The ISME Journal (20 13 ), 1 -1 0© 20 13 International Society for Microbial Ecology All rights reserved 1751-7362 /13

www.nature.com /ism ej

ORIGINAL ARTICLEChanges in assembly processes in soil bacterial communities following a wildfire disturbanceScott Ferrenberg^, Sean P O’NeilP ’̂ , Joseph E Knelman^’̂ , Bryan Todd^, Sam Duggan^, Daniel Bradley^, Taylor Robinson^, Steven K Schmidt^, Alan R Townsend^’̂ ,Mark W Williams^’̂ , Cory C Cleveland"^, Brett A Melbourne^, Lin Jiang'^ and Diana R Nemergut^ ®^Department o f Ecology and Evolutionary Biology University o f Colorado, Boulder, CO, USA; ^Institute o f Arctic and Alpine Besearch, University o f Colorado, Boulder, CO, USA; ^Department o f Geography, University o f Colorado, Boulder, CO, USA; ‘̂ Department o f Ecosystem and Conservation Sciences, University o f Montana, Missoula, MT, USA; ^School o f Biology, Georgia Institute o f Technology, Atlanta, GA, USA and ‘̂ Environmental Studies Program, University o f Colorado, Boulder, CO, USA

Although recent work has shown that both deterministic and stochastic processes are important in structuring microbiai communities, the factors that affect the reiative contributions of niche and neutrai processes are pooriy understood. The macrobioiogicai iiterature indicates that ecoiogicai disturbances can infiuence assembiy processes. Thus, we sampied bacteriai communities at 4 and 16 weeks foiiowing a wiidfire and used nuii deviation anaiysis to examine the roie that time since disturbance has in community assembiy. Fire dramaticaiiy aitered bacteriai community structure and diversity as weii as soii chemistry for both time-points. Community structure shifted between 4 and 16 weeks for both burned and unburned communities. Community assembiy in burned sites 4 weeks after fire was significantiy more stochastic than in unburned sites. After 16 weeks, however, burned communities were significantiy iess stochastic than unburned communities. Thus, we propose a three-phase modei featuring shifts in the reiative importance of niche and neutrai processes as a function of time since disturbance. Because neutrai processes are characterized by a decoupiing between environmentai parameters and community structure, we hypothesize that a better understanding of community assembiy may be important in determining where and when detaiied studies of community composition are vaiuabie for predicting ecosystem function.The ISME Journal advance online publication, 14 February 2013; doi:10.1038/ismej.2013.11 Subject Category: Microbiai popuiation and community ecoiogyKeywords: niche vs neutrai processes; community assembly: 16S rRNA gene pyrosequencing; Betaproteobacteria; Firmicutes; beta diversity

IntroductionOver the last few decades, phylogenetic approaches have revealed that microhes exhibit hiogeographic patterns in diversity and distrihution (Martiny et ah, 2006; Hanson et ah, 2012) which often mirror those observed for macro-organisms (Langenheder and Prosser, 2008; Langenheder et ah, 2010). For example, ample evidence suggests that different ecosystems host distinct types of microhes (Lozupone and Knight, 2007; Nemergut et ah, 2011), and that these community differences likely reflect selection [sensu Vellend, 2010) acting on trait differences between suites of organisms. Indeed,

C orrespondence: DR Nem ergut, Institu te of Arctic and A lpine Research, Environm ental Studies Program, U niversity of Colorado, Boulder, CO 80309-0450, USA.E-mail: D iana.Nemergut@ colorado.eduR eceived 1 M ay 2012; revised 28 Decem ber 2012; accepted 9 January 2013

evidence supports the role of environmental para­meters, including pH, salinity and the abundance and quality of carbon (C) in structuring microbial communities (Fierer and Jackson, 2006; Lozupone and Knight, 2007; Logue and Lindstrom, 2010; Nemergut et ah, 2010). Community assembly pro­cesses driven by environmental parameters acting on traits are often called ‘niche-based’ (the term we use here) hut they have also been referred to as ‘habitat filters’ and ‘deterministic processes’ in the literature.

However, recent work suggests that ‘historical filters’ or stochastic processes also affect microbial hiogeography (Martiny et ah, 2006). In particular, there is increasing evidence that dispersal lim ita­tions may have a more important role in structuring microbial communities than previously thought (Telford and Vandvik, 2006; Peay et ah, 2010; Chytry et ah, 2012). Indeed, the neutral theory of biodiversity (Bell, 2001; Huhhell, 2001) in which dispersal is a key determinant of community

W ildfire sh ifts a ssem b iy p r o c esse s in a soli b acteriai com m unityS Ferrenberg et at

structure, has been shown to explain a significant portion of bacterial community variation in a variety of systems ranging from tree hole aquatic habitats to wastewater treatment facilities (Sloan et ah, 2006; Woodcock et ah, 2007; Ofiteru et ah, 2010; Caruso et ah, 2011; Langenheder and Szekely, 2011). Although the relative contribution of niche and neutral processes in determining microbial commu­nity structure may vary across systems, evidence is mounting that both can he important (Ostman et al., 2009; Ofiteru et a l, 2010).

Results suggesting that both dispersal and selec­tion can influence microbial community assembly raise an important question: what regulatesthe relative role o f niche vs neutral processes in structuring microbiai communities^ Work from macrohial systems suggests that a suite of factors including ecosystem productivity, metacommunity (defined here as a set of commu­nities linked by dispersing and interacting taxa) diversity, and dispersal rates are important in the relative balance of these assembly processes (Chase, 2003). Additionally, empirical studies have demonstrated that disturbance can cause an increase in the importance of niche-based processes in structuring communities (Chase, 2007; Jiang and Patel, 2008). Yet, other research suggests that disturbance may promote neutral processes (Didham et al., 2005; Didham and Norton, 2006). Although the specifics of community assembly in response to disturbance may vary with the type and intensity of disturbance, as well as the ecosystem examined, disturbance events frequently kill or severely impact many members of a community. This can ‘reset’ assembly processes and may create temporal gradients that provide excellent opportu­nities for examining general rules about community assembly.

Here, we examined bacterial community assembly processes in response to a wildfire. Fires are ecologically im portant disturbances (Bond et a!.,2005) and their effects on plant and animal communities as well as soil hiogeochemistry have been widely studied (Certini, 2005; Wang and Kemhall, 2005; Ferrenberg et al., 2006). Recent work has also shown that fire induces microbial commu­nity shifts characterized by an increase in the relative abundance of Firmicutes and/or (3-proteo- hacteria and an increase in the ratio of bacteria to fungi (Yeager et ai., 2005; Smith et ai., 2008; Waldrop and Harden, 2008; Barcenas-Moreno et ai., 2011). However, the relative role of niche vs neutral assembly processes in driving these com­munity shifts is unknown. Fire-based disturbance can lead to major shifts in a variety of environmental parameters that are likely to have large direct and indirect effects on the soil microbial community through niche-based processes. For example, fires typically result in an ephemeral pulse of ammonium (NH4t ), creation of a reactive charcoal layer, and subsequent changes in pH (Peitikainen et al., 2000;

DeLuca and Sala, 2006; Wardle et al., 1997, 1998). On the other hand, because fire causes large reductions in the standing soil microbial biomass (Hart et ai., 2005; Wang et ai., 2012), dispersal, which can he largely stochastic, may lead to an increase in the relative importance of neutral processes in early community assembly.

In this study, we used pyrosequencing of bacterial 16S rRNA genes and null deviation analysis (Chase and Myers, 2011) to examine changes in bacterial community assembly processes following a wildfire that killed all vegetation and consumed the surface litter layer of a conifer forest. We sampled the bacterial community and chemistry of soils from a burned site and from an adjacent unhurned forest stand at 4 and 16 weeks after the fire. On the basis of previous studies, we expected that burned soils would demonstrate an increase in soil pH and ammonium pools, a decrease in soil organic matter and alpha (local) and gamma (regio­nal) diversity, and shifts in bacterial community composition. Given the importance of dispersal in early recovery processes, we hypothesized that the burned communities at 4 weeks would show a greater relative importance of neutral processes than unhurned communities. By 16 weeks, we hypothe­sized that niche-based processes, driven by the major shifts in soil chemistry, would become more important for microbial community assembly in the burned sites. Our data yielded insights into micro­bial community assembly following disturbance, which may he im portant for a better understanding of the relationships between assembly processes, microbial community structure and ecosystem function.

Materials and methodsstudy site and sample coiiection A total of 100 samples, 25 each from burned and unhurned soils, collected at both 4 (October 2010) and 16 weeks postfire (January 2011) were analyzed in this study. Soils were sampled roughly 1 m from the base of living (unhurned) and dead (burned) tree trunks near the southeastern edge of the Fourmile Fire (40.039N, 105.391W), that was ignited on 6 September 2010 on the eastern slope of the Colorado Front Range, Boulder County, CO, USA. Forests were dominated by ponderosa pines [Pinus ponderosa scopuiorum) and Douglas firs [Pseudot- sugae menziesii giauca) on similar northeastern aspects, between 2100-2285 m asl. The climate, fire history and soils of these forests were described by Schoennagel et ai., (2011) and Vehlen et ai., (2000). Unhurned and burned sites were ~ 300 m apart with each site roughly 150 m from the dividing fire-line between burned and unhurned forest. Trees in both treatments were located w ithin a 650 m^ plot w ith a minimum of 3 m and a maximum of 25 m separating individual trees; soils were collected from under the

The ISME Jo u rn a l

W ildfire sh ifts a ssem b iy p r o c esse s in a soli bacteriai com m unityS Ferrenberg et al

same trees on both dates. Samples were a composite of three 130.5 cm^ cores from the top 5 cm of mineral soil with litter and visible organic material excluded, and were transported on ice, sieved through 2 mm mesh, and stored at — 80 °C for DNA extraction or at + 4 °C for hiogeochemical analyses.

D N A e x tr a c tio n a n d p y r o s e q u e n c in g o f p a r t ia l 16 S rR N A g e n e sDNA was isolated using the MO BIG Power Soil DNA Extraction kit (MG BIO Lahoratories, Carlshad, CA, USA), and was processed as described in Nemergut e t a l. , (2 0 1 0 ) and Knelman e t a l., (2 0 1 2 ). A fragment of the 16S rRNA gene encoding the VI- V2 region was amplified using modified primers of 27F and 338R adapted for Titanium chemistry (454 Life Sciences, Bradford, CT, USA). PCR reactions were performed in triplicate with 1 0 p i of sterile H 2 G, 1 0 p i of 5 PRIME hot master mix (5 PRIME, Gaithersburg, MD, USA), 2 p i (5pM ) of the reverse primer, 1 p i ( 1 0 piu) of the forward primer and 2 p i of the sample DNA. Samples were denatured for 3 min at 94 °C followed by 25 cycles at 94 °C for 45 s, 50 °C for 30s, 72 °C for 90s and a final elongation step at 70 °C for 1 0 min. Three replicate PCR products were quantified, pooled and cleaned using MG BIG UltraClean-htp PCR Clean-up kits and 16S rRNA gene amplicons were sent to the Environmental Genomics Core Facility (Engencore) at University of South Carolina for 454 Life Sciences CS FLX Titanium pyrosequencing.

Sequence AnalysisPyrosequencing data were screened using the QIIME (version 1.2.1) toolkit (Caporaso et al., 2010) with the following parameters: quality score >25,sequence length >200 and <400, maximum hom o­polymer of 6, 0 maximum ambiguous bases and 0 mismatched bases in the primer. GTUs were denoised using Denoiser (Reeder and Knight, 2010) and were picked at the 97% identity level using UCLUST (Edgar, 2010) in QIIME. GTUs were randomly suhsampled in QIIME so each library contained 1142 sequences (the fewest in a single sample). Quality data were not obtained from five samples which were excluded from analyses. The taxonomic identity of GTUs was assigned using RDP in QIIME, and QIIME was used to generate a weighted UniFrac distance matrix (Lozupone and Knight, 2005; Lozupone et al., 2006) and a Bray- Curtis distance matrix (Bray and Curtis, 1957). QIIME was also used to generate a and y diversity metrics (GTU richness (unique GTUs), Shannon diversity, phylogenetic diversity, Pielou’s evenness and dominance (probability of randomly sampling two individuals of the same GTU, Caparaso et al.,2010)). All sequencing data have been deposited in the MC-RAST database (http://m etagenom ics.anl. gov/).

Soil analysis and microbial biomass Soil moisture, pH and total C and nitrogen (N) were m easured on samples collected on both dates. Soil moisture was determined with the gravimetric m ethod after drying soils at 60 °C for 48 h. Soil pH was measured using a 1:5 ratio of soil to de-ionized H 2 G. Total C and N were determined by grinding and combustion in an elemental analyzer as described by (Knelman et al., 2 0 1 2 ). NH4/ , dissolved organic N (DGN), dissolved organic C (DGC), and microbial biomass were m easured for only the 16 week samples by adding 40 ml of 0.5 m K2 SO4 to 10 g of soil, shaking the mix for 1 h, and filtering through W hatman no .l paper (Whatman Incorporated, Flor- ham Park, NJ, USA). N H / concentrations were determined using the sodium salicylate method and absorbance at 650 nm on a microplate reader (Mulvaney, 1996). DGC and DGN were determined using a TIC/TGC analyzer. For DGC, biomass C = EC/kEC, where EC = extractahle C from soil, and kEC (extractahle C from microbial biomass) was esti­mated at 0.45 (Beck et al., 1997). DGN was determined by Kjeldahl Digestion of 2 0 ml of extract; N = EN/kEN where kEN was estimated at 0.54 (Brookes et al., 1985). Microbial C and N pools were calculated as the difference between DGC and DGN from non-fumigated and 5-day chloroform fumi­gated soils (Brookes et al., 1985; Beck et al., 1997).

D a ta A n a ly s isBurned and unhurned soil chemistry, a diversity measures, and weighted UniFrac matrices were compared with SAS-JMP 9.0.0 (JMP 2 0 1 1 ) using one-way analysis of variance followed by Tukey’s HSD means comparisons (Kruskal-Wallace tests followed by Steel-Dwass means comparisons when test assumptions were not met). To avoid violating assumptions of sample independence, sample GTU dissimilarities (p-diversity, calculated as mean group Bray-Curtis dissimilarity) were compared using ADGNIS followed by the perm utation method of ‘hetadisper’ in the Vegan package for the R platform (Gksanen e t a l., 2 0 1 1 ; Team RDC 2 0 1 1 ). Because the hypotheses tested here focus on the difference between burned and unhurned soils and less on differences between geographically collo­cated samples, and because samples were removed from the site, repeated measures analyses or paired tests were not used. Variables that were measured only in January (NH4t, DGC/DGN, microbial bio­mass) were compared via /-tests or M ann-W hitney U-tests. W hen effective, log or logio transformations were applied to meet test assumptions. Gur figures and tables contain hack-transformed values with statistical comparisons based on transformed data as noted.

PERMANGVA and non-metric m ultidim ensional scaling were completed on the Bray-Curtis distance matrix in PC-GRD (McCune and Mefford, 2 0 1 1 ) and used to compare community composition in burned

The ISME J o u rn a l

W ildfire sh ifts a ssem b ly p r o c esse s in a soil bacterial com m unityS Ferrenberg et al

and unburned samples. Mantel tests for correlations between environmental factors and the soil bacterial commnnity of bnrned and nnbnrned soils at 4 and 16 weeks were calcnlated in PC-ORD. Edaphic data are available in the MIMARKS database (Yilmaz et a l, 2011).

We nsed the nnll deviation approach (Chase and Myers, 2011) to examine bacterial commnnity assembly. This techniqne nses a nnll model to create stochastically assembled commnnities from the regional species pool to determine the degree to which observed p diversity patterns deviate from stochastic assembly. The nnll deviation approach disentangles variation in commnnity compositional dissimilarity across sites from variation dne to changes in oc (local) and y (regional) diversity (Chase and Myers, 2011). This approach assesses changes in p diversity that resnlt from the relative inflnence of niche and nentral processes and not from changes in oc diversity. We measnred the nnll deviation as the relative difference of the observed p diversity from the nnll-model p diversity, (Pobs-Pnuii)/ Pnuib where p diversity was measnred as Sorenson- Czekanowski dissimilarity. For each sample, the expected p diversity nnder the nnll model was calcnlated from 10 000 stochastically assembled commnnities. Camma diversity for each nnll model was calcnlated from each w ithin-treatm ent (bnrned vs nnbnrned at 4 and 16 weeks) species pool. As this analysis reqnires presence-absence data and does not weight species by their abnndance (nnlike the non-metric m nltidim ensional scaling and Mantel tests described above) it is sensitive to noise from rare species. Therefore, taxa w ith very low abnndances (<1% of seqnences per commn­nity) were removed from pyroseqnencing data before analyses (Ofitern et al., 2010). To test for treatment differences in the nnll deviation we condncted perm ntation tests by first randomly permnting treatment labels, then resimnlating nnll models and recalcnlating nnll deviations for each of 5000 permntations.

ResultsBacterial Community Diversity and Structure After rarefaction to an eqnal seqnencing depth, we fonnd a total of 4760 nniqne OTUs across all samples (Snpplementary Fignres S i and S2a). Unbnrned soils contained 2596 OTUs at 4 weeks and 2627 OTUs at 16 weeks postfire. Bnrned soils had 1889 OTUs at 4 weeks and 1656 at 16 weeks— 28 and 37% lower y diversity than was observed in nnbnrned soils, respectively. Unbnrned soil had 850 and 1002 nniqne (that is, not fonnd in samples from any other treatment/date) OTUs and bnrned soil had 357 and 273 nniqne OTUs at 4 and 16 weeks, respectively. A total of 698 ‘generalist’ OTUs were fonnd in both bnrned and nnbnrned soils on both dates.

Bnrning significantly rednced oc diversity for both dates (Snpplementary Fignre S2 b-f) regardless of the diversity metric (richness. Shannon, phyloge­netic diversity, evenness or dominance) applied. Alpha diversity w ithin bnrned or nnbnrned soils did not change significantly between sampling dates. Specifically, bnrned soils had an average of 31 and 50% lower OTU richness than nnbnrned soil at 4 and 16 weeks, respectively (Snpplementary Fignre S2b; Fg84 = 37.49, P < 0.0001). The Shannon diversity index of bnrned soil (fonr weeks = 6.12, 16 weeks = 5.46) was significantly lower than that of nnbnrned soil (4 weeks = 7.68, 16 weeks = 7.92) (Snpplementary Fignre S2c, Fg 84 = 26.69, P < 0.0001). Finally, bnrning rednced phylogenetic diversity by 28% at 4 weeks (49.61 vs 68.76) and 42% at 16 weeks (41.99 vs 72.65) (Snpplementary Fignre S2d; Fg8 4 = 33.98, P < 0.0001). OTU evenness was lower in bnrned soil than in nnbnrned at 4 (0.74 vs 0.87) and 16 weeks (0.69 vs 0.89) (Snpplementary Fignre S2e; Fg 8 4 = 22.23, P < 0.0001), while domi­nance was higher in bnrned soils at 4 (0.10 vs 0.02) and 16 weeks (0.09 vs 0.01) (Snpplementary Fignre S2f; Fg 8 4 = 10.17, P < 0.0001). Average p diversity, in this case mean pairwise Bray-Cnrtis dissimilarity and UniFrac distance, was also altered by bnrning (Snpplementary Fignres S2g and h). Fire cansed a significant increase in Bray-Cnrtis dissimilarity at 4 and 16 weeks (PERMANOVA, P = 0.0002), and a significant increase in UniFrac for both dates (PERMANOVA, P = 0.0002).

Bnrned and nnbnrned soils harbored significantly different bacterial commnnities 4 and 16 weeks postfire (PERMANOVA, P < 0.001) and w ithin both treatments between dates (PERMANOVA, P < 0.001). Non-metric m nltidim ensional scaling clnstered samples by treatment and date, w ith bnrned soils displaying a greater spread between samples than for nnbnrned commnnities (Fignre 1), consistent with the observed increases in p diversity.

2D Stress = 0.07

Figure 1 Non-m etric m ultid im ensional scaling o rd ination based on Bray-Curtis d istances show ing the change in bacterial com ­m unity com position and increase in P diversity in bu rn ed soil bacterial com m unities (4 w eek com m unities = gray squares; 16 w eek com m unities = black squares) com pared w ith un b u rn ed soil bacterial com m unities (4 w eek = open triangles; 16 w eek = solid triangles) 4 and 16 weeks after the fire.

The ISME Jo u rn a l

W ildfire sh ifts a ssem b iy p r o c esse s in a soli bacteriai com m unityS Ferrenberg et al

The relative abundances of seven dominant bacter­ial phyla (or subpbyla in the case of Proteobacteria) differed significantly between burned and unburned soils at 4 (Supplementary Figure S3; x^e = 58.07, P < 0 .0 0 0 1 ) and 16 weeks (x^s = 75.72, P < 0 .0 0 0 1 ); and w ithin the burned samples between dates (X̂ 6 = 28.73, P < 0 .0 0 0 1 , Supplementary Figure S3). Proportional phyla and subpbyla abundances in unburned soils were not different between dates (x^ 6 = 1.29, P>0.95). Firmicutes were in low relative abundance in unburned soils for both dates, but dominated burned soils. Betaproteobacteria were in low relative abundance in the 4 week burned samples, but increased at 16 weeks to become the second most abundant taxon.

Soil chemistryBurning reduced soil C and N (F3 _9 s = 36.61, P < 0 .0 0 0 1 ), with 65% less C and 48% less N in burned than unburned samples from 4 weeks, and 69% less C and 60% less N in the burned soil from 16 weeks (Table 1 ). Burning reduced microbial biomass (microbial C; m easured only at 16 weeks) by 77% compared with unburned samples (Supplementary Table Si). The C:N ratio of burned soils was 39% lower than that of unburned soils at 4 weeks, and 24% lower at 16 weeks (F3 _9 s = 107.94, P < 0 .0 0 0 1 ; Table 1 ). Burning increased pH for both sample dates, while pH showed an overall decrease between 4 and 16 weeks regardless of treatment (F3 , 9 7 = 20.5, P < 0 .0 0 0 1 , Table 1 ). Soil moisture was lower in burned soils at 4 weeks (F3 9 3 = 10.38, P < 0 .0 0 0 1 ), with no difference between treatments at 16 weeks (Table 1 ). Burned soils bad greater than a 1 2 -fold increase in mean N H / concentration (only measured at 16 weeks), averaging 3.12|rg^^g in unburned soils and 38.87|rgg^^ in burned soils (U = 623, Z = -6 .0 1 , P < 0 .0 0 0 1 , Table 1 ). Burning did not significantly change DOC, but increased DGN (both m easured only at 16 weeks) by 60% over unburned soil (1 2 4 ,2 4 = —6 .0 , P < 0 .0 0 0 1 , Table 1 ).

degree to which observed (3 diversity patterns deviate from stochastic assembly. A null deviation close to zero suggests that neutral processes are more important in structuring the community, whereas larger positive or negative null deviations suggest that niche-based processes are more impor­tant. After 4 weeks, burned communities deviated significantly less from the stochastic assembly model than unburned communities (permutation test, P = 0 .0 2 ; Figure 2 ). After 16 weeks, however, burned communities (relative null deviation = — 0.17) deviated significantly more from the sto­chastic assembly model than unburned commu­nities (P= 0 .0 0 1 ; Figure 2 ). Importantly, the unburned sites showed a moderate but consistent deviation from the stochastic assembly model (relative null deviation = —0 .1 2 ; Figure 2 ), w ith no significant changes in the null deviation value between the 4 and 16 week samples (P=0.36). Burned sites, by contrast, were significantly more stochastic at 4 weeks than at 16 weeks (P< 0 .0 0 1 ).

Community structure and soil environmental characteristicsGur analysis revealed that the relationship between environmental characteristics and soil microbial community structure varied with sampling date and disturbance. We found no significant correla­tions between environmental characteristics and community structure for the 4 week samples (Table 2 ). For the 16 week burned samples we observed significant correlations between commu­nity composition and pH. By contrast, we observed significant correlations between soil C:N and com­m unity composition for the 16 week unburned soils. Tests of correlations between the bacterial commu­nity and environmental factors yielded similar results regardless of whether weighted or unweighted community metrics were considered (data not shown).

Community Assembly ProcessesThe null deviation approach (Chase and Myers, 2 0 1 1 ) created stochastically assembled communities from the regional species pool to determine the

DiscussionA severe wildfire provided an opportunity to examine the relative roles of niche vs neutral assembly processes in recently disturbed soil

Table 1 C om parison of u n b u rn ed and b u rn ed soil p roperties at 4 and 16 w eeks after a stand-replacing w ildfire

Sam ple Treatm ent % M oisture p H %C % N C:N N H t DOC D O N

4 weeks

16 weeks

U nburnedB urned

U nburnedB urned

P

21.03 (2.07)“ 7.30 (0.12)'’ 5.75 (0.59)“ 0.23 (0.02)“ 26.07 (0.71)“ 9.70 (1.05)'’ 8.00 (0.25)“ 2.03 (O.lOp 0.12 (O.Olp 16.00 (0.39)“ 7.91 (1.32)'’ 6.92 (0.09)“ 7.96 (1.04)“ 0.30 (0.04)“ 26.07 (0.45)“ 8.18 (0.85)'’ 7.34 (0.12)'’ 2.44 (O.OOp 0.12 (0.02)'’ 20.02 (O.OOp < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001

3.12 (0.37) 38.87 (3.48)

< 0.0001

0.34 (0.06) 0.28 (0.04)

ns

0.02 (0.003) 0.06 (0.005)

< 0.0001

Abbreviations: DOC, dissolved organic C; DON, dissolved organic N; ns, not significant.Untransformed means ( ± 1 s.e.), P from analysis of variance with transformed values for variables measure in both sample dates; P from t-test or M ann-W hitney U for variables measured for one sample date. Means followed by different letters represent significant differences from Tukey’s HSD comparisons (P<0.05).

The ISME J o u rn a l

W ildfire sh ifts a ssem b ly p r o c esse s in a soil bacterial com m unityS Ferrenberg et al

>(U

I

On

-0.04

-0.08

- 0.12

-0.16

- 0.2

neutral

B

Table 2 B acterial com m unity struc tu re association w ith env ironm ental factors in u n b u rn ed and bu rn ed soils at 4 and 16 weeks after fire

burned - • unburned

nicheC

4 6 8 10 12 16Weeks since disturbance

Figure 2 Plot show ing the n u ll deviation (Chase and M yers, 2011) of bu rn ed and un h u rn ed com m unities 4 and 16 weeks after the fire. A n u ll dev iation close to zero suggests th a t neu tra l processes are m ore im portan t in structu ring the com m unity, w hereas larger positive or negative n u ll deviations suggest that n iche-based processes are m ore im portant. D ifferent letters ind icate significant differences betw een sam ple dates based on perm utation tests (P<0.05).

bacterial communities. As expected, we found that burning caused substantial changes in soil bacterial diversity (Supplementary Figure S2 ), community structure (Figure 1 , Supplementary Figures S i and S3) and soil chemistry (Table 1 ). We also found that bacterial secondary succession proceeded very rapidly in the postdisturbance landscape, as com­munities from 4 and 16 weeks postburn were significantly different, not only in terms of the OTUs present, but also w ith respect to the phyla/ subphyla proportional abundances (Figure 1 , Supplementary Figures S i and S3). Seasonal effects are known to influence soil microbial community abundances and activities (Monson et al., 2006; Schmidt et al., 2007) and, consistent w ith these observations, we observed a small but significant difference in the bacterial community structure (Figure 1 ) of unburned samples from 4 (fall) and 16 (winter) weeks. As well, pH and soil moisture were different between the two sampling time- points (Table 1 ). However, the magnitude of the differences in burned vs unburned soil community structure, along with significant reductions in y and oc diversity in both sample dates suggest that fire effects (direct or indirect) on microbial community assembly and secondary succession are m uch stronger than seasonal shifts over this time period.

A null deviation value close to zero suggests that community assembly is highly stochastic and neutral processes are more important in structuring the community. Larger positive or negative null deviations suggest that niche-based processes are more important, and environmental filters, for example, could have strong influences on commu­nity assembly. Regardless of disturbance or sam­pling time, null processes were important in structuring soil microbial communities (Figure 2 ).

Treatm ent group Soil factor M antel r F ^O .05

4 w eeks un b u rn ed All factors com bined 0.117 0.175C:N ratio -0 .0 7 1 0.300% N - 0.020 0.531%C - 0.020 0.552pH -0 .1 3 8 0.213H 2Q 0.129 0.155

4 w eeks b u rned All factors com bined 0.116 0.130C:N ratio -0 .0 2 5 0.459% N -0 .0 5 2 0.349%C -0 .0 3 3 0.431pH -0 .0 9 8 0.218H 2Q 0.125 0.125

16 weeks un b u rn ed All factors com bined 0.267 0.059C:N ratio 0.320 0.007*% N 0.121 0.162%C 0.119 0.144pH 0.083 0.226H 2Q 0.246 0.069NHy -0 .0 1 5 0.496DON -0 .0 8 6 0.306DOC -0 .0 7 8 0.337

16 weeks b u rned All factors com bined 0.208 0.109C:N ratio -0 .0 4 8 0.380% N 0.027 0.295%C -0 .0 0 3 0.649pH 0.303 0.013*H 2Q -0 .0 8 2 0.335NH4+ 0.214 0.101DON 0.202 0.071DOC 0.099 0.210

Abbreviations: DOC, dissolved organic C; DON, dissolved organic N.Mantel tests were completed w ith a Bray-Curtis distance matrix for GTU counts and a Euclidean distance matrix for soil factors. GTUs with ^1 0 sequences per sample were included. Significance for each test was determined from 5000 randomized Monte Carlo runs. *indicates significant relationships (P<0.05).

However, our analysis also revealed that fire caused a quantifiable change in assembly processes (that is, the relative importance of niche vs neutral pro­cesses) that shifted w ith time since disturbance. As with the observed changes in diversity and commu­nity structure, these shifts were evident over very short time frames: communities in the soils 4 weeks postburn were shaped by neutral processes (smaller null deviations) significantly more so than unburned communities, while burned communities at 16 weeks were shaped by niche processes (larger null deviations) more than unburned communities.

Interestingly, we also observed an increase in p diversity among postburn communities (Figure 1 , Supplementary Figure S2 e,f). Such increases in p diversity have been interpreted as support for neutral processes in community assembly (Kraft et al., 2007; Chase and Myers, 2 0 1 1 ). However, our null deviation analysis supports that the observed increases in p diversity are due to both increases (4 weeks) and decreases (16 weeks) in neutral processes in burned soils relative to unburned sites (Figure 2 ). The initial increase in p diversity may

The ISME Jo u rn a l

W ildfire sh ifts a ssem b ly p r o c esse s in a soil bacterial com m unityS Ferrenberg et al

thus reflect the stochastic nature of dispersal, and the fact that ‘seed’ microbes from air and precipita­tion vary in both space and time. As succession proceeds, however, onr data suggest that this increase in p diversity reflects an increase in niche-based processes. Other work has demon­strated that fire tends to increase hiogeochemical heterogeneity, likely dne to variation in the severity of the burn across the landscape (for example. Turner et ah, 2007; Hamman et al., 2008). Thus, these increases in environmental variation may be reflected, at least to some degree, in the changes in commnnity composition observed in the bnrned soils, combined with an increase in niche-based processes in shaping commnnities. These shifts in commnnity assembly may also be reflected in the trait differences of the dominant taxa. For instance, spore-forming Firmicutes —which may be more easily dispersed— are most abundant in the 4 week postbnrn soils. As well, Betaproteobacteria, which commonly dominate early snccessional landscapes (Nemergut et al., 2007; Sattin et al., 2009) are more abundant at 16 weeks, when the decrease in soil organic matter (Table 1 ) may select for more oligotrophic taxa.

We performed Mantel tests to examine correla­tions between environmental parameters and com­m nnity structure. We measnred a suite of standard soil chemical parameters (Table 1 ) but did not observe any significant relationships between these variables and commnnity structure, (regardless of whether comparisons were completed w ith abnn­dance based analyses or unweighted analyses), w ithin treatments for the 4 week samples (Table 2 ). For the 16 week samples, p diversity in the bnrned samples was correlated w ith pH while nnbnrned commnnities were correlated w ith soil C:N. Correla­tion coefficients for these relationships were roughly similar, and some have interpreted these as evidence for the amount of variation in commn­nity structure that is explained by niche-based processes. However, as noted by (Anderson et al.,2 0 1 1 ), extreme caution should be taken in interpret­ing these relationships in terms of assembly mechanisms because of the potential for unm ea­sured environmental variation as well as the possibility of spatial structure in environmental parameters. Thus, these analyses provide hypoth­eses about the potential sources of local variation in bacterial commnnity structure, but are not incon­sistent w ith onr nnll deviation analyses.

These observed differences in commnnity assem­bly over very short time scales may reconcile the fact that different researchers have found support for increases in both niche (Chase, 2007; Jiang and Patel, 2008) and nentral (Didham et al., 2005; Didham and Norton, 2006; Leibold and McPeek,2006) processes following disturbance. Indeed, onr results suggest dynamic shifts in commnnity assem­bly processes in postdistnrbance landscapes, lead­ing ns to propose a conceptual model describing

b u rn ed

Time since disturbance

Figure 3 The three hypo thesized phases of com m unity assem bly follow ing disturhance. Phase 1 is characterized hy m ore neu tra l assem bly processes; Phase 2 is m ore n iche-hased and Phase 3 is increasingly m ore neutral.

these changes. Specifically, we hypothesize that time since disturbance features at least three distinct phases in commnnity assembly (Fignre 3). Phase 1

immediately follows disturbance, and is character­ized by a brief increase in the relative role of nentral processes in commnnity assembly, perhaps because stochastic dispersal processes are strongly affecting commnnity structure. This is supported by other work that suggests that ecological equivalence may be more likely immediately following a severe disturbance event or at the onset of primary succession when immigrants face less competition (Leibold and McPeek, 2006). During phase 2 , organisms begin to grow and divide, and niche- based processes in the postdistnrbance landscape act as strong filters on microbial commnnity com­position. This can be characterized by increases in p diversity if the disturbance was heterogeneous at the landscape level, or decreases if it was more homo­genous. Similar results have been observed in other experimental systems as niche-based processes were shown to be important following drought as well as density-independent mortality disturbance events (Chase, 2007; Jiang and Patel, 2008). Finally, over longer time scales (phase 3), the environment becomes less harsh and neutral processes again become more important in shaping community structure. To some degree this supports other work that suggests that neutral processes may dominate community assembly w ithin snccessional stages while niche processes may dominate during transi­tion periods between snccessional stages (Denslow, 1980; Ellner and Fussmann, 2003; Cadotte, 2007).

An important caveat of our data is that we lack an understanding of community assembly from a functional level. It is possible that examining assembly processes using metagenomics or meta- transcriptomics would reveal different patterns in the relative importance of niche vs neutral pro­cesses. For example, Burke et al. (2 0 1 1 ) recently showed that microbial community succession on marine algae displayed functional convergence but

T he ISME J o u rn a l

W ildfire sh ifts a ssem b iy p r o c esse s in a soli b acteriai com m unityS Ferrenberg et al

lacked taxonomic coherence. This suggests a high degree of functional redundancy in microhial com­munities, which may decouple structure and func­tion. If these same processes are at work in the postdistnrbance landscape that we examined, ana­lysis of the 16S rRNA gene data may suggest that communities are assembled hy a larger predom i­nance of neutral processes than is actually the case. Future studies should use both SSU rRNA and metagenomic approaches to examine community assembly processes, as trait-hased approaches may yield deeper understandings of the mechanisms driving assembly.

As well, such trait-hased approaches may he important for guiding approaches for how and where to sample microhial communities to understand ecosystem processes. We now possess the tools to reveal high-resolution details about temporal and structural changes in microhial community structure. As microhial community structure drives function, some argue that there is value in knowing ‘who does what’ to understand and predict ecosystem processes (Zak et a l, 2003; Monson et a l, 2006; Van Der Heijden et al., 2008). However, as mentioned above, many studies reveal that environmental factors are important determinants of microhial community structure (Fierer and Jackson, 2006; Lozupone and Knight, 2007; Logue and Lindstrom, 2 0 1 0 ; Nemergut et a l, 2 0 1 0 ). Also, these same parameters are vital in regulating ecosystem processes (Bonan and Shugart 1989; Paul and Clark, 1996), raising the question: how much added value is provided hy detailed investigations of microhial structure data? Indeed, a better understanding of microhial community assem­bly processes, and where and when they may change in response to disturbances, could he fundamental to understanding links between structure and function. We hypothesize that if communities are largely structured hy neutral processes, then while environ­mental factors will still affect ecosystem processes in these communities hy influencing the physiologies of individual microorganisms, soil communities should exhibit less of a direct link between edaphic factors and processes. In other words, the degree to which niche vs neutral processes guide microhial commu­nity assembly will affect the strength of the relation­ship between environmental factors and ecosystem processes.

AcknowledgementsT his research w as supported , in part, th rough grants from the N ational Science F o u n d ah o n to DN, SS and CC (DEB1545913), DEB1120281 to LJ an d the N iw ot Ridge LTER grant (DEB-1027341) to MW, AT, and SS. T his research u sed the Janus supercom puter, w h ich is suppo rted by th e N ahonal Science F o undahon (aw ard num ber CNS-0821794) an d the U niversity of Colorado Boulder. The Janus supercom puter is a jo in t effort of the U niversity of C olorado Boulder, the U niversity of Colorado D enver an d the N ahonal C enter for A tm ospheric Research. We th an k T im othy Seastedt for com m ents on our m anuscrip t and the U niversity of

C olorado’s D epartm ent of Ecology and Evoluhonary Biology for generous support. We acknow ledge Bob H ead an d Bill N em ergut (WFN) for assisting w ith sam ple collechon and processing. O ur m anuscrip t w as greahy im proved by com ­m ents an d suggeshons from three anonjm ious review ers and the editor.

ReferencesO liveros JC. (2007). VENNY; A n in te rac tiv e too l for

co m p arin g lis ts w ith V enn D iagram sh ttp :/ /b io in - fo g p .c n b .c s ic .e s /to o ls /v e n n y /in d e x .h tm l.

A n d e rso n MJ, C ris t TO, C hase JM, V ellend M, In ouye BD, F rees to n e AL e t al. (2011). N av iga ting th e m u ltip le m ean in g s o f be ta -d iv e rsity : a ro ad m ap for th e p ra c t i­c ing eco log ist. E col L e tt 14: 19-28 .

B eck T, Joergensen G, K an d e le r E, M ak esch in F, N u ss E, O berh o lze r HR e t al. (1997). A n in te r-lab o ra to ry c o m p a riso n of te n d iffe ren t w ays of m easu rin g so il m ic ro b ia l b io m ass C. S o il B io l B io ch em 29: 1023-1032 .

B ell G. (2001). N eu tra l m acroecology. S c ien c e 293: 2413-2418 .

B onan GB, S h u g a rt HH. (1989). E n v iro n m en ta l fac to rs an d eco log ica l p ro cesses in bo rea l fo rests. A n n a B e v E col S y s t 20: 1 -28 .

B ond WJ, W oodw ard FI, M idg ley GF. (2005). T he g lobal d is tr ib u tio n of ecosystem s in a w o rld w ith o u t fire. N e w P h y to l 165: 525 -5 3 8 .

B ray JR, C u rtis JT. (1957). A n o rd in a tio n of th e u p la n d fo rest co m m u n itie s o f so u th e rn W isconsin . E col M onogr 27: 325-349 .

B rookes PC, L an d m an A, P ru d e n G, Je n k in so n DS. (1985). C h lo ro fo rm fu m ig a tio n a n d th e re lease of so il n it ro ­gen: a ra p id d ire c t ex trac tio n m e th o d to m easu re m ic ro b ia l b io m ass n itro g en in so il. S o il B io l B io ch em 17: 8 37 -8 4 2 .

B urke C, T h o m as T, L ew is M, S te inberg P, K jelleberg S. (2011). C om p o sitio n , u n iq u e n e s s a n d v a riab ility o f the e p ip h y tic b ac te ria l co m m u n ity o f th e g reen alga U lva au s tra lis . ISM E J 5: 590 -600 .

B h ce n a s-M o re n o G, G arcfa-O renes F, M ata ix -S o le ra J, M ata ix -B eney to J, B aath E. (2011). S o il m icro b ia l re c o lo n isa tio n after a fire in a M ed ite rran ean forest. B io l E ert S o ils 47: 261 -272 .

C ado tte MW. (2007). C o n c u rre n t n ic h e a n d n e u tra l p ro cesse s in th e c o m p e tit io n -c o lo n iz a tio n m o d e l of sp ec ies co ex is ten ce . P roc B S o c B 274: 2739-2744 .

C aporaso JG, K u czy n sk i J, S to m b au g h J, B ittin g er J, B u sh m an FD, C oste llo EK e t al. (2010). QIIME allow s an a ly s is of h ig h - th ro u g h p u t co m m u n ity seq u en c in g data . N a t M e th o d s 7: 335-336 .

C aruso T, C h an Y, L au MCY, M cK ay CP, P o in tin g SB.(2011). S to ch astic an d d e te rm in is tic p ro ce sse s in te ra c t in th e assem b ly of d e se r t m ic ro b ia l co m m u n itie s on a g lobal scale. ISM E J 5: 1406-1413 .

C ertin i G. (2005). E ffects of fire on p ro p e r tie s of fo rest so ils: a review . O ecologia 143: 1 -10 .

C hase JM, M yers JA. (2011). D isen tan g lin g th e im p o rtan ce of eco log ica l n ic h e s from s to ch astic p ro cesses across scales. P hilo s Trans B S o c L o n d B B io l S c i 366: 2351-2363 .

C hase JM. (2003). C o m m u n ity assem bly : w h e n sh o u ld h is to ry m atter? O ecologia 136: 4 8 9 -4 9 8 .

The ISME Jo u rn a l

W ildfire sh ifts a ssem b iy p r o c esse s in a soli bacteriai com m unityS Ferrenberg et al

C hase JM. (2007). D rough t m ed ia te s th e im p o rtan ce of s to ch astic co m m u n ity assem bly . Proc N a tl A c a d S c i U SA 104: 17430-17434 .

C h y try M, L ososova Z, H o rsak M, U h e r B, Cejka T, D an ih e lk a J e t al. (2012). D isp ersa l lim ita tio n is stronger in co m m u n itie s of m ic ro o rg an ism s th a n m acro o rg an ism s across C en tra l E u ro p ean cities. J B iogeography 39: 1101-1111 .

D eL uca TH, S ala A. (2006). F re q u e n t fire a lte rs n itro g en tran sfo rm a tio n s in p o n d e ro sa p in e s ta n d s of th e In la n d N orthw est. E co logy 87: 2511-2522 .

D en slo w JS. (1980). G ap p o rtio n in g am ong tro p ic a l ra in ­fo rest trees. B io trop ica 12: 4 7 -5 5 .

D id h am RK, N o rto n DA. (2006). W h en are a lte rn a tiv e stab le s ta tes m ore lik e ly to occur? O ikos 113: 357-362 .

D id h am RK, W atts CH, N o rto n DA. (2005). A re system s w ith strong u n d e rly in g ab io tic reg im es m ore lik e ly to ex h ib it a lte rn a tiv e stab le sta tes? O ikos 110: 4 0 9 -4 1 6 .

E dgar RC. (2010). S earch a n d c lu s te rin g o rd ers of m ag n itu d e fas te r th a n BLAST. B io in fo rm a tic s 26: 2460-2461 .

E lln e r SP, F u ssm a n n G. (2003). E ffects o f sn ccess io n a l d y n am ics on m e ta p o p u la tio n p e rs is ten ce . E co logy 84: 882 -889 .

F erren b erg SM , S ch w ilk DW, K n ap p EE, G ro th E, K eeley JE. (2006). fire dec reases a rth ro p o d a b u n d a n c e h u t in c rea se s d iv ersity : ea rly a n d la te seaso n p re sc rib e d fire effects in a s ie rra n e v a d a m ix ed -co n ife r fo rest. Eire E col 2: 79-102 .

F ie re r N, Jackson R. (2006). T he d iv e rs ity a n d h iogeogra­p h y of so il b a c te ria l co m m u n itie s . P roc N a tl A c a d S c i U SA 103: 626 -631 .

H am m an ST, B urke IC, K n ap p EE. (2008). S o il n u tr ie n ts an d m ic ro h ia l ac tiv ity a fter ea rly a n d la te seaso n p re sc rib e d b u rn s in a S ie rra N evada m ix ed con ife r forest. E orest E co l M anag 256: 367 -374 .

H an so n CA, F u h rm a n J, H orner-D ev ine MG, M artin y JBH.(2012). B eyond h iogeog raph ic p a tte rn s : p ro cesses sh ap in g th e m ic ro h ia l lan d scap e . N a t B e v M icrob io l 10: 497 -5 0 6 .

H art SC, D eL uca TH, N ew m an GS, M acK enzie MD, B oyle SI. (2005). P ost-fire vegeta tive d y n am ics as d riv e rs of m ic ro h ia l co m m u n ity s tru c tu re a n d fu n c tio n in fo rest so ils. Eor E col M anage 220: 166-184 .

H u h h e ll SP. (2001). T h e U n ified N eu tra l T h eo ry O f B io d ive rs ity A n d B iogeography. P rin c e to n U n iv e rs ity P ress: P rin ce to n , NJ.

Jiang L, P a te l SN. (2008). C o m m u n ity a ssem b ly in th e p re sen ce of d is tu rh an ce : a m ic ro co sm exp erim en t. E co logy 89: 1931-1940 .

JM PV ersion 10 (2011). SA S In s titu te Inc.: Cary, NC.K n e lm an JE, Legg TM , O ’N eill SP, W ashenherger CL,

G onzalez A, C lev e lan d CC e t al. (2012). B acteria l co m m u n ity s tru c tu re a n d fu n c tio n change in a sso c ia ­tio n w ith co lo n ize r p la n ts d u rin g early p rim ary su cce ss io n in a g lac ie r fo refie ld . S o il B io l B io ch em 46: 172-180 .

K raft NJB, C o rn w ell WK, W ehh CO, A ckerly DD. (2007). T rait ev o lu tio n , co m m u n ity assem bly , a n d th e p h y lo ­g en e tic s tru c tu re of eco log ica l co m m u n itie s . A m N a t 170: 271-283 .

L an g en h ed e r S, B u llin g MT, S o lan M , P ro sse r JI. (2010). B ac te ria l h io d iv e rs ity -eco sy stem fu n c tio n in g re la tio n s are m o d ified h y e n v iro n m e n ta l com plex ity . P lo S one 5: 10834.

L an g en h ed e r S, P ro sse r JI. (2008). R esource av a ilab ility in f lu e n c e s th e d iv e rs ity o f a fu n c tio n a l g ro u p of

h e te ro tro p h ic so il bac te ria . E nviron M icrob io l 10: 2245-2256 .

L an g en h ed e r S, S zekely AJ. (2011). S p ec ies so rting a n d n e u tra l p ro cesses are b o th im p o rta n t d u rin g th e in itia l a ssem b ly of b a c te ria l co m m u n itie s . ISM E J 5: 1086-1094 .

L e ibo ld M A , M cP eek MA. (2006). C oex istence o f th e n ic h e a n d n e u tra l p e rsp ec tiv e s in co m m u n ity ecology. E co logy 8 : 1399-1410 .

Logue JB, L in d stro m ES. (2010). S p ec ies so rtin g affects h ac te r io p la n k to n co m m u n ity c o m p o s itio n as d e te r­m in e d h y 16S rD NA a n d 16S rR N A fin g e rp rin ts . ISM E J 4: 729 -738 .

L o zu p o n e C, K n igh t R. (2005). ‘U n iF rac: a n e w p h y lo g e ­n e tic m e th o d for co m p arin g m ic ro h ia l c o m m u n itie s ’. A p p l E nv iron M icrob io l 71: 8 228 -8235 .

L o zu p o n e C, H am ad y M, K n igh t R. (2006). U n iF rac - an o n lin e to o l for co m p arin g m ic ro h ia l co m m u n ity d iv e rs ity in a p h y lo g en e tic con tex t. B M C b io in fo r­m a tic s 7: 371.

L o zu p o n e CA, K n igh t R. (2007). G lobal p a tte rn s in b ac te ria l d iversity . P roc N a tl A c a d S c i U SA 104: 1 1436-11440 .

M artin y JBH, B o h an n an BJM, B row n JH, C o lw ell RK, F u h rm a n JA, G reen JL e t al. (2006). M icroh ia l h iogeography : p u tt in g m ic ro o rg an ism s o n th e m ap . N a t B e v M icrob io l 4: 102-112 .

M cC une B, M efford MJ. (2011). PC-OBD. M u ltiva ria te A n a ly s is o f E co logica l D ata. Version 6.0. M jM S oft­w are: G len ed en B each , OR.

M o n so n RK, L ip so n DL, B u rn s SP, T u rn ip se e d AA, D elany AC, W illiam s M K e t al. (2006). W in te r fo rest so il re sp ira tio n co n tro lled h y c lim a te a n d m ic ro h ia l co m ­m u n ity co m p o s itio n . N a tu re 439: 711-714 .

M u lv an ey RL. (1996). N itrogen - Ino rgan ic form s. In: S parks, DL, e t al. (ed). M e th o d s o f so il ana lysis, P art 3, S S S A B o o k Series 5. SSSA : M ad iso n , WI, p p 1123-1184 .

N em erg u t DR, A n d e rso n SP, C lev e lan d CC, M artin AP, M ille r AE, S e im o n A e t al. (2007). M icroh ia l co m m u ­n ity su cce ss io n in an u n v eg e ta ted , recen tly -deg la - c ia ted soil. M icrob E col 53: 110 -122 .

N em erg u t DR, C lev e lan d CC, W iede r WR, W ashenherger CL, T o w n sen d AR. (2010). P lo t-sca le m a n ip u la tio n s of o rganic m a tte r in p u ts to so ils co rre la te w ith sh ifts in m ic ro h ia l c o m m u n ity co m p o s itio n in a lo w la n d tro p ic a l ra in forest. S o il B io l B io ch em 42: 2153-2160 .

N em erg u t DR, C oste llo EK, H am ad y M, L o zu p o n e C, Jiang L, S c h m id t SK e t al. (2011). G lobal p a tte rn s in th e h io geog raphy of b a c te ria l taxa. E nviron M icrob io l 13: 135-144 .

O fite ru ID, L u n n M, C urtis TP, W ells GF, G ridd le CS, F ran c is CA e t al. (2010). C o m b in ed n ic h e a n d n e u tra l effects in a m ic ro h ia l w a s tew a te r tre a tm e n t co m m u ­nity . P roc N a tl A c a d S c i U SA 107: 15345-15350 .

G k san en J, B lan ch e t FG, K in d t R, L egendre P, M in c h in PR, O ’H ara RB e t al. (2011). Vegan: C o m m u n ity Ecology Package. B p a cka g e version 2: 0 -1 .

O stm an O, D rakare S, K ritzherg ES, L an g en h ed e r S, Logue JB, L in d s tro m ES. (2009). R eg ional in v a rian ce am ong m ic ro h ia l co m m u n itie s . E co logy le tters 13: 118-127 .

P au l EA, C lark FE. (1996). S o il M icrob io logy a n d B iochem istry , 2 n d E d ition . A cadem ic Press: N ew York, NY.

P eay KG, G arhelo tto M , B ru n s TD. (2010). E v id en ce of d isp e rsa l lim ita tio n in so il m icroo rgan ism s: iso la tio n

The ISME J o u rn a l

W ildfire sh ifts a ssem b iy p r o c esse s in a soli b acteriai com m unityS Ferrenberg et al

10re d u c e s sp ec ies r ic h n e ss on m y co rrh iz a l tree is lan d s . E co logy 91: 3631-3640 .

P e itik a in e n J, H iu k k a R, F ritze H. (2000). D oes sh o r t­te rm h ea tin g of fo rest h u m u s change its p ro p e rtie s as a su b s tra te for m icrohes? S o il B io l B io ch em 32: 277 -288 .

R eeder J, K n igh t R. (2010). R ap id ly d en o is in g p y ro se q u e n ­cing a m p lic o n read s hy e x p lo itin g ran k -ah u n d an ce d is tr ib u tio n s . N a t M e th o d s 7: 688 -689 .

S a ttin SR, C lev e lan d CC, H ood E, R eed SC, K ing AJ, S c h m id t SK e t al. (2009). F u n c tio n a l sh ifts in u n v eg e ta ted , p e rh u m id , re cen tly d eg lac ia ted so ils do n o t co rre la te w ith sh ifts in so il b a c te ria l c o m m u n ity co m p o s itio n . / M icrob io l 47: 673 -6 8 1 .

S c h m id t SK, C oste llo EK, N em erg u t DR, C lev e lan d CC, R eed SC, W ein trau h M N e t al. (2007). B iogeochem ical co n seq u en ces o f ra p id m ic ro h ia l tu rn o v e r a n d seaso ­n a l su ccess io n in so il. E co logy 88: 1379-1385 .

S ch o en n ag e l T, S h e rr iff RL, V ehlen TT. (2011). F ire h is to ry an d tree re c ru itm e n t in th e C o lo rado F ro n t R ange u p p e r m o n tan e zone: im p lic a tio n s for fo rest re s to ra ­tio n . E col A p p l 21: 2210-2222 .

S loan WT, M ary L u n n M, W oodcock S, H ead IM , N ee S, C u rtis TP. (2006). Q u an tify in g th e ro les o f im m ig ra tio n an d ch an ce in sh a p in g p ro k ary o te co m m u n ity s tru c ­tu re . E n v M icrob io l 8: 732-740 .

S m ith NR, K ish c h u k BE, M o h n WW. (2008). E ffects of w ild fire a n d h a rv e s t d is tu rb a n c e s o n fo rest so il b a c te ria l co m m u n itie s . A p p l E nv iron M icrob io l 74: 216 -224 .

Team RDC. (2011). B: A L anguage A n d E n v iro n m e n t Eor S ta tis tica l C om pu ting . R F o u n d a tio n for S ta tis tica l C om pu ting : V ienna, A ustria .

T elford RJ, V andv ik V Birks HJB (2006). D ispersa l l im ita ­tio n m a tte rs for m ic ro h ia l m o rp h o sp ec ie s . Sc ien ce 312: 1015.

T u rn er MG, S m ith w ick EAH , M etzger KL, T in k er DB, R om m e W H. (2007). Ino rgan ic n itro g en av a ilab ility fo llo w in g severe s tan d -rep lac in g fire in th e G reater Y ellow stone E cosystem . Proc N a tl A c a d S c i U SA 104: 4 7 8 2 ^ 7 8 9 .

V an D er H e ijd en M C A , B ardgett RD, V an S traa len NM. (2008). T he u n se e n m ajority : so il m icrohes as d rivers

of p la n t d iv e rs ity a n d p ro d u c tiv ity in te rre s tr ia l ecosystem s. E col L e tt 11: 296 -310 .

V ehlen TT, K itzherger T, D o nnegan J. (2000). C lim atic an d h u m a n in f lu e n c e s on fire reg im es in p o n d e ro sa p in e fo rests in th e C o lo rado F ro n t R ange. E col A p p l 10: 1178-1195 .

V ellend M. (2010). C o n cep tu a l sy n th e s is in c o m m u n ity ecology. Q B e v B io l 85: 183-206 .

W ald rop MP, H a rd en JW. (2008). In te rac tiv e effects of w ild fire an d p e rm afro s t on m ic ro h ia l co m m u n itie s a n d so il p ro cesses in an A lask an b lack sp ru ce forest. G lobal C hange B io l 14: 2591-2602 .

W ang GG, K em hall KJ. (2005). E ffects o f fire sev e rity on early d e v e lo p m e n t of u n d e rs to ry vegeta tion . Can J E orest B es 35: 254-262 .

W ang Q, Z h o n g M, W ang S. (2012). A m e ta -an a ly s is on the re sp o n se of m ic ro h ia l b io m ass, d is so lv ed organ ic m atter, re sp ira tio n , an d N m in e ra liz a tio n in m in e ra l so il to fire in fo rest ecosystem s. E orest E co l M anag 271: 9 1 -97 .

W ardle DA, Z ack risso n O, H ornherg G, G alle t C. (1997). T he in f lu e n c e of is la n d a rea on eco sy stem p ro p ertie s . S c ien c e 277: 1296-1299 .

W ardle DA, Z ack risso n O, N ilsso n MC. (1998). T he ch a rco a l effect in B oreal fo rests: m ech a n ism s an d eco log ica l co n seq u en ces . O ecologia 115: 4 1 9 -4 2 6 .

W oodcock S, V an D er C ast CJ, B ell T, L u n n M, C urtis TP, H ead IM e t al. (2007). N eu tra l assem b ly of bac te ria l co m m u n itie s . E LM S M icrob io l E co l 62: 171-180 .

Y eager CM, N o rth ru p DE, G row CC, B arns SM , K uske CR. (2005). C hanges in n itro g en -fix in g a n d am m on ia- o x id iz in g b ac te ria l co m m u n itie s in so il of a m ix ed co n ife r fo rest a fter w ild fire . A p p l E nv iron M icrobio l 71: 2713-2722 .

Y ilm az P, K o ttm an n R, F ie ld D, K n igh t R, C ole JR, A m aral- Z e ttle r L e t al. (2011). M in im u m in fo rm a tio n ab o u t a m ark e r gene seq u en ce (M IMARKS) a n d m in im u m in fo rm a tio n ab o u t any (x) seq u en ce (M IxS) sp ec ific a ­tio n s. N a t B io tech n o l 29: 415 -4 2 0 .

Z ak DR, H o lm es WE, W hite DC, P eaco ck AD, T ilm an D. (2003). P la n t d iversity , so il m ic ro h ia l co m m u n itie s , a n d ecosystem fu n c tio n : are th e re an y lin k s? E co logy 84: 2042-2050 .

Supplementary Information accompanies the paper on The ISME Journal wehsite (http://www.nature.com/ismej)

The ISME Jo u rn a l