SHORT COMMUNICATION
Bacterial community composition of divergent soil habitats in a polar desert Kevin M. Geyer1, Adam E. Altrichter1, Cristina D. Takacs-Vesbach2, David J. Van Horn2, Michael N. Gooseff3 & John E. Barrett1 1
Department of Biological Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA; 2Department of Biology, University of New Mexico, Albuquerque, NM, USA; and 3Department of Civil and Environmental Engineering, Pennsylvania State University, University Park, PA, USA
Correspondence: Kevin M. Geyer, Department of Biological Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA. Tel.: (540) 231 3827; fax: (540) 231 9307; e-mail: [email protected] Present address: Michael N. Gooseff, Department of Civil and Environmental Engineering, Colorado State University, Fort Collins, CO 80523, USA
MICROBIOLOGY ECOLOGY
Received 18 October 2013; revised 13 February 2014; accepted 13 February 2014. DOI: 10.1111/1574-6941.12306 Editor: Max H€ aggblom Keywords Antarctic Dry Valleys; environmental gradients; microbial ecology; geochemical severity.
Abstract Edaphic factors such as pH, organic matter, and salinity are often the most significant drivers of diversity patterns in soil bacterial communities. Desert ecosystems in particular are model locations for examining such relationships as food web complexity is low and the soil environment is biogeochemically heterogeneous. Here, we present the findings from a 16S rRNA gene sequencing approach used to observe the differences in diversity and community composition among three divergent soil habitats of the McMurdo Dry Valleys, Antarctica. Results show that alpha diversity is significantly lowered in high pH soils, which contain higher proportions of the phyla Acidobacteria and Actinobacteria, while mesic soils with higher soil organic carbon (and ammonium) content contain high proportions of Nitrospira, a nitrite-oxidizing bacteria. Taxonomic community resolution also had a significant impact on our conclusions, as pH was the primary predictor of phylum-level diversity, while moisture was the most significant predictor of diversity at the genus level. Predictive power also increased with increasing taxonomic resolution, suggesting a potential increase in niche-based drivers of bacterial community composition at such levels.
Soil microbial communities, like macroscopic communities, are subject to a variety of niche-based and neutralbased mechanisms which contribute to observed biogeographic patterns (Martiny et al., 2006; Fierer & Lennon, 2011; Hanson et al., 2012). Examples of niche-based drivers come from laboratory manipulations of edaphic properties like organic carbon (Orwin et al., 2006; Eilers et al., 2010) as well as across natural (e.g. pH, moisture) gradients spanning local (Zhou et al., 2002; Zeglin et al., 2011) to continental (Fierer & Jackson, 2006) scales. Although examples of neutral-based mechanisms also exist for microorganisms (Takacs-Vesbach et al., 2008; Sokol et al., 2013), evidence of environmental controls dominates the soils literature (Hanson et al., 2012). Consequently, understanding the effects of environmental drivers on microbial communities is an important step toward developing an integrated understanding of biodiversity FEMS Microbiol Ecol && (2014) 1–5
patterns in soil systems, particularly as we increasingly recognize the intimate connection between certain soil functions and the community composition of mediating microorganisms (Strickland et al., 2009). Desert ecosystems are ideal locations to examine environmental controls over the diversity and composition of microbial communities because of strong spatial heterogeneity in soil properties and generally low food web complexity, relative to temperate systems. Soils of low habitat suitability (saline, low organic carbon) are interspersed with localized ‘resource islands’ that display more mesic conditions due to the flushing effects of available moisture and primary production (e.g. NoyMeir, 1973). In the McMurdo Dry Valleys of Antarctica, such locations are hot spots of rapid biogeochemical cycling where extremes in pH (> 9.0) and electrical conductivity (> 10 000 lS cm1) are ameliorated, organic ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
2
carbon is more abundant, and biologic activity is increased (Barrett et al., 2009). Although the overall diversity and abundance of Antarctic eukaryotic organisms are reduced because of severe climatic constraints (Treonis et al., 1999; Adams et al., 2006), diverse communities of soil microorganisms have been revealed by molecular sequencing (Cary et al., 2010; Takacs-Vesbach et al., 2010). Recent studies have demonstrated that moisture availability (Zeglin et al., 2011), geochemistry (e.g. pH, soluble ions; Lee et al., 2012), carbon concentration (Aislabie et al., 2009; Geyer et al., 2013), or some combination of these (Niederberger et al., 2008; Smith et al., 2010; Stomeo et al., 2012; Sokol et al., 2013; Van Horn et al., 2013) are key drivers of bacterial diversity and community composition. Our previous work in this region has focused on the identification and characterization of soil habitats associated with the full range of observable primary production as determined by chlorophyll a quantification from moss and cyanobacterial cryptogamic mats (Geyer et al., 2013). This effort documented soil habitats from among the more extreme in geochemical conditions (e.g. pH, electrical conductivity) as well as resource availability (e.g. moisture, organic carbon) locally known. Multivariate (canonical correspondence) analysis revealed three divergent soil habitats: those with extreme pH (> 9.40), high conductivity (> 1000 lS cm1), and mesic habitats with increased moisture, soil organic carbon (SOC), and generally reduced pH and conductivity. Based on these previous results, twelve samples from among these three habitats (four from each) were selected for pyrosequencing of the 16S rRNA gene to address the following questions: (1) How does bacterial community diversity and composition vary between geochemical habitats? (2) What environmental properties are the best predictors of bacterial community diversity? (3) How does taxonomic resolution influence these conclusions, that is, do phyla- and genera-level assessments change our understanding of the importance of environmental drivers? Soils were collected from Wright and Taylor Valleys of southern Victoria Land, Antarctica, in December 2010. Triplicate samples were taken from within a 2.5 m2 area to a depth of 5 cm, pooled into a single (c. 500 g) composite sample, and immediately frozen at 20 °C to be shipped to Virginia Tech for further analysis. Soils were sourced from five sampling sites distributed throughout the region (CAST – Canada Stream; HYPO – hypolithic uplands; GRCR – Green Creek; BRBB – Lake Brownworth pond; ONYX – Onyx River) found near open water (except the uplands) as described in Geyer et al. (2013). These sites constrain a natural gradient in dry valley soil productivity as determined by visual inspection of cryptogamic mat density while also capturing broad ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
K.M. Geyer et al.
gradients in soil geochemistry. Distance-decay effects for bacterial communities were previously examined for these soils and found to be significant, although of much less importance than environmental effects. Geochemical properties were assayed from soil/liquid extract slurries using standard methods for this region (Nkem et al., 2006). Chlorophyll a was measured using a spectrophotometric analysis adapted from Metting (1994) and Castle et al. (2011). DNA was extracted from soil using a modified cetyltrimethylammonium bromide (CTAB) and phenol/chloroform isoamyl alcohol procedure (Giovannoni et al., 1990). Barcoded amplicon pyrosequencing of 16S rRNA genes was performed as described previously (Van Horn et al., 2013) using the universal bacterial primers 939F (50 TTG ACG GGG GCC CGC ACA AG) and 1492R (50 -GTT TAC CTT GTT ACG ACT T-30 ) containing variable regions V6–V9. Briefly, 50 ng of template DNA was amplified per sample with Roche-specific sequencing adapters and an error-correcting barcode (Hamady et al., 2008). Amplicons were purified using Agencourt Ampure beads and combined in equimolar concentrations for pyrosequencing on a Roche 454 FLX using Roche titanium reagents and titanium procedures. Raw sequences were quality-filtered, denoised, screened for PCR errors, and
Fig. 1. Distance-based (Bray) redundancy analysis of log (x+1)normalized phylum-level bacterial communities, constrained by the independent, square-root-normalized soil properties soil organic carbon (SOC), pH, and electrical conductivity. Species scores (red text), environmental parameter biplots, and 95% standard deviation confidence intervals (of geochemical associated groups) are overlain. Axes indicate the percent of variation explained by nonindependent, constrained eigenvalues. Site labels indicate region of sampling followed by a visual estimate of surface primary production noted during sampling (e.g. high, medium, low).
FEMS Microbiol Ecol && (2014) 1–5
Bacterial communities of polar desert soil habitats
chimera-checked using AmpliconNoise and Perseus to minimize potential artifacts (Quince et al., 2011). Sequences were then processed using the QIIME PIPELINE v. 1.6 (Caporaso et al., 2010). After quality control and demultiplexing, 125 812 sequences were clustered into 7343 OTUs at 97% similarity using uclust (Edgar, 2010). Taxonomic assignments were made using the RDP classifier (Wang et al., 2007). Raw sequence files for the 12 samples analyzed here have been deposited in the NCBI Sequence Read Archive and are accessible through a BioProject database (Accession: PRJNA228946). All univariate and multivariate statistical analyses were performed with R statistical freeware (R Core Team, 2011) using square-root-transformed environmental data and rarefied, log (x + 1)-normalized community data. Distance-based (Bray) redundancy analysis (RDA) was used to ordinate the twelve sites in space constrained by the independent environmental parameters pH, conductivity, and SOC (capscale {vegan}). Significance of geochemical groupings was assessed using analysis of similarities (ANOSIM {vegan}) and the contribution of taxa to average overall pairwise sample dissimilarity with similarity percentages (SIMPER {vegan}; Clarke, 1993). Relative abundance heatmap (heatmap.2 {gplots}) was created from phylum-level bacterial annotations.
3
Rarefaction demonstrated that sequencing depth was sufficient to generate decreasing slopes of collector curves for all but one sample (CAST-H; Supporting Information, Fig. S1), suggesting most samples had low to intermediate diversity. Sequencing results also corroborate findings from previous TRFLP analysis of the same sites (Geyer et al., 2013), in that communities from habitats of divergent soil properties appear highly dissimilar (Fig. 1). Sites grouped by their geochemical association (SOC, pH, and conductivity) are significantly different in community composition from one another (ANOSIM P-value = 0.001). Also apparent are the distinct phyla which appear strongly associated with particular geochemical habitats, such as the phyla Acidobacteria and Actinobacteria in high pH regions, genus Nitrospira in high carbon regions, and phylum Firmicutes in high carbon and conductivity regions (confirmed by SIMPER analysis; Fig. 2). Acidobacteria are common inhabitants of oligotrophic environments and across the range of soil pH examined here (pH > 9.4, SOC < 700 mg kg1 dry soil, gravimetric moisture < 0.30%), providing support for the ecological classification of soil bacteria as suggested by Fierer et al. (2007). Although initially isolated from acidic mineral environments, this phylum is increasingly recognized to constitute c. 20% of most soil bacterial communities
Fig. 2. Relative abundance heatmap of log (x+1)-normalized phylum-level bacterial communities. Dendrogram of sample sites (average clustering technique) based on community similarity along left axis; locations color-coded by geochemical association (light blue = conductivity; blue = soil organic carbon; dark blue = pH). Upper dendrogram depicts clustering of phyla by co-occurrence (average clustering technique).
FEMS Microbiol Ecol && (2014) 1–5
ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
K.M. Geyer et al.
4
regardless of acidity (Cary et al., 2010; Naether et al., 2012). Our sequencing results indicate the presence of seven Acidobacteria subgroups (1, 3, 4, 6, 7, 10, and 25), along with 21 genera of Actinobacteria (e.g. Arthrobacter and Pseudonocardia) and 18 genera of Firmicutes (e.g. Alkaliphilus and Clostridium). The abundance of Nitrospira in mesic habitats is also consistent with our understanding of dry valley biogeochemistry, as these wetted areas often display increased ammonium availability (mean = 0.3 lg g1 dry soil; e.g. Barrett et al., 2009). Analysis of variance between geochemical groupings indicates a significantly reduced richness and diversity of bacterial communities at the phylum, order, and genus levels, as well as Chao1 estimated richness, from locations with high soil pH. Bacterial richness of high conductivity and mesic habitats were not significantly different. Correlations with soil properties suggest few parameters with potential explanatory power of phylum-level richness, but a very strong relationship between soil moisture and genus-level richness (Supporting Information, Table S1). Multiple linear regression (mixed stepwise selection) was used to examine these results further and confirms bacterial community richness is primarily explained by pH (adjusted R2 = 0.57; SOC and conductivity as co-parameters) at the phylum level, shifting to water (adjusted R2 = 0.91; chlorophyll a and conductivity as co-parameters) at the order level, and moisture alone (R2 = 0.91) at the genus level (Table S2). Van Horn et al. (2013) also found context-dependent roles for edaphic factors in explaining dry valley bacterial richness, in this case across kilometer spatial scales. Soil habitats of the McMurdo Dry Valleys have divergent properties that harbor distinct bacterial communities which in some cases have significantly lowered diversity (e.g. alkaline locations). The explanation of c. 60% of phylum-level richness by soil pH agrees with the continental-scale assessment of bacterial biogeography by Fierer & Jackson (2006). Adjusting the taxonomic resolution to a finer (order or genus) scale provided much more explanatory power here (91%). This result is logical considering the extensive diversity among prokaryotes at the phylum level and potential difficulty in explaining properties of any such complex group by only a few parameters. However, such increasing explanatory power with increasing taxonomic resolution may also suggest a shift from neutral to niche-based controls for bacteria at the genus level. Additional sequencing effort will be needed to understand the complex response among diverse microbial taxa and resulting diversity patterns to edaphic drivers, with particular attention needed to constrain potential variability caused by spatial relationships among communities. An important consideration for the maintenance and management of bacterial communities ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
and associated functions should therefore be the taxonomic level at which different soil properties may elicit effects (Martiny et al., 2013). As our understanding of the functional redundancy of microorganisms continues to be refined, information linking environmental conditions to specific taxa, or even communities of different inherent diversities, may provide avenues for the intentional selection of communities capable of desirable and predictable (e.g. low variability) functioning using knowledge of environmental controls.
References Adams BJ, Bardgett RD, Ayres E et al. (2006) Diversity and distribution of Victoria Land biota. Soil Biol Biochem 38: 3003–3018. Aislabie JS, Jordan S, Ayton J, Klassen JL, Barker GM & Turner S (2009) Bacterial diversity associated with ornithogenic soil of the Ross Sea region, Antarctica. Can J Microbiol 55: 21–36. Barrett JE, Gooseff MN & Takacs-Vesbach C (2009) Spatial variation in soil active-layer geochemistry across hydrologic margins in polar desert ecosystems. Hydrol Earth Syst Sci 13: 2349–2358. Caporaso JG, Kuczynski J, Stombaugh J et al. (2010) QIIME allows analysis of high- throughput community sequencing data. Nat Methods 7: 335–336. Cary CS, McDonald IR, Barrett JE & Cowan DA (2010) On the rocks: the microbiology of Antarctic Dry Valley soils. Nat Rev Microbiol 8: 129–138. Castle SC, Morrison CD & Barger NN (2011) Extraction of chlorophyll a from biological soil crusts: a comparison of solvents for spectrophotometric determination. Soil Biol Biochem 43: 853–856. Clarke KR (1993) Non-parametric multivariate analyses of changes in community structure. Aust J Ecol 18: 117–143. Edgar RC (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461. Eilers KG, Lauber CL, Knight R & Fierer N (2010) Shifts in bacterial community structure associated with inputs of low molecular weight carbon compounds to soil. Soil Biol Biochem 42: 896–903. Fierer N & Jackson RB (2006) The diversity and biogeography of soil bacterial communities. P Natl Acad Sci USA 103: 626–631. Fierer N & Lennon JT (2011) The generation and maintenance of diversity in microbial communities. Am J Bot 98: 439–448. Fierer N, Bradford MA & Jackson RB (2007) Toward an ecological classification of soil bacteria. Ecology 88: 1354– 1364. Geyer KM, Altrichter AE, Van Horn DJ, Takacs-Vesbach CD, Gooseff MN & Barrett JE (2013) Environmental controls over bacterial communities in polar desert soils. Ecosphere 4: art127.
FEMS Microbiol Ecol && (2014) 1–5
5
Bacterial communities of polar desert soil habitats
Giovannoni SJ, Delong EF, Schmidt TM & Pace NR (1990) Tangential flow filtration and preliminary phylogenetic analysis of marine picoplankton. Appl Environ Microbiol 56: 2572–2575. Hamady M, Walker JJ, Harris JK, Gold NJ & Knight R (2008) Error-correcting barcoded primers allow hundreds of samples to be pyrosequenced in multiplex. Nat Methods 5: 235–237. Hanson CA, Fuhrman JA, Horner-Devine MC & Martiny JBH (2012) Beyond biogeographic patterns: processes shaping the microbial landscape. Nat Rev Microbiol 10: 497–506. Lee CK, Barbier BA, Bottos EM, McDonald IR & Cary CS (2012) The Inter-Valley Soil Comparative Survey: the ecology of Dry Valley edaphic microbial communities. ISME J 6: 1046–1057. Martiny JBH, Bohannan BJM, Brown JH et al. (2006) Microbial biogeography: putting microorganisms on the map. Nat Rev Microbiol 4: 102–112. Martiny AC, Treseder K & Pusch G (2013) Phylogenetic conservatism of functional traits in microorganisms. ISME J 7: 830–838. Metting FB Jr (1994) DMSO extraction method for quantification of chlorophyll in soil. Part 2: Microbiological and Biochemical Properties (Bigham JM, ed), pp. 443–444. Soil Science Society of America, Inc., Madison, WI. Naether A, Foesel BU, Naegele V et al. (2012) Environmental factors affect Acidobacterial communities below the subgroup level in grassland and forest soils. Appl Environ Microbiol 78: 7398–7406. Niederberger TD, McDonald IR, Hacker AL, Soo RM, Barrett JE, Wall DH & Cary CS (2008) Microbial community composition in soils of Northern Victoria Land, Antarctica. Environ Microbiol 10: 1713–1724. Nkem JN, Virginia RA, Barrett JE, Wall DH & Li G (2006) Salt tolerance and survival thresholds for two species of Antarctic soil nematodes. Polar Biol 29: 643–651. Noy-Meir I (1973) Desert ecosystems: environment and producers. Annu Rev Ecol Syst 4: 25–51. Orwin KH, Wardle DA & Greenfield LG (2006) Ecological consequences of carbon substrate identity and diversity in a laboratory study. Ecology 87: 580–593. Quince C, Lanzen A, Davenport RJ & Turnbaugh PJ (2011) Removing noise from pyrosequenced amplicons. BMC Bioinformatics 12: 38. R Core Team (2011) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org. Smith JL, Barrett JE, Tusnady G, Rejtoe L & Cary CS (2010) Resolving environmental drivers of microbial community structure in Antarctic soils. Antarct Sci 22: 673–680. Sokol ER, Herbold C, Lee CK, Cary SC & Barrett JE (2013) Local and regional influences over soil microbial metacommunities in the Transantarctic Mountains. Ecosphere 4: art136.
FEMS Microbiol Ecol && (2014) 1–5
Stomeo F, Makhalanyane TP, Valverde A, Pointing SB, Stevens MI, Cary CS, Tuffin MI & Cowan DA (2012) Abiotic factors influence microbial diversity in permanently cold soil horizons of a maritime-associated Antarctic Dry Valley. FEMS Microbiol Ecol 82: 326–340. Strickland MS, Lauber C, Fierer N & Bradford MA (2009) Testing the functional significance of microbial community composition. Ecology 90: 441–451. Takacs-Vesbach C, Mitchell K, Jackson-Weaver O & Reysenbach A (2008) Volcanic calderas delineate biogeographic provinces among Yellowstone thermophiles. Environ Microbiol 10: 1681–1689. Takacs-Vesbach C, Zeglin L, Barrett JE, Gooseff MN & Priscu JC (2010) Factors promoting microbial diversity in the McMurdo Dry Valleys, Antarctica. Life in Antarctic Deserts and Other Cold Dry Environments: Astrobiological Analogs (Doran P, Lyons WB, & McKnight DM, eds), pp. 221–257. Cambridge University Press, Cambridge, UK. Treonis AM, Wall DH & Virginia RA (1999) Invertebrate biodiversity in Antarctic dry valley soils and sediments. Ecosystems 2: 482–492. Van Horn DJ, Van Horn ML, Barrett JE, Gooseff MN, Altrichter AE, Geyer KM, Zeglin LH & Takacs-Vesbach CD (2013) Factors controlling soil microbial biomass and bacterial diversity and community composition in a cold desert ecosystem: role of geographic scale. PLoS One 8: e66103. Wang Q, Garrity GM, Tiedje JM & Cole JR (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 73: 5261–5267. Zeglin LH, Dahm CN, Barrett JE, Gooseff MN, Fitpatrick SK & Takacs-Vesbach CD (2011) Bacterial community structure along moisture gradients in the parafluvial sediments of two ephemeral desert streams. Microb Ecol 61: 543–556. Zhou JZ, Xia BC, Treves DS, Wu LY, Marsh TL, O’Neil RV, Palumbo AV & Tiedje JM (2002) Spatial and resource factors influencing high microbial diversity in soil. Appl Environ Microbiol 68: 326–334.
Supporting Information Additional Supporting Information may be found in the online version of this article: Fig. S1. Rarefaction curves of 12 pyrosequenced samples. Table S1. Spearman nonparametric correlation matrix between soil parameters (square-root-transformed), Chao1 estimated richness, and bacterial community richness at the phylum (Phy R), order (Ord R), and genus (Gen R) taxonomic levels. Table S2. Multiple linear regression results for the prediction of phylum, order, and genus-level richness through mixed stepwise selection of model effects (a = 0.15).
ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
Bacterial community composition of divergent soil habitats in a polar desert Kevin M. Geyer1, Adam E. Altrichter1, Cristina D. Takacs-Vesbach2, David J. Van Horn2, Michael N. Gooseff3 & John E. Barrett1 1
Department of Biological Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA, USA; 2Department of Biology, University of New Mexico, Albuquerque, NM, USA; and 3Department of Civil and Environmental Engineering, Pennsylvania State University, University Park, PA, USA
Correspondence: Kevin M. Geyer, Department of Biological Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA. Tel.: (540) 231 3827; fax: (540) 231 9307; e-mail: [email protected] Present address: Michael N. Gooseff, Department of Civil and Environmental Engineering, Colorado State University, Fort Collins, CO 80523, USA
MICROBIOLOGY ECOLOGY
Received 18 October 2013; revised 13 February 2014; accepted 13 February 2014. DOI: 10.1111/1574-6941.12306 Editor: Max H€ aggblom Keywords Antarctic Dry Valleys; environmental gradients; microbial ecology; geochemical severity.
Abstract Edaphic factors such as pH, organic matter, and salinity are often the most significant drivers of diversity patterns in soil bacterial communities. Desert ecosystems in particular are model locations for examining such relationships as food web complexity is low and the soil environment is biogeochemically heterogeneous. Here, we present the findings from a 16S rRNA gene sequencing approach used to observe the differences in diversity and community composition among three divergent soil habitats of the McMurdo Dry Valleys, Antarctica. Results show that alpha diversity is significantly lowered in high pH soils, which contain higher proportions of the phyla Acidobacteria and Actinobacteria, while mesic soils with higher soil organic carbon (and ammonium) content contain high proportions of Nitrospira, a nitrite-oxidizing bacteria. Taxonomic community resolution also had a significant impact on our conclusions, as pH was the primary predictor of phylum-level diversity, while moisture was the most significant predictor of diversity at the genus level. Predictive power also increased with increasing taxonomic resolution, suggesting a potential increase in niche-based drivers of bacterial community composition at such levels.
Soil microbial communities, like macroscopic communities, are subject to a variety of niche-based and neutralbased mechanisms which contribute to observed biogeographic patterns (Martiny et al., 2006; Fierer & Lennon, 2011; Hanson et al., 2012). Examples of niche-based drivers come from laboratory manipulations of edaphic properties like organic carbon (Orwin et al., 2006; Eilers et al., 2010) as well as across natural (e.g. pH, moisture) gradients spanning local (Zhou et al., 2002; Zeglin et al., 2011) to continental (Fierer & Jackson, 2006) scales. Although examples of neutral-based mechanisms also exist for microorganisms (Takacs-Vesbach et al., 2008; Sokol et al., 2013), evidence of environmental controls dominates the soils literature (Hanson et al., 2012). Consequently, understanding the effects of environmental drivers on microbial communities is an important step toward developing an integrated understanding of biodiversity FEMS Microbiol Ecol && (2014) 1–5
patterns in soil systems, particularly as we increasingly recognize the intimate connection between certain soil functions and the community composition of mediating microorganisms (Strickland et al., 2009). Desert ecosystems are ideal locations to examine environmental controls over the diversity and composition of microbial communities because of strong spatial heterogeneity in soil properties and generally low food web complexity, relative to temperate systems. Soils of low habitat suitability (saline, low organic carbon) are interspersed with localized ‘resource islands’ that display more mesic conditions due to the flushing effects of available moisture and primary production (e.g. NoyMeir, 1973). In the McMurdo Dry Valleys of Antarctica, such locations are hot spots of rapid biogeochemical cycling where extremes in pH (> 9.0) and electrical conductivity (> 10 000 lS cm1) are ameliorated, organic ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
2
carbon is more abundant, and biologic activity is increased (Barrett et al., 2009). Although the overall diversity and abundance of Antarctic eukaryotic organisms are reduced because of severe climatic constraints (Treonis et al., 1999; Adams et al., 2006), diverse communities of soil microorganisms have been revealed by molecular sequencing (Cary et al., 2010; Takacs-Vesbach et al., 2010). Recent studies have demonstrated that moisture availability (Zeglin et al., 2011), geochemistry (e.g. pH, soluble ions; Lee et al., 2012), carbon concentration (Aislabie et al., 2009; Geyer et al., 2013), or some combination of these (Niederberger et al., 2008; Smith et al., 2010; Stomeo et al., 2012; Sokol et al., 2013; Van Horn et al., 2013) are key drivers of bacterial diversity and community composition. Our previous work in this region has focused on the identification and characterization of soil habitats associated with the full range of observable primary production as determined by chlorophyll a quantification from moss and cyanobacterial cryptogamic mats (Geyer et al., 2013). This effort documented soil habitats from among the more extreme in geochemical conditions (e.g. pH, electrical conductivity) as well as resource availability (e.g. moisture, organic carbon) locally known. Multivariate (canonical correspondence) analysis revealed three divergent soil habitats: those with extreme pH (> 9.40), high conductivity (> 1000 lS cm1), and mesic habitats with increased moisture, soil organic carbon (SOC), and generally reduced pH and conductivity. Based on these previous results, twelve samples from among these three habitats (four from each) were selected for pyrosequencing of the 16S rRNA gene to address the following questions: (1) How does bacterial community diversity and composition vary between geochemical habitats? (2) What environmental properties are the best predictors of bacterial community diversity? (3) How does taxonomic resolution influence these conclusions, that is, do phyla- and genera-level assessments change our understanding of the importance of environmental drivers? Soils were collected from Wright and Taylor Valleys of southern Victoria Land, Antarctica, in December 2010. Triplicate samples were taken from within a 2.5 m2 area to a depth of 5 cm, pooled into a single (c. 500 g) composite sample, and immediately frozen at 20 °C to be shipped to Virginia Tech for further analysis. Soils were sourced from five sampling sites distributed throughout the region (CAST – Canada Stream; HYPO – hypolithic uplands; GRCR – Green Creek; BRBB – Lake Brownworth pond; ONYX – Onyx River) found near open water (except the uplands) as described in Geyer et al. (2013). These sites constrain a natural gradient in dry valley soil productivity as determined by visual inspection of cryptogamic mat density while also capturing broad ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
K.M. Geyer et al.
gradients in soil geochemistry. Distance-decay effects for bacterial communities were previously examined for these soils and found to be significant, although of much less importance than environmental effects. Geochemical properties were assayed from soil/liquid extract slurries using standard methods for this region (Nkem et al., 2006). Chlorophyll a was measured using a spectrophotometric analysis adapted from Metting (1994) and Castle et al. (2011). DNA was extracted from soil using a modified cetyltrimethylammonium bromide (CTAB) and phenol/chloroform isoamyl alcohol procedure (Giovannoni et al., 1990). Barcoded amplicon pyrosequencing of 16S rRNA genes was performed as described previously (Van Horn et al., 2013) using the universal bacterial primers 939F (50 TTG ACG GGG GCC CGC ACA AG) and 1492R (50 -GTT TAC CTT GTT ACG ACT T-30 ) containing variable regions V6–V9. Briefly, 50 ng of template DNA was amplified per sample with Roche-specific sequencing adapters and an error-correcting barcode (Hamady et al., 2008). Amplicons were purified using Agencourt Ampure beads and combined in equimolar concentrations for pyrosequencing on a Roche 454 FLX using Roche titanium reagents and titanium procedures. Raw sequences were quality-filtered, denoised, screened for PCR errors, and
Fig. 1. Distance-based (Bray) redundancy analysis of log (x+1)normalized phylum-level bacterial communities, constrained by the independent, square-root-normalized soil properties soil organic carbon (SOC), pH, and electrical conductivity. Species scores (red text), environmental parameter biplots, and 95% standard deviation confidence intervals (of geochemical associated groups) are overlain. Axes indicate the percent of variation explained by nonindependent, constrained eigenvalues. Site labels indicate region of sampling followed by a visual estimate of surface primary production noted during sampling (e.g. high, medium, low).
FEMS Microbiol Ecol && (2014) 1–5
Bacterial communities of polar desert soil habitats
chimera-checked using AmpliconNoise and Perseus to minimize potential artifacts (Quince et al., 2011). Sequences were then processed using the QIIME PIPELINE v. 1.6 (Caporaso et al., 2010). After quality control and demultiplexing, 125 812 sequences were clustered into 7343 OTUs at 97% similarity using uclust (Edgar, 2010). Taxonomic assignments were made using the RDP classifier (Wang et al., 2007). Raw sequence files for the 12 samples analyzed here have been deposited in the NCBI Sequence Read Archive and are accessible through a BioProject database (Accession: PRJNA228946). All univariate and multivariate statistical analyses were performed with R statistical freeware (R Core Team, 2011) using square-root-transformed environmental data and rarefied, log (x + 1)-normalized community data. Distance-based (Bray) redundancy analysis (RDA) was used to ordinate the twelve sites in space constrained by the independent environmental parameters pH, conductivity, and SOC (capscale {vegan}). Significance of geochemical groupings was assessed using analysis of similarities (ANOSIM {vegan}) and the contribution of taxa to average overall pairwise sample dissimilarity with similarity percentages (SIMPER {vegan}; Clarke, 1993). Relative abundance heatmap (heatmap.2 {gplots}) was created from phylum-level bacterial annotations.
3
Rarefaction demonstrated that sequencing depth was sufficient to generate decreasing slopes of collector curves for all but one sample (CAST-H; Supporting Information, Fig. S1), suggesting most samples had low to intermediate diversity. Sequencing results also corroborate findings from previous TRFLP analysis of the same sites (Geyer et al., 2013), in that communities from habitats of divergent soil properties appear highly dissimilar (Fig. 1). Sites grouped by their geochemical association (SOC, pH, and conductivity) are significantly different in community composition from one another (ANOSIM P-value = 0.001). Also apparent are the distinct phyla which appear strongly associated with particular geochemical habitats, such as the phyla Acidobacteria and Actinobacteria in high pH regions, genus Nitrospira in high carbon regions, and phylum Firmicutes in high carbon and conductivity regions (confirmed by SIMPER analysis; Fig. 2). Acidobacteria are common inhabitants of oligotrophic environments and across the range of soil pH examined here (pH > 9.4, SOC < 700 mg kg1 dry soil, gravimetric moisture < 0.30%), providing support for the ecological classification of soil bacteria as suggested by Fierer et al. (2007). Although initially isolated from acidic mineral environments, this phylum is increasingly recognized to constitute c. 20% of most soil bacterial communities
Fig. 2. Relative abundance heatmap of log (x+1)-normalized phylum-level bacterial communities. Dendrogram of sample sites (average clustering technique) based on community similarity along left axis; locations color-coded by geochemical association (light blue = conductivity; blue = soil organic carbon; dark blue = pH). Upper dendrogram depicts clustering of phyla by co-occurrence (average clustering technique).
FEMS Microbiol Ecol && (2014) 1–5
ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
K.M. Geyer et al.
4
regardless of acidity (Cary et al., 2010; Naether et al., 2012). Our sequencing results indicate the presence of seven Acidobacteria subgroups (1, 3, 4, 6, 7, 10, and 25), along with 21 genera of Actinobacteria (e.g. Arthrobacter and Pseudonocardia) and 18 genera of Firmicutes (e.g. Alkaliphilus and Clostridium). The abundance of Nitrospira in mesic habitats is also consistent with our understanding of dry valley biogeochemistry, as these wetted areas often display increased ammonium availability (mean = 0.3 lg g1 dry soil; e.g. Barrett et al., 2009). Analysis of variance between geochemical groupings indicates a significantly reduced richness and diversity of bacterial communities at the phylum, order, and genus levels, as well as Chao1 estimated richness, from locations with high soil pH. Bacterial richness of high conductivity and mesic habitats were not significantly different. Correlations with soil properties suggest few parameters with potential explanatory power of phylum-level richness, but a very strong relationship between soil moisture and genus-level richness (Supporting Information, Table S1). Multiple linear regression (mixed stepwise selection) was used to examine these results further and confirms bacterial community richness is primarily explained by pH (adjusted R2 = 0.57; SOC and conductivity as co-parameters) at the phylum level, shifting to water (adjusted R2 = 0.91; chlorophyll a and conductivity as co-parameters) at the order level, and moisture alone (R2 = 0.91) at the genus level (Table S2). Van Horn et al. (2013) also found context-dependent roles for edaphic factors in explaining dry valley bacterial richness, in this case across kilometer spatial scales. Soil habitats of the McMurdo Dry Valleys have divergent properties that harbor distinct bacterial communities which in some cases have significantly lowered diversity (e.g. alkaline locations). The explanation of c. 60% of phylum-level richness by soil pH agrees with the continental-scale assessment of bacterial biogeography by Fierer & Jackson (2006). Adjusting the taxonomic resolution to a finer (order or genus) scale provided much more explanatory power here (91%). This result is logical considering the extensive diversity among prokaryotes at the phylum level and potential difficulty in explaining properties of any such complex group by only a few parameters. However, such increasing explanatory power with increasing taxonomic resolution may also suggest a shift from neutral to niche-based controls for bacteria at the genus level. Additional sequencing effort will be needed to understand the complex response among diverse microbial taxa and resulting diversity patterns to edaphic drivers, with particular attention needed to constrain potential variability caused by spatial relationships among communities. An important consideration for the maintenance and management of bacterial communities ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
and associated functions should therefore be the taxonomic level at which different soil properties may elicit effects (Martiny et al., 2013). As our understanding of the functional redundancy of microorganisms continues to be refined, information linking environmental conditions to specific taxa, or even communities of different inherent diversities, may provide avenues for the intentional selection of communities capable of desirable and predictable (e.g. low variability) functioning using knowledge of environmental controls.
References Adams BJ, Bardgett RD, Ayres E et al. (2006) Diversity and distribution of Victoria Land biota. Soil Biol Biochem 38: 3003–3018. Aislabie JS, Jordan S, Ayton J, Klassen JL, Barker GM & Turner S (2009) Bacterial diversity associated with ornithogenic soil of the Ross Sea region, Antarctica. Can J Microbiol 55: 21–36. Barrett JE, Gooseff MN & Takacs-Vesbach C (2009) Spatial variation in soil active-layer geochemistry across hydrologic margins in polar desert ecosystems. Hydrol Earth Syst Sci 13: 2349–2358. Caporaso JG, Kuczynski J, Stombaugh J et al. (2010) QIIME allows analysis of high- throughput community sequencing data. Nat Methods 7: 335–336. Cary CS, McDonald IR, Barrett JE & Cowan DA (2010) On the rocks: the microbiology of Antarctic Dry Valley soils. Nat Rev Microbiol 8: 129–138. Castle SC, Morrison CD & Barger NN (2011) Extraction of chlorophyll a from biological soil crusts: a comparison of solvents for spectrophotometric determination. Soil Biol Biochem 43: 853–856. Clarke KR (1993) Non-parametric multivariate analyses of changes in community structure. Aust J Ecol 18: 117–143. Edgar RC (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461. Eilers KG, Lauber CL, Knight R & Fierer N (2010) Shifts in bacterial community structure associated with inputs of low molecular weight carbon compounds to soil. Soil Biol Biochem 42: 896–903. Fierer N & Jackson RB (2006) The diversity and biogeography of soil bacterial communities. P Natl Acad Sci USA 103: 626–631. Fierer N & Lennon JT (2011) The generation and maintenance of diversity in microbial communities. Am J Bot 98: 439–448. Fierer N, Bradford MA & Jackson RB (2007) Toward an ecological classification of soil bacteria. Ecology 88: 1354– 1364. Geyer KM, Altrichter AE, Van Horn DJ, Takacs-Vesbach CD, Gooseff MN & Barrett JE (2013) Environmental controls over bacterial communities in polar desert soils. Ecosphere 4: art127.
FEMS Microbiol Ecol && (2014) 1–5
5
Bacterial communities of polar desert soil habitats
Giovannoni SJ, Delong EF, Schmidt TM & Pace NR (1990) Tangential flow filtration and preliminary phylogenetic analysis of marine picoplankton. Appl Environ Microbiol 56: 2572–2575. Hamady M, Walker JJ, Harris JK, Gold NJ & Knight R (2008) Error-correcting barcoded primers allow hundreds of samples to be pyrosequenced in multiplex. Nat Methods 5: 235–237. Hanson CA, Fuhrman JA, Horner-Devine MC & Martiny JBH (2012) Beyond biogeographic patterns: processes shaping the microbial landscape. Nat Rev Microbiol 10: 497–506. Lee CK, Barbier BA, Bottos EM, McDonald IR & Cary CS (2012) The Inter-Valley Soil Comparative Survey: the ecology of Dry Valley edaphic microbial communities. ISME J 6: 1046–1057. Martiny JBH, Bohannan BJM, Brown JH et al. (2006) Microbial biogeography: putting microorganisms on the map. Nat Rev Microbiol 4: 102–112. Martiny AC, Treseder K & Pusch G (2013) Phylogenetic conservatism of functional traits in microorganisms. ISME J 7: 830–838. Metting FB Jr (1994) DMSO extraction method for quantification of chlorophyll in soil. Part 2: Microbiological and Biochemical Properties (Bigham JM, ed), pp. 443–444. Soil Science Society of America, Inc., Madison, WI. Naether A, Foesel BU, Naegele V et al. (2012) Environmental factors affect Acidobacterial communities below the subgroup level in grassland and forest soils. Appl Environ Microbiol 78: 7398–7406. Niederberger TD, McDonald IR, Hacker AL, Soo RM, Barrett JE, Wall DH & Cary CS (2008) Microbial community composition in soils of Northern Victoria Land, Antarctica. Environ Microbiol 10: 1713–1724. Nkem JN, Virginia RA, Barrett JE, Wall DH & Li G (2006) Salt tolerance and survival thresholds for two species of Antarctic soil nematodes. Polar Biol 29: 643–651. Noy-Meir I (1973) Desert ecosystems: environment and producers. Annu Rev Ecol Syst 4: 25–51. Orwin KH, Wardle DA & Greenfield LG (2006) Ecological consequences of carbon substrate identity and diversity in a laboratory study. Ecology 87: 580–593. Quince C, Lanzen A, Davenport RJ & Turnbaugh PJ (2011) Removing noise from pyrosequenced amplicons. BMC Bioinformatics 12: 38. R Core Team (2011) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org. Smith JL, Barrett JE, Tusnady G, Rejtoe L & Cary CS (2010) Resolving environmental drivers of microbial community structure in Antarctic soils. Antarct Sci 22: 673–680. Sokol ER, Herbold C, Lee CK, Cary SC & Barrett JE (2013) Local and regional influences over soil microbial metacommunities in the Transantarctic Mountains. Ecosphere 4: art136.
FEMS Microbiol Ecol && (2014) 1–5
Stomeo F, Makhalanyane TP, Valverde A, Pointing SB, Stevens MI, Cary CS, Tuffin MI & Cowan DA (2012) Abiotic factors influence microbial diversity in permanently cold soil horizons of a maritime-associated Antarctic Dry Valley. FEMS Microbiol Ecol 82: 326–340. Strickland MS, Lauber C, Fierer N & Bradford MA (2009) Testing the functional significance of microbial community composition. Ecology 90: 441–451. Takacs-Vesbach C, Mitchell K, Jackson-Weaver O & Reysenbach A (2008) Volcanic calderas delineate biogeographic provinces among Yellowstone thermophiles. Environ Microbiol 10: 1681–1689. Takacs-Vesbach C, Zeglin L, Barrett JE, Gooseff MN & Priscu JC (2010) Factors promoting microbial diversity in the McMurdo Dry Valleys, Antarctica. Life in Antarctic Deserts and Other Cold Dry Environments: Astrobiological Analogs (Doran P, Lyons WB, & McKnight DM, eds), pp. 221–257. Cambridge University Press, Cambridge, UK. Treonis AM, Wall DH & Virginia RA (1999) Invertebrate biodiversity in Antarctic dry valley soils and sediments. Ecosystems 2: 482–492. Van Horn DJ, Van Horn ML, Barrett JE, Gooseff MN, Altrichter AE, Geyer KM, Zeglin LH & Takacs-Vesbach CD (2013) Factors controlling soil microbial biomass and bacterial diversity and community composition in a cold desert ecosystem: role of geographic scale. PLoS One 8: e66103. Wang Q, Garrity GM, Tiedje JM & Cole JR (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 73: 5261–5267. Zeglin LH, Dahm CN, Barrett JE, Gooseff MN, Fitpatrick SK & Takacs-Vesbach CD (2011) Bacterial community structure along moisture gradients in the parafluvial sediments of two ephemeral desert streams. Microb Ecol 61: 543–556. Zhou JZ, Xia BC, Treves DS, Wu LY, Marsh TL, O’Neil RV, Palumbo AV & Tiedje JM (2002) Spatial and resource factors influencing high microbial diversity in soil. Appl Environ Microbiol 68: 326–334.
Supporting Information Additional Supporting Information may be found in the online version of this article: Fig. S1. Rarefaction curves of 12 pyrosequenced samples. Table S1. Spearman nonparametric correlation matrix between soil parameters (square-root-transformed), Chao1 estimated richness, and bacterial community richness at the phylum (Phy R), order (Ord R), and genus (Gen R) taxonomic levels. Table S2. Multiple linear regression results for the prediction of phylum, order, and genus-level richness through mixed stepwise selection of model effects (a = 0.15).
ª 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
