m in one direction and 15 m in the other. The sampling sites were located under trees in treatments AT and T and under shrubs in the control S. At each sampling site, soil samples were collected from three soil depths: 0e5 cm, 5e20 cm, and 20e25 cm

Table 1 Soil properties of the topsoil (0e5 cm depth) in AT (afforested þ organic amendment), T (afforested), and S (shrubland). Soil properties

Treatments

1

Organic carbon (g kg ) Total N (%) Available P (mg kg1) Available K (meq 100 g1 soil) pH Carbonates (%) Bulk density (g cm3) Water holding capacity (%) Field capacity (33 kPa) Permanent wilting point (1500 kPa) Available water content (%) Texture

S

T

AT

12.8 ± 0.7a 0.19 ± 0.01ab 8.9 ± 0.1b 0.72 ± 0.03c

12.5 ± 0.7a 0.15 ± 0.02a 4.8 ± 0.1a 0.39 ± 0.01a

22.6 0.23 24.9 0.57

8.1 ± 0.18a 29.5 ± 1.1a 1.13 ± 0.06a

8.0 ± 0.1a 41.4 ± 1.5b 1.26 ± 0.01a

7.9 ± 0.1a 47.2 ± 1.3b 0.84 ± 0.19b

21.9 ± 1.4b 10.5 ± 1.2a

16.2 ± 0.6a 10.8 ± 0.7a

22.0 ± 1.0b 12.7 ± 0.6ª

11.5 ± 1.6b Loam

5.4 ± 0.5a Loam

9.3 ± 0.5b Silt loam

± ± ± ±

2.5b 0.02b 1.0c 0.02b

Numerical values are means ± standard errors for n ¼ 6. Different letters in rows indicate significant differences between treatments (Tukey's test, P < 0.05).

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 3/12

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

(54 sampling points for the whole experiment). At each sampling point, soil samples were randomly collected as a composite of three subsamples. Soil DNA extractions, aimed at analyzing the community structure of soil bacteria and fungi, were performed in composite surface samples (0e5 cm) collected in three sampling sites per treatment. 2.3. Analytical methods 2.3.1. Soil organic C fractionation procedure To test the changes in SOC fractions with different turnover rates (hypothesis 1), the organic fractions were separated with a combined physical and chemical method, according to Zimmermann et al. (2007). Thirty grams of soil (<2000 mm) were added to 150 mL of water and dispersed using a calibrated ultrasonic probetype with an output energy of 22 J mL1. Application of more energy may disrupt the coarse sand-sized SOM (Amelung and Zech, 1999). The dispersed suspension was then wet-sieved over a 63mm-aperture sieve until the rinsing water was clear. The fraction >63 mm, containing the sand fraction and stable aggregates (S þ A) together with particulate organic matter (POM), was dried at 40  C and weighed. The suspension <63 mm was filtered through a 0.45mm-aperture nylon mesh and the material >0.45 mm was dried at 40  C and weighed. The POM was separated by stirring the fraction >63 mm with sodium polytungstate at a density of 1.6 g cm3 (Cerli et al., 2012). The mixture was centrifuged at 1000 g for 15 min and the light fraction was decanted. Both fractions were washed with deionized water to remove all the sodium polytungstate, dried at 40  C, and weighed. The OC concentrations were determined in every fraction using an Elemental Analyzer (LECO TRUSPEC CN, Michigan, USA). The samples were analyzed in triplicate. The OC content for each fraction was calculated with the following equation:

  ðOCÞ ¼ ðOCÞfraction *ðfraction proportionÞsoil g C kg1 soil In this study, we have processed the SOC fractions in two main groups according to their turnover rates: i) a sensitive fraction (OCs) formed by the free particulate organic matter, which correspond to the light fraction separated by stirring the fraction >63 mm with sodium polytungstate at a density of 1.6 g cm3, and ii) a slow fraction (OCsw) constituted by the OC stabilized in aggregates (heavy fraction >63 mm) combined with the OC associated with clay and silt (<63 mm fraction). The separation of these two groups was based in the study of Zimmermann et al. (2007), in which a strong correlation was found between the separate SOC fractions (separated with the above procedure) and the sensitive and slow SOC pools used in Roth C. 2.3.1.1. Characterization of the OCs fraction. The molecular composition of the OCs fraction at 0e5 cm depth was determined by 13CNMR spectroscopic analysis, using a Varian Unity 300 spectrometer. Samples were filled into zirconium dioxide rotors (7 mm diameter) and spun in a magic angle spinning probe, at a rotation speed of 4.0 kHz to minimize chemical anisotropy. A ramped 1H pulse was used during a contact time of 1 ms to prevent Hartmann-Hahn mismatches. The contact time was 1500 ms. For each sample, 20,000 runing was carried out before the final spectrum. For integration, chemical shift regions were used as follows: i) aliphatic or alkyl-C (0e45 ppm) of lipids, fatty acids, and plant aliphatic polymers; (ii) O-alkyl-C (45e110 ppm) deriving primarily from polysaccharides (cellulose and hemicelluloses), but also from proteins and side chains of lignin; (iii) aromatic or aryl-C (110e162 ppm), deriving from lignin and/or protein; and finally (iv) carbonyl-C (162e190 ppm) from aliphatic esters, carboxyl groups, and amide

3

carbonyls. Integration of the peaks within each of the chemical shift regions allowed estimation of the relative C contents, expressed as percentages of the total area (Helfrich et al., 2006). 2.3.2. Soil aggregate-size distribution To test the changes in soil aggregate-size distribution due to the formation of new macroaggregates and organic C enriched microaggregates within macroaggregates (hypothesis 2), aggregate-size separation was performed by the wet-sieving method of Elliott (1986). Briefly, 100 g of air-dried (5-mm-sieved) soil were placed on top of a 2000-mm sieve and submerged for 5 min in deionized water at room temperature. The sieving was performed manually by moving the sieve up and down 3 cm, 50 times, for 2 min e to achieve aggregate separation. A series of three sieves (2000, 250, and 63 mm) was used to obtain four aggregate fractions: i) >2000 mm (large-macroaggregates; LM), ii) 250e2000 mm (smallmacroaggregates; SM), iii) 63e250 mm (microaggregates; m), and iv) <63 mm (silt plus clay-size particles; s þ c). The aggregate-size classes were oven dried (50  C), weighed, and stored in glass jars at room temperature (21  C). Sand correction was performed for each aggregate-size class because sand was not considered to be part of the aggregates (Elliott et al., 1991). Microaggregates contained within both the large- and smallmacroaggregates (LMm and SMm, respectively) were mechanically isolated according to the methodology described by Six et al. (2000b) and Denef et al. (2004). Briefly, a 10-g macroaggregate subsample was immersed in deionized water on top of a 250-mm mesh screen, inside a cylinder. The macroaggregates were shaken together with 50 glass beads (4 mm diameter) until complete macroaggregate disruption was observed. Once the macroaggregates had been broken up, microaggregates and other <250 mm material passed through the mesh screen with the help of a continuous water flow. The material retained on the 63-mm sieve was wet sieved to ensure that isolated microaggregates were water stable (Six et al., 2000b). The OC determination was performed separately for all aggregate size classes, using an Elemental Analyzer (LECO TRUSPEC CN. Michigan, USA). Samples were analyzed in triplicate. The OC content for the water-stable aggregate-size classes was calculated, at the soil level, with the following equation:

OC ¼ ðOCÞfraction *ðagg:proportionÞsoil ðg C=kgsoil Þ 2.3.3. Diversity and structure of soil microorganisms To test the changes in the microbial populations structure due to the increase in ectomycorrhizal fungi associated with P. halepensis (hypothesis 4), soil DNA was extracted by using the FastDNA Kit (Qbiogene Inc., Irvine, USA) and purified on Low Melting Point agarose gel 1.25 (wt/vol) containing 2% polyvinylpyrrolidone (PVP) (Young et al., 1993). The DNA extracts were analyzed on 1% agarose gels and their final concentrations quantified with a Nanodrop 2000 (Thermo Scientific, Wilmington, USA). The universal Eubacterial primers BSF8 (50 -TCAGAGTTTGATCCTGGCTCAG-30 ) and USR515 (50 -CACCGCCGCKGCTGGCA-3) were used for amplifying the 16S rRNA V3 fragment (Bibby et al., 2010). The universal eukaryotic primers nu-SSU-0817-59 (50 TTAGCATGGAATAA 0 0 TRRAATAGGA-3 ) and nu-SSU-1196-39 (5 -TCTGGACCTGGTGAG TTTCC-30 ) were used for amplifying the 18S rRNA V4 fragment (Borneman and Hartin, 2000). The amplifications were performed in a MyCyclerTm Thermal cycler (Bio-Rad Laboratories Inc., Hercules, USA). The PCR conditions used for amplification of the 16S rRNA fragment were: 94  C for 5 min, followed by 30 cycles consisting of 30 s at 94  C, 30 s at 56  C, and 90 s at 72  C, and a final elongation step at 72  C for 10 min. The 18S rRNA gene was amplified using the

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 4/12

4

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

following protocol: 94  C for 2 min then 35 cycles of 94  C for 10 s, 56  C for 10 s, and 72  C for 30 s, followed by 72  C for 2 min. The amplicon length and concentration were estimated, and an equimolar mix of all amplicon libraries was used for tag-encoded FLXtitanium amplicon pyrosequencing (Roche, Basel, Switzerland). It should be noticed that this PCR-based technique could underestimate low abundant groups in the bulk soil such as Glomeromycota. The sequences were processed using QIIME v. 1.5.0 (Caporaso et al., 2010) e by selecting sequences with a quality score >25, containing no ambiguous bases, without any primer mismatches, and with a sequence length between 200 and 400 bp. After quality checking, counting, sorting, and denoising, chimeras were identified using ChimeraSlayer (Haas et al., 2011). Clustering of the sequences into operational taxonomic units (OTUs) was performed using UCLUST (Edgar, 2010) and a cutoff value of 97% sequence identity. The most abundant sequence type within each OTU was selected to represent the respective OTU in further analysis. The taxonomic assignment was performed according to RDP (Wang et al., 2007) and the BLAST data base of the NCBI. The Chao1 and Shannon indices were calculated from 5986 to 5575 seqs/samples for bacteria and fungi, respectively, to estimate taxon richness and diversity. Principal coordinate analysis (PCoA), using weighted UniFrac distances, was used to visualize differences in microbial community structure across the treatments (Hamady et al., 2010). Sequences generated in this study were deposited in EMBL within the study with accession number PRJEB7535 (http://www.ebi.ac. uk/ena/data/view/PRJEB7535). 2.3.4. Soil respiration measurements in macroaggregates (>250 mm) To test the increase in basal respiration within the macroaggregates as an indicator of higher microbial activity (hypothesis 3), the amount of CO2eC released daily per kg of soil macroaggregate was measured during the incubation under controlled conditions (Nannipieri et al., 1990). Soil respiration was analyzed by placing 15 g of soil macroaggregates, moistened to 60% of their waterholding capacity, in hermetically sealed 125-ml flasks during a 31-day incubation period at 28  C (Bastida et al., 2007). Three repetitions were made per sample. The CO2 released was measured periodically (every day for the first 4 days and then weekly) using an infrared gas analyzer (Toray PG-100, Toray Engineering Co. Ltd. Japan). After each measurement, the stoppers were removed for 1 h to balance the atmosphere inside and outside the bottles. 2.4. Statistical analyses Prior to the analyses, the normality of the data was proved using the KolgomoroveSmirnov test and the homogeneity of variances with the Levene test. Data that were not distributed normally (LM, SM, LMm, OC-LM, OC-SM, OC-SM, and OC-SMm) were lntransformed. To compare all the soil variables between treatments, a General Linear Models (GLM) procedure was carried out e considering treatment and depth as fixed factors. To test the hypothesis 5, Pearson correlations were used in order to explore the relationships between the functional OC pools (OCs and OCsw) in bulk soil and the distribution of aggregates and their associated organic carbon, as well as the relationships between basal respiration and the distribution of aggregates and their associated organic carbon. The analyses were computed with SPSS 19.0 (Chicago, IL, USA) and the significance was set at p < 0.05. To compare the soil microbial communities of the treatments, analysis of variance using distance matrices was performed with the vegan package for R (ADONIS, R Development Core Team 2011; Oksanen et al., 2013). The statistical significance was tested against 999 null permutations. The effect of the treatments on the relative

Table 2 Sensitive OC pool (OCs) and slow OC pool (OCsw) concentrations (g C kg1soil) in AT (afforested þ organic amendment), T (afforested), and S (shrubland), in the bulk soil Q2 at 0e5, 5e20, and 20e25 cm depth. Depth (cm)

S

Sensitive pool (OCs) 0e5 4.02 5e20 2.59 20e25 1.74 Slow pool (OCsw) 0e5 8.78 5e20 9.40 20e25 7.28

T

AT

± 0.10aC ± 0.21aB ± 0.24aA

5.90 ± 0.78aB 3.10 ± 0.26aA 2.45 ± 0.22aA

11.65 ± 1.20bB 3.46 ± 0.34aA 1.72 ± 0.31aA

± 0.45bAB ± 0.73cB ± 0.46bA

5.87 ± 0.26aA 5.09 ± 0.17aA 5.61 ± 0.52aA

10.92 ± 0.59cB 7.02 ± 0.45bA 5.62 ± 0.35aA

Numerical values are means ± standard errors for n ¼ 6. Different lowercase letters in rows indicate significant differences between treatments at each depth within each OC pool. Different uppercase letters in columns indicate significant differences between depths within each treatment (Tukey's test, P < 0.05).

abundance of bacterial and fungal phyla (arcsine-transformed data), as well as richness (Chao 1), number of unique OTUs and diversity (Shannon index) was analyzed with a GLMs using R. To test the influence of soil parameters (bulk density, available water content, clay, silt, and sand contents, pH, carbonates, available P, and labile OC) on the microbial communities, we computed correlations between the soil physical-chemical and (bacterial or fungal) OTU abundance distance matrices through Mantel tests, with the vegan package for R. Similarly, matrix correlations between OTU abundance and aggregate-size distance matrices were performed to test for the influence of the microbial communities on the distribution of soil aggregates. 3. Results 3.1. Changes in SOC fractions after afforestation A significant increase in the sensitive OC fraction (OCs) and the slow OC fraction (OCsw) was found in AT e compared to S e in the 0e5 cm layer, while on the contrary, OCsw was significantly decreased in AT compared to the S treatment in deeper layers. When afforestation was performed without amendment (T) a significant decrease of the OCsw fraction was found e compared to treatment S e through the whole profile, while no changes were observed in the OCs fraction (Table 2). In addition, significant differences in the chemical composition of the OCs fraction at 0e5 cm depth were found between the afforested (AT and T) and the shrubland (S) soils, the former showing higher percentages of Oalkyl C and lower percentages of aryl C materials (Table 3). No differences were detected among treatments with regard to the relative contents of the alkyl-C or carbonyl-C materials (Table 3). 3.2. Distribution of water-stable aggregate-size classes Macroaggregates (>250 mm) was the predominant aggregatesize class for all treatments and depths, representing between 65 Table 3 Relative contents (%) of alkyl-C, O-alkyl-C, aryl-C, and carbonyl-C in the sensitive pool for the afforested treatments (AT and T) and shrubland (S), at 0e5 cm depth. Functional group

Chemical shift regions (ppm)

Relative content (%)

Alkyl-C O-Alkyl-C: Aryl-C: Carbonyl-C

0e45 45e110 110e165 165e220

33.2 45.1 12.5 9.25

S

T ± ± ± ±

0.05a 0.25a 0.05b 0.15a

31.7 49.2 10.2 8.8

AT ± ± ± ±

0.8a 0.7b 0.2a 0.4a

30.3 51.1 10.4 8.2

± ± ± ±

0.6a 1.1b 0.3a 0.6a

Numerical values are means ± standard errors for n ¼ 6. Different letters in rows indicate significant differences between treatments (Tukey's test, P < 0.05).

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 5/12

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

5

and 80% of the bulk soil. In general, in the afforested soils a decrease in the percentage of macroaggregates and an increase in that of microaggregates were observed with depth. Also, in soil S an increase in the microaggregates percentage with depth was found. Likewise, the silteclay particles percentage increased with depth in all treatments. The T treatment generally gave the lowest proportion of largeand small-macroaggregates (LM and SM, respectively) along the soil profile, while soils in AT and S treatments tended to have similar percentages of both macroaggregates. The afforested soils (AT and T) showed a lower proportion of microaggregates, compared to S, across the depths, while the opposite trend occurred with silteclay particles (Fig. 1). 3.3. Organic carbon associated with aggregate-size classes The OC associated with macroaggregates (>250 mm) represented 47% of the total SOC in the topsoil and increased to 76% at the deepest layer (20e25 cm) in AT. The opposite trend was observed in T and S, in which the OC in macroaggregates reached 66% and 73%, respectively, of the total SOC in the top soil and decreased to about 30% and 50%, respectively, below 5 cm depth. A reduction in the OC concentrations associated with all the aggregate-size classes was found in soil T compared to S, at all depths. Soil AT had similar OC-LM (>2000 mm) and OC-m (250e63 mm) concentrations to S along the soil profile (Fig. 2). The OC associated with the silteclay fraction (<63 mm) was higher in AT than in T and S in the top layer, whereas AT showed the lowest OC in silteclay content below 5 cm depth (Fig. 2). The OC-LM (>2000 mm) decreased with depth in all treatments, whereas the OC-SM (2000e250 mm) content only decreased with depth in the afforested treatments. 3.4. Proportion of microaggregates within macroaggregates and the associated organic carbon The AT treatment produced a higher proportion of microaggregates, within both large and small-macroaggregates, than S and T at 0e5 cm depth. In turn, T and S did not show any differences. Below 5 cm depth, no differences were found between AT and S, while T showed the lowest percentages (Table 3). The percentage of microaggregates occluded within macroaggregates decreased with depth in the afforested soils while an increase of small macroaggregates was observed in soil S. The OC associated with microaggregates within macroaggregates showed the same pattern among treatments as the percentage of microaggregates mentioned above e soil AT showing higher OC associated with large and small macroaggregates, at the surface, with respect to S and T (Table 3). Below 5 cm depth, the T treatment exhibits the lowest OC concentrations in macroaggregates, while no differences were found between AT and S. Decreases with depth of OC in macroaggregates were observed in the afforestation treatments, these reductions being more drastic in T than in AT (Table 3).

Fig. 1. Weight percentage of the water-stable aggregate-size classes distribution (g aggregate 100 g1 soil): >2000 mm (large-macroaggregates; LM), 250e2000 mm (small-macroaggregates; SM), 63e250 mm (microaggregates; m), and <63 mm (silt þ clay fraction; s þ c) in the 0e5, 5e20, and 20e25 cm soil layers of the AT (afforested þ organic amendment), T (afforested), and S (shrubland) soils. Numerical values are means ± standard errors for n ¼ 6. Bars with different lowercase letters indicate significant differences between treatments at each depth and different uppercase letters indicate significant differences between depths within each treatment (Tukey's test, P < 0.05).

3.5. Soil respiration measurements Significant differences in the basal respiration in macroaggregates were found between treatments. The AT treatment gave a higher basal respiration than treatments S and T, at all depths (Fig. 3). At the surface, basal respiration was higher in soil T than in S, while below 5 cm no significant differences were observed between them.

3.6. Correlations between functional OC pools, basal respiration, and microaggregates within macroaggregates Due to the similar characteristics and behavior of large and small macroaggregates in all the treatments, we grouped both sizeclasses together as macroaggregates >250 mm (M), to facilitate

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 6/12

6

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

Fig. 3. Basal respiration in macroaggregates (BR-M: mg CO2eC kg1 day1) in the 0e5, 5e20, and 20e25 cm soil layers, as affected by the AT (afforested þ organic amendment), T (afforested), and S (shrublands) treatments. Numerical values are means ± standard errors. Bars with different lowercase letters indicate significant differences between treatments at each depth and different uppercase letters indicate significant differences between depths within each treatment (Tukey's test, P < 0.05).

correlation with the percentage of macroaggregates and OC in macroaggregates. In a similar way, OCs and OCsw were correlated with the percentage of microaggregates within macroaggregates and its associated OC only for the AT treatment. The OCs fraction was correlated with micro within macroaggregates in treatment T, but in S no correlations were found between these parameters. Negative correlations between both SOC fractions and the

Table 4 Weight percentage (%) and organic carbon concentration (g C kg1 soil) of microaggregates within macroaggregates: LMm (microaggregates within largemacroaggregates) and SMm (microaggregates within small-macroaggregates) at 0e5, 5e20, and 20e25 cm soil depth, as affected by the AT (afforested þ organic amendment), T (afforested), and S (shrubland) treatments. Treatments

Fig. 2. Organic carbon content (g kg1 soil) of soil aggregates: >2000 mm (largemacroaggregates; LM), 250e2000 mm (small-macroaggregates; SM), 63e250 mm (microaggregates; m), and <63 mm (s þ c) in the 0e5, 5e20, and 20e25 cm soil layers of the AT (afforested þ organic amendment), T (afforested), and S (shrubland) soils. Numerical values are means ± standard errors for n ¼ 6. Bars with different lowercase letters indicate significant differences between treatments at each depth and different uppercase letters indicate significant differences between depths within each treatments (Tukey's test, P < 0.05).

the interpretation of the correlations. Significant differences were found between the treatments (Table 5). Significant, positive correlations were found between the sensitive OC fraction (OCs) and the percentage of macroaggregates and OC associated with macroaggregates, for all treatments. However, for the slow fraction (OCsw), only in treatment AT was there a tight

Weight (%) LMm 0e5 cm 5e20 cm 20e25 cm SMm 0e5 cm 5e20 cm 20e25 cm OC (g C kg1 soil) OC-LMm 0e5 cm 5e20 cm 20e25 cm OC-SMm 0e5 cm 5e20 cm 20e25 cm

S

T

AT

14.7 ± 2.0aA 24.2 ± 1.8cB 17.4 ± 1.5bA

17.1 ± 1.5aC 11.7 ± 1.5aB 3.0 ± 0.1aA

21.6 ± 2.1bB 20.7 ± 2.4bAB 17.8 ± 1.4bA

14.4 ± 1.5aA 22.3 ± 1.9bB 23.7 ± 1.5bB

15.4 ± 1.5aA 14.5 ± 1.5aA 14.0 ± 0.9aA

27.5 ± 1.5bC 20.4 ± 1.0bB 16.4 ± 3.1aA

1.3 ± 0.2aA 2.1 ± 0.2bB 1.2 ± 0.2bA

1.3 ± 0.2aB 0.5 ± 0.1aA 0.1 ± 0.03aA

2.0 ± 0.2bB 1.3 ± 0.1abA 1.1 ± 0.3bA

1.3 ± 0.2aA 1.9 ± 0.2bA 1.5 ± 0.3bA

1.4 ± 0.2aB 0.7 ± 0.2AaB 0.5 ± 0.2aA

2.7 ± 0.3bB 1.8 ± 0.2bA 1.3 ± 0.1bA

Numerical values are means ± standard errors for n ¼ 6. Different lowercase letters in rows indicate significant differences between treatments at each depth. Different uppercase letters in columns indicate significant differences between depths within each treatment (Tukey's test, P < 0.05).

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 7/12

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

Q3

7

Table 5 Pearson correlation coefficients between the sensitive pool (OCs), slow pool (OCsw), basal respiration (BR), aggregates percentage, and aggregate-associated OC. Correlations:

Treatments S

T

AT

0.816** 0.471 0.733 0.780** 0.341 0.182 0.877**

0.684** ¡0.714** 0.683** 0.659** 0.046 0.451 0.784**

0.828** ¡0.836** 0.860** 0.931** 0.433 0.864** 0.925**

0.567* 0.345 0.018 0.552* 0.273 0.474 0.245

0.093 0.281 0.048 0.316 0.032 0.235 0.261

0.786** ¡0.867** 0.752** 0.743** 0.449 0.746** 0.896**

0.798** 0.486 ¡0.838** 0.719** 0.138 0.435

0.817** ¡0.922** 0.855** 0.985** 0.098 0.823**

0.933** ¡0.918** 0.911** 0.917** ¡0.481* 0.819**

1

OCs (g kg ): M (%) m (%) Mm (%) OC-M (g kg1) OC-m (g kg1) OC-Mm (g kg1) BR (mg CO2eC kg1 d1) OCsw (g kg1): M (%) m (%) Mm (%) OC-M (g kg1) OC-m (g kg1) OC-Mm (g kg1) BR (mg CO2eC kg1 d1) BR-M (mg CO2eC kg1 d1) M (%) m (%) Mm (%) OC-M (g kg1) OC-m (g kg1) OC-Mm (g kg1)

Fig. 5. Means ± standard errors (n ¼ 3) for the relative abundance of fungal taxa in the soil surface layer (0e5 cm), in AT (afforested þ organic amendment), T (afforested), and S (shrubland) soils. Bars with different lowercase letters indicate significant differences between treatments (P < 0.05).

3.7. Changes in the soil microbial community

percentage of microaggregates not occluded within macroaggregates were found in the afforested soils, mainly in AT. The basal respiration in macroaggregates showed strong correlations with the percentage of macroaggregates and its associated OC in all treatments. It is important to note the strong correlations between basal respiration and the percentage of micro within macroaggregates: positive in treatments AT and T and negative in S. In turn, basal respiration and OC associated with micro within macroaggregates showed close correlations in the afforested soils. As occurred with OCs, the basal respiration was negatively correlated with the percentage of free microaggregates in treatments AT and T. Finally, basal respiration was correlated with OCs in all treatments, but with OCsw only in AT.

3.7.1. Soil bacterial community The bacterial community structure did not differ significantly across treatments, based on the OTU distance matrices (F ¼ 1.16, R2 ¼ 0.279, P ¼ 0.104). However, PCoA based on weighted Unifrac distance matrices, which include a phylogenetic component, suggested a distinct bacterial community structure in soils under shrubland, compared to afforested areas (Fig. 6a). No relationship was found between a battery of nine physical and chemical parameters and the bacterial community structure in the soils (r ¼ 0.076, P ¼ 0.283). The same result was found when each parameter was tested individually. The bacterial richness Chao1, the number of unique OTUs, and Shannon's index did not differ among treatments either (Table 6). The relative abundance of most of the dominant bacterial phyla (Proteobacteria, Actinobacteria, Chloroflexi, Bacteroidetes, Gemmatimonadetes, Planctomycetes, and Acidobacteria) was not statistically different among treatments (Fig. 4). The abundance of other phyla, comprising those with relative abundances below 1% or not assigned with sufficient confidence to any known bacterial phylum, did not differ significantly among treatments (Fig. 4).

Fig. 4. Means ± standard errors (n ¼ 3) for the relative abundance of bacterial phyla in the soil surface (0e5 cm), in AT (afforested þ organic amendment), T (afforested), and S (shrubland). No significant differences were detected across treatments.

3.7.2. Soil fungal community The fungal community structure differed significantly across treatments (t ¼ 1.60, R2 ¼ 0.348, P < 0.05; Fig. 6b). This was mediated by the differences in the soil physical and chemical parameters (r ¼ 0.468, P < 0.05). Particularly important were the individual effects of carbonate (r ¼ 0.408, P < 0.05) and sand contents (r ¼ 0.439, P < 0.05). In addition, fungal richness was significantly higher in AT compared to S (F ¼ 2.90, P < 0.05), while that in T did not differ significantly from the other two treatments (Table 6). No differences were found in the number of unique fungal OTUs or Shannon's index (Table 6). The relative abundance of the dominant fungal taxon (Saccharomyceta) was not statistically different among treatments (Fig. 5). Mitosporic Ascomycota was more abundant in S than in AT, while their abundance in T did not differ significantly from that in S or AT. Other Ascomycota were significantly more abundant in AT than in T and S. The Basidiomycota (Agaricomycotina) were significantly more abundant in AT than in S, while their abundance in T did not differ significantly from that in S or AT. The Chytridiomycota were significantly higher in S than in the other treatments.

*P < 0.05; **P < 0.01. M: percentage of macroaggregates (>250 mm: LM þ SM); m: microaggregates (250e63 mm), Mm: percent of microaggregates within macroaggregates; OC-M: organic carbon content in macroaggregates; OC-m: organic carbon content in microaggregates; OC-Mm: organic carbon content in microaggregates within macroaggregates; BR: basal respiration in macroaggregates.

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 8/12

8

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12 Table 6 Soil bacterial and fungal richness and diversity in afforested (AT and T) and shrubland (S) soils at 0e5 cm depth.

Bacteria Number of unique OTUs Chao1 Shannon Fungi Number of unique OTUs Chao1 Shannon

S

T

AT

3230 ± 82.8a 11,071 ± 372a 10.7 ± 0.1a

3271 ± 138a 10,647 ± 869a 10.78 ± 0.2a

3509 ± 57.6a 12,286 ± 386a 11.01 ± 0.1a

1737 ± 11.4a 4374 ± 243a 8.69 ± 0.1a

1707 ± 138.2a 5519 ± 498ab 7.8 ± 0.6a

1934 ± 64.5a 6046 ± 439b 8.7 ± 0.2a

Numerical values are means ± standard errors for n ¼ 3. Different letters in rows indicate significant differences between treatments (P < 0.05).

4. Discussion 4.1. Changes in soil aggregation

Fig. 6. Dot-plots obtained from principal coordinate analyses, representing (a) the bacterial and (b) the fungal community structure in every treatment: AT (afforested þ organic amendment), T (afforested), and S (shrubland).

Finally, the abundance of other fungal taxa, comprising those with relative abundances below 1% (i.e. Glomeromycota, and Neocallimastigomycota) or not assigned with sufficient confidence to any known fungal taxon, was significantly higher in T and AT than in S (Fig. 5). In addition, the distribution of aggregate-size classes was significantly correlated with the community structure of soil fungi (r ¼ 0.378, P < 0.05), but not with that of soil bacteria (r ¼ 0.003, P ¼ 0.52). The same trend was found for the basal respiration measured in soil macroaggregates (BR-M), which correlated significantly with the community structure of soil fungi (r ¼ 0.462, P ¼ 0.01), but not with that of bacteria (r ¼ 0.056, P ¼ 0.35).

Twenty years after the afforestation, the percentage of waterstable aggregates differed significantly across the treatments. Areas afforested with P. halepensis and receiving no organic amendments (T) showed the lowest percentages of large- and smallmacroaggregates, suggesting mechanical disturbance of the soil structure owing to the terracing works. This initial impact was followed by an active process of new macroaggregates formation (>250 mm) in the organically-amended, afforested soils (AT), which behaved similarly to the shrubland soils e particularly in the topsoil layers. This suggests that the organic amendments offset the negative impact of the terraces. Similar soil structure deterioration following mechanical terracing was described by other authors (Barber and Romero, 1994). Likewise, other studies reported increases in soil aggregation due to both soil afforestation (Caravaca et al., 2002; Khale et al., 2005) and soil organic amendment (Díaz et al., 1994). Afforestation and amendment (treatment AT) increased not only the percentage of macroaggregates (>250 mm) but also the percentage of microaggregates occluded inside the macroaggregates. In addition, the OC concentration in these new microaggregates formed in AT soil was higher than in the non-occluded microaggregates existing in the soil before afforestation, as was hypothesized above. Overall, these results suggest a hierarchical order of aggregation in AT, in which macroaggregates were the nucleus for microaggregate formation in the center of macroaggregates (Oades, 1984). Similar to previously described models of soil aggregation (see Six et al., 2004, for a review), we suggest that in a first stage following afforestation the organic amendment quickly induced the formation of macroaggregates due to: a) an increase in microbial activity, and b) inputs of binding agents like polysaccharides (Díaz et al., 1994; Golchin et al., 1994). Over time, this initial effect of the organic amendment was gradually maintained by fresh plant material entering the soil, derived from the growth of the planted vegetation (see discussion below). In a second stage, inside these macroaggregates, the presence of decomposed organic matter, metabolites and biogenic products, polyvalent cations, and other binding agents promoted the solidphase reaction between organic matter and clay and silt particles e leading to the formation of stable microaggregates (Edwards and Bremner, 1967; Golchin et al., 1994). In treatment T, the soil conditions e very low soil organic matter, little microbial activity, and few biomass inputs e were totally unfavorable for aggregates formation and the process unfolded very slowly. 4.2. Factors controlling changes in soil aggregation In order to know the relative contribution of each factor to the soil aggregation and the carbon stabilization and sequestration

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 9/12

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

capacity in the afforested areas, we analyzed the changes which occurred after each treatment. 4.2.1. Quality and amount of soil organic inputs Organic amendments significantly increased the percentage of stable aggregates in the first years following the AT treatment (Querejeta et al., 2000). Similar results have been reported worldwide (Bartoli et al., 1992; Clark et al., 2009) and, specifically, under the same environmental conditions and with the same organic amendments as ours (Roldan et al., 1996). It is worth mentioning, however, that not all organic amendments have the same effect on soil aggregation (Clark et al., 2009). The ability to promote the formation of new macroaggregates depends on: a) the content of transient binding agents such as polysaccharides or fresh plant material, and b) the amounts of microbially available C that can promote fungal proliferation (Lucas et al., 2014). In this experiment, the organic amendment had a high content of carbohydrates (see Querejeta et al., 1998); it is likely that this was a key factor in the initial stages of macroaggregate formation due to its double action as a binding agent and as a stimulating microbial resource. Several studies have shown that the effects of soil organic amendments are temporary, mainly appearing in the first week after addition (Debosz et al., 2002; Clark et al., 2009). This is attributed to the rapid turnover of labile organic pools, as Díaz et al. (1994) reported for polysaccharides. In this study, afforestation plus a single initial amendment (AT) led to an increase in the soil OCs and OCsw fractions, while afforestation without organic matter addition did not increase the OCs fraction and led to a reduction of OCsw compared to the shrubland. These changes probably are a response to the higher biomass production, and thus litter inputs, in AT compared to the other treatments (Garcia-Franco et al., 2014). Other authors obtained similar results after afforestation on
re et al., 2010; Wei et al., abandoned agricultural soils (Laganie 2013). In this way, the initial effects of the organic amendment on soil aggregation continued for treatment AT due to the increases in OC fractions. Besides quantitative changes in OC fractions, the quality of OCs plays an important role in soil aggregation and C sequestration (Steffens et al., 2009). We found that the composition of OCs in afforested soils showed an increase in O-alkyl C, such as that found in carbohydrates, with respect to the shrubland soil. The O-alkyl C compounds are considered as indicators of the amount of organic binding agents, which improve macroaggregate formation (Degens, 1997). Several authors showed that carbohydrates were significantly correlated with aggregate stability (Martens and Frankenberger, 1992). Our results agree with those of Steffens et al. (2009), who found greater aggregate stability to be associated with large contributions of O-alkyl C from the labile fraction of soil organic matter. Altogether, our results suggest that the increase of the SOC fractions, mainly OCs and the higher carbohydrates concentration in these fraction, after afforestation plus organic amendment led to an increase of the soil carbon sequestration capacity, induced by new aggregates formation. An important proportion of this OC came from microaggregates enriched in OC and occluded within macroaggregates e as can be deduced from the positive correlation between this occluded OC and both Oc fractions, especially OCs (Table 4). Other authors also obtained positive correlations between labile C fractions and macroaggregates (Bhattacharyya et al., 2012). 4.2.2. Changes in microbial activity and structure With regard to the implications of the microbial community for soil aggregate formation and C sequestration, we must discuss two aspects: a) microbial activity (hypothesis 3), and b) community

9

structure (hypothesis 4). Basal respiration, an estimate of the total microbial activity in soils (Vanhala et al., 2005), was significantly higher in macroaggregates of afforested soils receiving amendments (AT), compared to the other treatments. Afforestation performed without amendments (T), however, led to a lower macroaggregates basal respiration than in the shrubland. Similarly, in the literature the responses of microbial activity to afforestation are various. Several studies showed decreases in microbial activity in afforested areas (Goberna et al., 2007; Chen et al., 2008); in contrast, other authors reported increases in microbial activity after afforestation (Mao and Zeng, 2010). This apparent discrepancy is mainly due to the previous use and characteristics of the soils, the afforestation method, or the stand age. The tight correlations between basal respiration in macroaggregates and the percentage of macroaggregates (>250 mm) in all treatments suggest that this microbial activity could be an important factor in the new macroaggregates formation. In accordance with these results, many studies have indicated the key role of microbial populations in soil aggregate formation (Siddiky et al., 2012; Daynes et al., 2013). However, the correlations between basal respiration in macroaggregates with the percentage of microwithin macroaggregates (BR-Mm) and basal respiration with OC associated to micro-within macroaggregates (BR-OCMm), showed very different behavior between the native shrubland and the afforested soils, with respect to aggregation processes and C sequestration. The strong, positive correlations between BR-Mm and BR-OCMm in soils AT and T suggest an active, microbialinduced process of microaggregates formation inside larger aggregates e much more active in AT as discussed above e which protects the OC associated with these microaggregates, increasing its turnover time and leading to present C sequestration in the afforested ecosystems. These changes in SOM turnover induced by easily available C inputs are widely known as the priming effect (Blagodatsky et al., 2010). As a result of priming, the decomposition of SOM by soil microbes is stimulated by the supply of fresh C as a source of energy (Fontaine et al., 2011; Van Groenigen et al., 2014). Paradoxically, in this study the increase in microbial activity in the macroaggregates led to an increase in C sequestration potential. This indicates that the SOM dynamic in the macroaggregates is controlled, at least in part, by microbial populations e as has been shown by other authors (Baumann et al., 2013). Two mechanisms have been suggested to explain how this priming effect leads to C sequestration: i) the existence of a bank mechanism that regulates nutrients and C sequestration in soil: priming effect is low when nutrient availability is high, allowing sequestration of nutrients and carbon; in contrast, microbes release nutrients from SOM when nutrient availability is low (Fontaine et al., 2011), and ii) the Microbial Efficiency-Matrix Stabilization, which integrates plant litter decomposition with SOM stabilization: labile plants constituents are the dominant source of microbial products. These microbial products of decomposition would thus become the main precursors of stable SOM by promoting aggregation and through strong chemical bonding to the mineral matrix (Cotrufo et al., 2013). Our results, in agreement with both mechanisms, may contribute to a more detailed knowledge of SOM stabilization pathways in macroaggregates of semiarid forest ecosystems. The bank mechanism can be corroborated by the comparison of nutrient availability and SOM accumulation between the AT and S treatments (Table 1). Due to the high nutrient availability, the stimulation of SOM decomposition (priming effect) in AT was low, allowing C sequestration. In addition, the increase in OCs in soil AT was a dominant source for microbial products of decomposition and fungal hyphae: these were the main precursors of SOM stabilization through promotion of macroaggregates formation. Inside

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 10/12

10

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

these macroaggragates, the strong chemical bonding between organic matter and mineral soil particles led to stable microaggregates formation, which was favored by the clay mineralogy and abundance of Ca2þ in the soil matrix. Therefore, our results suggest a pathway for the physicalchemical protection of SOM and C sequestration, through the formation of OC-enriched microaggregates occluded in macroaggregates (Fig. 7). We hypothesize that this is a self-protection system developed by the soil, inside the macroaggregates, to offset the potential effect on SOM decomposition induced by the increase of microbial activity. This process was activated in the soil following the increase in the inputs of fresh plant residues after afforestation with the AT treatment. Our results seem to confirm the hypothesis 5 and show the importance of microbial activity enhancement in the afforestation methodologies employed in semiarid areas with the purpose of C sequestration. By contrast, in the shrubland a negative correlation between basal respiration in macroaggregates and the percentage of micro within macroaggregates was found. This suggests that an aggregates breakdown process occurred, promoted by the microbial activity. The degree of conservation or loss of resources, such as soil nutrients, is considered to be crucial in dryland ecosystem dynamic (Scanlon et al., 2007). We hypothesized that this process was induced by a bank mechanism: due to the reduction of fresh litter inputs, because of degradation of shrubs by external stresses (Mayor et al., 2013), soil nutrient availability decreased (Bautista et al., 2007) e promoting microbial decomposition of stable SOM and organo-mineral bonds, which led to aggregates disruption. Therefore, we consider that the key factor e for the activation of either the self-protection organic carbon system or the aggregate disruption model e was the input of plant litter into the soil, as this regulates the soil food web (Fig. 7).

In addition, these results suggest that the correlations between the microbial activity in macroaggregates (measured as basal respiration) and the percentage of microaggregates within macroaggregates and the OC concentration in microaggregates within macroaggregates could be valid indicators of SOC gains or losses. Positive correlations indicate soil C sequestration and negative correlations indicate soil C emission processes. The advantage of these indicators is that the measurement at a point in time allows the determination of the trend of a dynamic process, not needing long time periods to establish the soil C dynamic in the ecosystem. Evidently, further research under different environmental conditions is needed to validate this hypothesis. The afforestation treatments (mainly AT), led to long-term shifts in the fungal community structure, diversity, and relative abundance of several major fungal taxa, compared to the reference shrubland. However, the soil bacterial communities were not so sensitive to afforestation. The changes in the fungal community structure seem to have been mediated mainly by the shifts in the plant cover, due to the afforestation with pine trees, as suggested by the increase in the relative abundance of Agaricomycotina. This taxon includes ectomycorrhizal fungi that are associated with P. halepensis (Roldan and Albaladejo, 1994). This is in agreement with previous studies which reported that bacterial communities primarily respond to changes in the physical-chemical characteristics of the substrate, while fungal communities are more sensitive to land uses changes (Ros et al., 2006; Macdonald et al., 2009). In our study, the distribution of aggregate-size classes correlated significantly with the community structure of the soil fungi but not with that of the soil bacteria. This agrees with previous surveys that attributed a stronger influence on soil aggregation to fungi, relative to bacteria (De Gryze et al., 2005; Zhang et al., 2012). Particularly, members of the Agaromycotina have been reported to produce

Fig. 7. Diagram showing the soil organic carbon dynamic in: (i) semiarid areas with a well-preserved vegetation (pathway 1), and (ii) semiarid areas which vegetation has been degraded (pathway 2).

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

SBB6179_proof ■ 11 May 2015 ■ 11/12

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12

considerable amounts of aggregate-stabilizing mycelia (Caravaca et al., 2002). We also detected that the fungal (but not the bacterial) community structure was significantly associated with the basal respiration measured in the macroaggregates. This agrees with the notion that fungi metabolize low quality substrates more efficiently (Six et al., 2006) than bacteria do (Holland and Coleman, 1987; Griffith and Bardgett, 2000). Faced with the question “were the changes in fungal community structure, after afforestation, involved in the promotion of soil aggregation?”, our findings suggest an affirmative response. We found the following changes in afforested areas, in comparison to shrubland: 1) an increase in Agaromycotina e a fungal group which can produce considerable amounts of aggregates-stabilizing mycelium (Caravaca et al., 2002), 2) a decrease in Chrytridiomycota e a unicellular fungus unable to produce mycelia, and 3) an increase in species richness. We think that these changes increased the amount and diversity of biotic bonding agents and fostered the formation of macroaggregates. This affirmation may be supported by the significant correlations found between the fungal community structure and aggregate size distribution, and between the fungal community structure and basal respiration in macroaggregates, at the surface layer, in all the treatments tested. 5. Conclusions Qualitative and quantitative changes in OC fractions linked to the shifts in microbial activity within macroaggregates and fungal community structure after afforestation, performed with an initial organic amendment, promoted the formation of macroaggregates occluding and protecting OC-rich microaggregates inside. The accrual and physical-chemical stabilization of OC in this hierarchical structure is a key aspect with respect to increasing or maintaining soil C stocks. Overall, our results revealed that land-use changes involving increases in: 1) sensitive OC fractions, and 2) fungal populations capable of producing large amount of mycelia, are suitable to prevent and mitigate the negative impacts of climatic change on soil quality. In addition, we suggest that the correlations between the basal respiration in macroaggregates and percentage of microaggregates within macroaggregates could be an indicator of the organic carbon dynamic in the soil. Further research is needed to validate this indicator. Acknowledgments This research was found by the Spanish Research, Development and Innovation Plan IþDþI 2008-2011 (Project AGL2010-20941). €lbl and Markus Steffens from We want to thanks Drs. Angelika Ko the Research Department Ecology and Ecosystem Management (Techische Munchen Universitat, Germany) who help us in the RMN results interpretation. Our thanks to the technical staff of Soil and Water Conservation Group from CEBAS-CSIC, who help us in the laboratory and field work. References Abiven, S., Menasseri, S., Angers, D.A., Leterme, P., 2007. Dynamics of aggregate stability and biological binding agents during decomposition of organic materials. European Journal of Soil Science 58, 239e247. Amelung, W., Zech, W., 1999. Minimisation of organic matter disruption during particle-size fractionation of grassland epipedons. Geoderma 92, 73e85. Barber, R.G., Romero, D., 1994. Effects of bulldozer and chain clearing on soil properties and crop yields. Soil Science Society of America Journal 58, 1768e1775. Bartoli, F., Burtin, G., Guerif, J., 1992. Influence of organic matter on aggregation in oxisols rich in gibbsite or in goethite. II. Clay dispersion, aggregate strength and water-stability. Geoderma 54, 259e274.

11


ndez, T., García, C., 2007. The long-term effects of the Bastida, F., Moreno, J.L., Herna management of a forest soil on its carbon content, microbial biomass and activity under a semi-arid climate. Applied Soil Ecology 37, 53e62. Baumann, K., Dignac, M.F., Rumple, C., Bardoux, G., Sarr, A., Steffens, M., Maron, P.A., 2013. Soil microbial diversity affects soil organic matter decomposition in a silty grassland soil. Biogeochemistry 144, 201e212. Bautista, S., Mayor, A.G., Bourakhouadar, J., Bellot, J., 2007. Plant spatial pattern predicts hillslope runoff and erosion in a semiarid Mediterranean landscape. Ecosystems 10, 987e998. Bhattacharyya, R., Tuti, M.D., Kundu, S., Bisht, J.K., Bhatt, J.C., 2012. Conservation tillage impacts on soil aggregation and carbon pools in a sandy clay loam soil of the Indian Himalayas. Soil Science Society of America Journal 76, 617e627. Bibby, K., Viau, E., Peccia, J., 2010. Pyrosequencing of the 16S rRNA gene to reveal bacterial pathogen diversity in biosolids. Water Research 44, 4252e4260. Blagodatsky, S., Blagodatskaya, E., Yuyukina, T., Kuzyakov, Y., 2010. Model of apparent and real priming effects: linking microbial activity with soil organic matter decomposition. Soil Biology & Biochemistry 42, 1275e1283. Borneman, J., Hartin, R.J., 2000. PCR primers that amplify fungal rRNA genes from environmental samples. Applied and Environmental Microbiology 66, 4356e4360. Cao, S., Tian, T., Chen, L., Dong, X., Tu, X., Wang, G., 2010. Damage caused to the environment by reforestation policies in arid and semiarid areas of China. Ambio 39, 279e283. Caporaso, J.G., et al., 2010. QIIME allows analysis of high-throughput community sequencing data. Nature Methods 7, 335e336.
ndez, M.T., Roldan, A., 2002. Aggregate stability Caravaca, F., García, C., Herna changes after organic amendment and mycorrhizal inoculation in the afforestation of a semiarid site with Pinus halepensis. Applied Soil Ecology 19, 199e208. Cerli, C., Celi, L., Kalbitz, K., Guggenberger, G., Kaiser, K., 2012. Separation of light and heavy organic matter fractions in soil-Testing for proper density cut-off and dispersion level. Geoderma 170, 403e416. Chen, C.R., Condron, L.M., Xu, Z.H., 2008. Impacts of grassland afforestation with coniferous trees on soil phosphorus dynamics and associated microbial processes: a review. Forest Ecology and Management 255, 396e409. Clark, G.J., Sale, P.W.G., Tang, C., 2009. Organic amendments initiate the formation and stabilisation of macroaggregates in a high clay sodic soil. Australian Journal of Soil Research 47, 770e780. Costa, R., Gotz, M., Mrotzek, N., Lottmann, J., Berg, G., Smalla, K., 2006. Effects of site and plant species on rhizosphere community structure as revealed by molecular analysis of microbial guilds. FEMS Microbiology Ecology 56, 236e249. Cotrufo, M.F., Wallenstein, M.D., Boot, C.M., Denef, K., Paul, E., 2013. The microbial efficiency-matrix stabilization (MEMS) framework integrates plant litter decomposition with soil organic matter stabilization: do labile plant inputs form stable soil organic matter? Global Change Biology 19, 988e995. Daynes, C.D., Field, D.J., Saleeba, J.A., Cole, M.A., McGee, P.A., 2013. Development and stabilisation of soil structure via interactions between organic matter arbuscular mycorrhizal fungi and plant roots. Soil Biology & Biochemistry 57, 683e694. Debosz, K., Petersen, S.O., Kure, L.K., Ambus, P., 2002. Evaluating effects of sewage sludge and household compost on soil physical, chemical and microbiological properties. Applied Soil Ecology 19, 237e248. Degens, B.P., 1997. Macro-aggregation of soils by biological bonding and binding mechanisms and the factors affecting these: a review. Australian Journal of Soil Research 35, 431e459. De Gryze, S., Six, J., Paustian, K., Morris, S.J., Paul, E., Merckx, R., 2004. Soil organic carbon pool changes followings land-use conversion. Global Change Biology 10, 1120e1132. De Gryze, S., Six, J., Brits, C., Merckx, R., 2005. A quantification of short-term macroaggregate dynamics: influences of wheat residue input and texture. Soil Biology & Biochemistry 37, 55e66. Denef, K.J., Six, J., Merckx, R., Paustian, K., 2004. Carbon sequestration in microaggregates of no-tillage soils with different clay mineralogy. Soil Science Society of America Journal 68, 1935e1944. Díaz, E., Roldan, A., Lax, A., Albaladejo, J., 1994. Formation of stable aggregates in degraded soils by amendment with urban refuse and peat. Geoderma 63, 277e288. Edgar, R.C., 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26, 2460e2461. Edwards, A.P., Bremner, J.M., 1967. Microaggregates in soils. Journal of Soil Science 18, 64e73. Elliott, E.T., 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native and cultivated soils. Soil Science Society of America Journal 50, 627e633. Elliott, E.T., Palm, C.A., Reuss, D.E., Monz, C.A., 1991. Organic matter contained in soil aggregates from a tropical chronosequence: correction for sand and light fraction. Agriculture Ecosystems & Environment 34, 443e451. FAO, 2006. Guidelines for Soil Description. Food and Agriculture Organization of the United Nations, Rome. Fontaine, S., Henault, C., Aamor, A., Bdioui, N., Bloor, J.M.G., Maire, V., Mary, B., Revaillot, S., Maron, P.A., 2011. Fungi mediate long term sequestration of carbon and nitrogen in soil through their priming effect. Soil Biology & Biochemistry 43, 86e96.
ndez, T., Albaladejo, J., Castillo, V., Roldan, A., 1998. Revegetation in García, C., Herna semiarid zones: influence of terracing and organic refuse on microbial activity. Soil Science Society of America Journal 670e676. Garcia-Franco, N., Wiesmeier, M., Goberna, M., Martínez-Mena, M., Albaladejo, J., 2014. Carbon dynamics after afforestation of semiarid shrublands: implications of site preparation techniques. Forest Ecology and Management 319, 107e115.

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

SBB6179_proof ■ 11 May 2015 ■ 12/12

12

N. Garcia-Franco et al. / Soil Biology & Biochemistry xxx (2015) 1e12


nchez, J., Pascual, J.A., García, C., 2007. Pinus halepensis Mill. PlanGoberna, M., Sa tations did not restore organic carbon, microbial biomass and activity levels in a semi-arid Mediterranean soil. Applied Soil Ecology 36, 107e115. Golchin, A., Oades, J.M., Skjemstad, J.O., Clarke, P., 1994. Study of free and occluded particulate organic matter in soils by solid state 13C P/MAS NMR spectroscopy and scanning electron microscopy. Australian Journal of Soil Research 32, 285e309. Griffith, G.S., Bardgett, R.D., 2000. Influence of resource unit distribution and quality on the activity of soil fungi in a particulate medium. New Phytologist 148, 143e151. Haas, B.J., Gevers, D., Earl, A.M., Feldgarden, M., Giannoukos, G., Ciulla, D., Tabbaa, D.,
, B., DeSantis, T.Z., The Human MicroHighlander, S.K., Sodergren, E., Methe biome Consortium, Petrosino, J.F., Knight, R., Birren, B.W., 2011. Chimeric 16S rRNA sequence formation and detection in Sanger and 454-pyrosequenced PCR amplicons. Genome Research 21, 494e504. Hamady, M., Lozupone, C., Knight, R., 2010. Fast UniFrac: facilitating highthroughput phylogenetic analyses of microbial communities including analysis of pyrosequencing and PhyloChip data. The ISME Journal 4, 17e27. Helfrich, M., Ludwig, B., Buurman, P., Flessa, H., 2006. Effect of land use on the composition of soil organic matter in density and aggregate fractions as revealed by solid-state 13C NMR spectroscopy. Geoderma 136, 331e341. Helfrich, M., Ludwig, B., Potthoff, M., Flessa, H., 2008. Effect of litter quality and soil fungi on macroaggregate dynamics and associated partitioning of litter carbon and nitrogen. Soil Biology & Biochemistry 40, 1823e1835. Holland, E.A., Coleman, D.C., 1987. Litter placement effects on microbial and organic matter dynamics in an agroecosystem. Ecology 68, 425e433. IPCC (Intergovernmental Panel on Climate Change), 2007. Fourth Assessment Report: Climate Change. Jastrow, J.D., 1996. Soil aggregate formation and the accrual of particulate and mineral eassociated organic matter. Soil Biology & Biochemistry 28, 665e676. Jastrow, J.D., Amonette, J.E., Bailey, V.L., 2007. Mechanisms controlling soil carbon turnover and their potential application for enhancing carbon sequestration. Climatic Change 80, 5e23. Khale, P., Baum, C., Boelcke, B., 2005. Effect of afforestation on soil properties and mycorrhizal formation. Pedosphere 15, 740e760. €gel-Knabner, I., 2002. The macromolecular organic composition of plant and Ko microbial residues as inputs to soil organic matter. Soil Biology & Biochemistry 34, 139e162. re, J., Angers, D.A., Pare
, D., 2010. Carbon accumulation in agricultural soils Laganie after afforestation: a meta-analysis. Global Change Biology 16, 439e453. Lucas, S.T., D'Angelo, E.M., Williams, M.A., 2014. Improving soil structure by promoting fungal abundance with organic soil amendments. Applied Soil Ecology 75, 13e23. Macdonald, C.A., Thomas, N., Robinson, L., Tate, K.R., Ross, D.J., Dando, J., Singh, B.K., 2009. Physiological, biochemical and molecular responses of the soil microbial community after afforestation of pastures with Pinus radiate. Soil Biology & Biochemistry 41, 1642e1651. Martens, D.A., Frankenberger Jr., W.T., 1992. Decomposition of bacterial saccharides in soil and their influence on soil structure. Biology and Fertility of Soils 13, 65e73. Mao, R., Zeng, D.H., 2010. Changes in soil particulate organic matter, microbial biomass, and activity following afforestation of marginal agricultural lands in a semi-arid area of Northeast China. Journal of Environmental Management 46, 110e116.
fi, S., Bautista, S., Rodríguez, F., Cartení, F., Rietkerk, M., 2013. Mayor, A.G., Ke Feedbacks between vegetation pattern and resource loss dramatically decrease ecosystem resilience and restoration potential in a simple dryland model. Landscape Ecology 28, 931e942. Nannipieri, P., Grego, S., Ceccanti, B., 1990. Ecological significance of the biological activity in soil. In: Bollag, J.M., Stotzky, G. (Eds.), Soil Biochemistry. Marcel Dekker Inc., New York, pp. 293e355. Oades, J.M., 1984. Soil organic matter and structural stability: mechanisms and implications for management. Plant and Soil 76, 319e337. Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H., 2013. In: . Vegan: Community Ecology Package. R Package Version 2, pp. 0e9.

Querejeta, J.I., Roldan, A., Albaladejo, J., Castillo, V., 1998. The role of mycorrhizae, site preparation, and organic amendment in the afforestation of a semi-arid Mediterranean site with Pinus halepensis. Forest Science 44, 203e211. Querejeta, J.I., Rold
an, A., Albaladejo, J., Castillo, V., 2000. Soil physical properties and moisture content affected by site preparation in the afforestation of a semiarid rangeland. Soil Science Society of America Journal 64, 2087e2096. Roldan, A., Albaladejo, J., 1994. Effect of mycorrhizal inoculation and soil restoration on the growth of Pinus halepensis seedlings in a semiarid soil. Biology and Fertility of Soils 18, 143e149. Roldan, A., Querejeta, I., Albaladejo, J., Castillo, V., 1996. Survival and growth of Pinus halepensis Miller seedlings in a semi-arid environment after forest soil transfer, terracing and organic amendments. Annals of Forest Science 53, 1099e1112. Ros, M., Klammer, S., Knapp, B., Aichberger, K., Insam, H., 2006. Long-term effects of compost amendment of soil on functional and structural diversity and microbial activity. Soil Use and Management 22, 209e218. Scanlon, T.M., Caylor, K.K., Levin, S.A., Rodriguez-Iturbe, I., 2007. Positive feedbacks promote power-law clustering of Kalahari vegetation. Nature 449, 209e212. Siddiky, R.K., Kohler, J., Cosme, M., Rillig, M.C., 2012. Soil biota effects on soil structure: interactions between arbuscular mycorrhizal fungal mycelium and collembola. Soil Biology & Biochemistry 50, 33e39. Six, J., Elliott, E.T., Paustian, K., 2000a. Soil macroaggregate turnover and microaggregate formation: a mechanism for C sequestration under no-tillage agriculture. Soil Biology & Biochemistry 32, 2099e2103. Six, J., Paustian, K., Elliot, E.T., Combrink, C., 2000b. Soil structure and organic matter: I distribution of aggregate-size classes and aggregate-associated carbon. Soil Science Society of America Journal 64, 681e689. Six, J., Callewaert, P., Lenders, S., De Gryze, S., Morris, S.J., Gregorich, E.G., Paul, E.A., Paustian, K., 2002. Measuring and understanding carbon storage in afforested soils by physical fractionation. Soil Science Society of America Journal 66, 1981e1987. Six, J., Bossuyt, H., Degryze, S., Denef, K., 2004. A history of research on the link between (micro) aggregates, soil biota, and soil organic matter dynamics. Soil & Tillage Research 79, 7e31. Six, J., Frey, S.D., Thiet, R.K., Batten, K.M., 2006. Bacterial and fungal contribution to carbon sequestration in agroecosystems. Soil Science Society of America Journal 70, 555e569. € gel-Knabner, I., 2009. Alteration of soil organic matter Steffens, M., Kolbl, A., Ko pools and aggregation in semi-arid steppe topsoils as driven by organic matter input. European Journal of Soil Science 60, 198e212. Van Groenigen, K.J., Qi, X., Osenberg, C.W., Luo, Y., Hungate, B.A., 2014. Faster decomposition under increased atmospheric CO2 limits soil carbon storage. Science 344, 508e509. Vanhala, P., Tamminen, P., Fritze, H., 2005. Relationship between basal soil respiration rate, tree stand and soil characteristics in boreal forests. Environmental Monitoring and Assessment 1, 85e92. Wang, Q., Garrity, G.M., Tiedje, J.M., Cole, J.R., 2007. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology 73, 5261e5267. Wei, X., Li, X., Jia, X., Shao, M., 2013. Accumulation of soil organic carbon in aggregates after afforestation on abandoned farmland. Biology and Fertility of Soils 49, 637e646. Wiesmeier, M., Dick, D.P., Rumpel, C., Dalmolin, R.S.D., Hilscher, A., Knicker, H., 2009. Depletion of soil organic carbon and nitrogen under Pinus taeda plantations in Southern Brazilian grasslands (Campos). European Journal of Soil Science 60, 347e359. Young, C.C., Burghoff, R.L., Keim, L.G., Minak-Bernero, V., Lute, J.R., Hinton, S.M., 1993. Polyvinylpyrrolidone-agarose gel electrophoresis purification of polymerase chain reaction-amplifiable DNA from soils. Applied and Environmental Microbiology 59, 1972e1974. Zhang, S., Li, Q., Zhang, X., Wei, K., Chen, L., Liang, W., 2012. Effects of conservation tillage on soil aggregation and aggregate binding agents in black soil of Northeast China. Soil & Tillage Research 124, 196e202. Zimmermann, M., Leifeld, J., Schmidt, N.W.I., Smith, P., Fuhrer, J., 2007. Measured soil organic matter fractions can be related to pools in the RothC model. European Journal of Soil Science 58, 658e667.

Please cite this article in press as: Garcia-Franco, N., et al., Changes in soil aggregation and microbial community structure control carbon sequestration after afforestation of semiarid shrublands, Soil Biology & Biochemistry (2015), http://dx.doi.org/10.1016/j.soilbio.2015.04.012

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108