Environmental Controls Over Bacterial Communities in Polar Desert Soils 1, 1 2 2 KEVIN M

Home , Hypolith

Environmental controls over bacterial communities in polar desert soils 1, 1 2 2 KEVIN M. GEYER, ADAM E. ALTRICHTER, DAVID J. VAN HORN, CRISTINA D. TAKACS-VESBACH, 3,4 1 MICHAEL N. GOOSEFF, AND J. E. BARRETT

1Department of Biological Sciences, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA 2Department of Biology, University of New Mexico, Albuquerque, New Mexico 87131 USA 3Department of Civil and Environmental Engineering, Pennsylvania State University, University Park, Pennsylvania 16802 USA

Citation: Geyer, K. M., A. E. Altrichter, D. J. Van Horn, C. D. Takacs-Vesbach, M. N. Gooseff, and J. E. Barrett. 2013. Environmental controls over bacterial communities in polar desert soils. Ecosphere 4(10):127. http://dx.doi.org/10.1890/ ES13-00048.1

Abstract. Productivity-diversity theory has proven informative to many investigations seeking to understand drivers of spatial patterns in biotic communities and relationships between resource availability and community structure documented for a wide variety of taxa. For soil bacteria, availability of organic matter is one such resource known to influence diversity and community structure. Here we describe the influence of environmental gradients on soil bacterial communities of the McMurdo Dry Valleys, Antarctica, a model ecosystem that hosts simple, microbially-dominated foodwebs believed to be primarily structured by abiotic drivers such as water, organic matter, pH, and electrical conductivity. We sampled 48 locations exhibiting orders of magnitude ranges in primary production and soil geochemistry (pH and electrical conductivity) over local and regional scales. Our findings show that environmental gradients imposed by cryptogam productivity and regional variation in geochemistry influence the diversity and structure of soil bacterial communities. Responses of soil bacterial richness to carbon content illustrate a productivity-diversity relationship, while bacterial community structure primarily responds to soil pH and electrical conductivity. This diversity response to resource availability and a community structure response to environmental severity suggests a need for careful consideration of how microbial communities and associated functions may respond to shifting environmental conditions resulting from human activity and climate variability.

Key words: Antarctic Dry Valleys; biogeography; environmental gradients; microbial ecology; productivity/diversity theory.

Received 15 February 2013; revised 17 July 2013; accepted 18 July 2013; final version received 17 September 2013; published 25 October 2013. Corresponding Editor: U. Nielsen. Copyright: Ó 2013 Geyer et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. http://creativecommons.org/licenses/by/3.0/ 4 Present address: Department of Civil and Environmental Engineering, Colorado State University, Fort Collins, Colorado 80523 USA. E-mail: [email protected]

INTRODUCTION ing a diverse microbial world exhibiting spatial patterns over environmental gradients spanning Despite early suggestions that microbial taxa meter, kilometer (Noguez et al. 2005, Zeglin et al. experience cosmopolitan distribution (Baas Beck- 2011) and regional to continental scales (Fierer ing 1934, Finlay 2002), recent evidence is reveal- and Jackson 2006, Yergeau et al. 2007, Bryant et

v www.esajournals.org 1 October 2013 v Volume 4(10) v Article 127 GEYER ET AL. al. 2008). Although geographic patterns in soil organic matter content and extreme ranges of microbial communities are now evident, the soil pH (.9.0) and salinity (.10,000 lS/cm) mechanisms driving them remain poorly under- (Bockheim 1997). Distinct biogeochemical gradi- stood (Soininen 2012). One promising avenue for ents span orders of magnitude in nutrient beginning to frame hypotheses behind microbial availability, major ion concentrations, and bio- biogeography is the application of macroecolog- mass (Barrett et al. 2004, Poage et al. 2008), ical theory, with which researchers may test characteristics of a landscape where abiotic whether controls over the spatial organization factors are the primary controls over the diversity of eukaryotic communities are equally appropri- and structure of microbial communities. Recent ate for microorganisms (Martiny et al. 2006, research conducted in this region has described Soininen 2012). responses of microbial communities to a number Environmental controls have been long recog- of abiotic drivers including: water availability nized as major drivers of a species’ presence/ (Zeglin et al. 2011); geochemistry (Lee et al. absence and abundance (Ricklefs and Schluter 2012); carbon concentration (Aislabie et al. 2009); 1994). For instance, productivity-diversity hy- or a combination of these factors (Niederberger et potheses predict that spatial or temporal varia- al. 2008, Smith et al. 2010, Stomeo et al. 2012). tion in resource availability (e.g., nitrogen, The anticipated response of dry valley biota to organic matter) influences communities by elic- environmental controls is also supported by iting niche-specialization of, and even competi- evidence from other ecosystems and experimen- tive exclusion by, particular species across ranges tal manipulations, which have demonstrated a of nutrient availability or productivity (Tilman community response to environmental variabil- 1982, Waide et al. 1999). Often a unimodal ity, such as positive influences of moisture levels (hump-shaped curve) relationship is observed on diversity (Zhou et al. 2002) and effects on between productivity and diversity, a result of community similarity due to carbon substrate positive effects of resource availability and (resource) diversity (Orwin et al. 2006, Eilers et negative effects of competition along a gradient al. 2010). of increasing ecosystem production (Michalet et Here we focus on the distinct effects of al. 2006). Field surveys (Abramsky and Rose- resource availability on both bacterial communi- nzweig 1984, Mittelbach et al. 2001) and exper- ty diversity and structure, as well as the imental resource manipulations (Silvertown et al. influences of geochemical severity (pH and 2006, Chase 2010) support these predictions for a salinity). To evaluate these abiotic drivers we variety of taxa in terrestrial and aquatic ecosys- studied soils representing a productivity gradient while additionally capturing an extensive range tems. Environmental severity (e.g., extreme pH in geochemical conditions. Based on evidence or salinity) may also generate a range of physical from productivity/diversity theory and other conditions that influences productivity, habitat field surveys, we hypothesize: (1) bacterial suitability, and community structure of organ- community diversity will exhibit a positive isms (Freckman and Virginia 1997, Lee et al. relationship with resource availability (organic 2012). Given the universal constraints to biolog- carbon and/or water) and a negative association ical diversity and activity which arise from both with geochemical severity, while (2) bacterial resource limitations and environmental severity, community structure will be influenced by both these mechanisms likely operate together to resource availability and geochemical severity. In control macro- and microorganismal community addition we quantify a gradient of primary structure alike (Prosser 2007). Indeed, examples production (chlorophyll a) for Antarctic soils of significant productivity/diversity relationships and examine the influence of aboveground have been reported for microorganisms (Horner- productivity on belowground biology and bio- Devine et al. 2003, Smith 2007, Logue et al. 2012). geochemistry. Antarctica’s polar deserts are a model system in which to address questions of controls over microbial biogeography. Resident organisms are sensitive to resource availability, and abiotic factors in general, given the exceptionally low

v www.esajournals.org 2 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Fig. 1. Location of sixteen regional sampling sites in Wright and Taylor Valley of the McMurdo Dry Valleys, Antarctica. See Table 1 for an explanation of site labels.

METHODS ents (Barrett et al. 2009) and promoting localized hotspots of primary production (Barrett et al. Site description 2006). The McMurdo Dry Valleys are an ice-free polar The soil food web is simple, microbially- desert in Southern Victoria Land, Antarctica. dominated, and at the base composed of various Aridity (generally less than 10cm of annual prokaryotic, photosynthetic bacteria (Families precipitation), temperature (mean annual be- Nostocaceae, Oscillatoriaceae), eukaryotic algae tween 168Cto218C), and low soil organic carbon availability (0.03% average by weight) (Phyla Chlorophyta, Bacillariophyta), and fewer together constrain the diversity and activity of than 10 species of moss (Family Bryaceae) native biota (Kennedy 1993, Burkins et al. 2000, (Broady 1996, Seppelt and Green 1998) that Hogg et al. 2006). Dry permafrost soils are poorly associate in cryptogamic communities. Several weathered and composed of .90% sand-sized species each of tardigrades, rotifers, and nema- particles with ice cement occurring within 0.5 m todes represent the apex of the soil foodweb of the surface (Ugolini and Bockheim 2008). (Freckman and Virginia 1997, Adams et al. 2006). Salinity and pH are generally high, a conse- Molecular analyses of microbial communities quence of limited vertical water movement from mineral soils have revealed higher than through soil layers that results in the accumula- expected diversity in comparison to non-polar tion of weathered carbonates and aerially-depos- soils, with an abundance of heterotrophic bacte- ited salts, particularly on old, exposed surfaces ria (;90% of total isolates) (Cary et al. 2010, (Bockheim 1997). Only during the austral sum- Takacs-Vesbach et al. 2010). This suggests a mer (November–February) does 24-hour incident radiation generate above-freezing temperatures, carbon (energy) limitation to terrestrial microbi- inducing melt and stimulating biological activity ota, although extremes in pH and conductivity (Fountain et al. 1999, McKnight et al. 2007). reported for this region can also exceed levels Liquid water creates environmental gradients known to limit the abundance and distribution of across fine and landscape scales by altering soil microorganisms, here and elsewhere (Fierer and geochemistry, e.g., environmental severity gradi- Jackson 2006, Poage et al. 2008).

v www.esajournals.org 3 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Table 1. Average soil characteristics of regional sampling sites (n ¼ 3) and ad hoc productivity classes: soil organic carbon (SOC, mg/kg dry soil); total nitrogen (TN, mg/kg dry soil); moisture (% gravimetric); electrical conductivity (EC, lS/cm). Values are means with SD in parentheses.

Site ID or class Landscape location SOC TN Moisture pH EC 1 Stream margin 260.09 (61.62) 35.65 (10.46) 6.73 (6.65) 8.37 (0.75) 186.18 (221.79) 2 Hypolith region 553.07 (76.93) 65.59 (12.22) 0.26 (0.01) 9.43 (0.12) 138.90 (22.27) 3 Pond margin 337.47 (198.91) 41.74 (9.92) 12.95 (0.51) 8.31 (0.46) 108.63 (37.88) 4 Stream margin 919.87 (477.46) 95.09 (37.81) 9.59 (3.36) 8.74 (0.14) 117.23 (27.71) 5 Pond margin 243.34 (58.78) 39.32 (8.27) 13.37 (2.28) 7.70 (0.34) 70.96 (53.55) 6 Pond margin 197.52 (61.14) 31.30 (5.72) 7.78 (4.50) 7.26 (0.06) 30.30 (8.42) 7 Pond margin 730.31 (111.60) 85.90 (12.11) 14.75 (0.74) 8.35 (0.08) 59.14 (18.88) 8 Snowpack margin 538.48 (212.98) 65.45 (25.18) 11.49 (1.88) 8.37 (0.05) 47.88 (5.04) 9 Stream margin 597.53 (157.46) 68.77 (20.19) 7.69 (4.34) 8.31 (0.37) 57.12 (13.06) 10 Pond margin 248.32 (41.89) 39.86 (4.97) 9.54 (5.14) 7.43 (0.17) 21.14 (3.33) 11 Stream margin 1638.15 (917.61) 192.87 (105.30) 5.62 (3.16) 8.11 (0.31) 57.17 (39.51) 12 Relict stream channel 504.63 (66.45) 78.09 (37.68) 7.18 (1.89) 8.51 (0.23) 59.93 (9.02) 13 Pond margin 1076.11 (145.21) 104.03 (14.69) 11.78 (3.24) 8.38 (0.17) 995.50 (663.00) 14 Pond margin 2111.03 (587.73) 299.07 (83.72) 11.52 (3.00) 8.50 (0.16) 197.80 (79.17) 15 Relict stream channel 577.13 (227.28) 73.40 (28.09) 4.69 (3.85) 8.00 (0.32) 44.09 (28.21) 16 Snowpack margin 1196.38 (667.70) 155.60 (91.12) 12.68 (2.35) 8.41 (0.23) 64.30 (3.22) Ad hoc productivity classes Low (n ¼ 8) 450.03 (271.83) 53.85 (26.05) 4.86 (5.30) 8.62 (0.65) 318.15 (524.20) 318.15 (524.20) Medium (n ¼ 15) 752.75 (571.85) 93.42 (84.53) 8.90 (4.69) 8.43 (0.37) 70.00 (37.98) 70.00 (37.98) High (n ¼ 9) 1543.48 (529.40) 191.07 (73.90) 11.33 (3.56) 8.28 (0.30) 228.75 (355.77) 228.75 (355.77)

SAMPLE COLLECTION ration ranged from 0.3 km to 60 km. To capture fine-scale gradients of soil production, 3 local 2 We collected samples from 16 regional sites (8 plots were chosen within a 25 m area of each each in Taylor and Wright Valleys; Fig. 1, Tables 1 regional site on the basis of surface cryptogam and 2) chosen with the intent of including cover. Thus, local plots (n ¼ 48) varied from dry, locations exhibiting the full range of soil primary mineral soil without obvious surface production production for these valleys. Regional site sepa- (e.g., Site 1) to intermittently saturated zones

Table 2. Average soil characteristics of regional sampling sites (n ¼ 3) and ad hoc productivity classes: microbial biomass carbon (MBC, mg/kg dry soil); ratio of beta- to alpha-glucosidase enzyme activity (bgluc/agluc,g SOC1kg dry soil1); invertebrate density (Total inverts, number of individuals/kg dry soil); TRFLP ribotype richness (Bacterial richness). Values are means with SD in parentheses.

Site ID or class Landscape location MBC bgluc/agluc Total inverts Bacterial richness 1 Stream margin 9.26 (4.68) 0.81 (0.22) 156 (177) 19.67 (2.52) 2 Hypolith region 18.62 (3.41) 1.05 (0.42) N/A 19.33 (0.58) 3 Pond margin 6.25 (1.98) 0.64 (0.02) 137 (137) N/A 4 Stream margin 24.43 (11.14) 0.99 (0.06) 432 (116) N/A 5 Pond margin 3.87 (2.15) 1.03 (0.16) 193 (244) N/A 6 Pond margin 7.97 (2.39) 0.91 (0.38) 460 (210) N/A 7 Pond margin 25.16 (7.46) 1.30 (0.20) 613 (96) 25.67 (1.53) 8 Snowpack margin 21.06 (3.97) 1.10 (0.16) 1418 (888) 23.33 (3.21) 9 Stream margin 19.35 (8.89) 1.00 (0.42) 1484 (1335) 26.00 (1.00) 10 Pond margin 7.39 (1.54) 1.23 (0.13) 482 (407) N/A 11 Stream margin 68.75 (68.71) 1.55 (0.39) 1233 (251) 25.67 (2.52) 12 Relict stream channel 24.77 (5.39) 1.05 (0.04) 476 (500) 22.00 (0.00) 13 Pond margin 35.15 (5.38) 1.22 (0.16) 2742 (2527) 23.00 (1.00) 14 Pond margin 53.62 (1.45) 1.43 (0.22) 5106 (1189) 25.67 (1.15) 15 Relict stream channel 26.83 (24.53) 1.05 (0.12) 1092 (154) 24.33 (1.15) 16 Snowpack margin 54.76 (43.84) 1.01 (0.17) 1972 (2260) 24.00 (4.00) Ad hoc productivity classes Low (n ¼ 8) 14.16 (8.89) 0.89 (0.18) 415 (462) 21.50 (3.38) 14.16 (8.89) Medium (n ¼ 15) 23.61 (9.96) 1.13 (0.25) 1221 (1106) 23.67 (2.41) 23.61 (9.96) High (n ¼ 9) 63.74 (36.58) 1.38 (0.29) 3218 (2081) 25.22 (2.17) 63.74 (36.58) Note: N/A means data unavailable.

v www.esajournals.org 4 October 2013 v Volume 4(10) v Article 127 GEYER ET AL. along stream, lake, and snowpack margins min. Samples were then centrifuged at 4000 rpm supporting dense cryptogamic mats composed for 15 min. to pelletize particulates and a portion primarily of either cyanobacteria or moss (e.g., of the supernatant used for spectrophotometric Site 16). One location (Site 2) consisted of rock- analysis using a Shimadzu UV-1601 UV-VIS associated cryptogams (e.g., hypoliths; Pointing spectrophotometer (Shimadzu, Columbia, MD, et al. 2009) found locally within an otherwise dry, USA). Absorbance at both 665 nm and 750 nm upland landscape. was recorded before and after dilute-acid degra- Samples were collected from local plots in dation of chlorophyll a to account for phaeopig- December 2010 to characterize the following: (1) ment content. All results were standardized to surface cryptogam chlorophyll a and (2) below- dry weight of starting material and thus ex- ground soil and associated bacterial communi- pressed in w/w of chlorophyll a per mass of ties. A rectangular prism of cryptogamic mat of starting material. A second round of extraction known surface area (,0.4 m2) was separated performed on a subset of samples indicated an from the mineral soil and collected into an average first-round chlorophyll extraction effi- opaque Nalgene bottle. The top 1 mm of mineral ciency of 85%, leaving many second round soil exposed by the removal of surface mat was extracts too dilute for spectrophotometry. One additionally collected for chlorophyll a analysis, extraction was used for all future samples. and in the absence of visual mat this top 1mm Belowground biogeochemical parameters are provided the sole estimate of producer biomass. reported per unit dry soil mass. A 1:2 and 1:5 For each local plot (n ¼ 48) this collection scheme soil/water slurry was used to assess soil pH and was repeated in triplicate within a 2.5 m2 area. electrical conductivity, respectively, following Finally, ;500 g of bulk mineral soil was collected standard procedures developed for this region and pooled from beneath the areas sampled for (Nkem et al. 2006). Soil water content was cryptogams (at each local plot) to a depth of 5 determined gravimetrically by oven-drying for cm. From this composite sample, ;10 g of soil 48 hours at 1058C. Total soil organic carbon was preserved in-field with a sucrose lysis buffer (SOC) and total nitrogen were estimated from for nucleic acid stabilization (Mitchell and ;300 mg of ground, dried, and acidified soil Takacs-Vesbach 2008). All samples were frozen using a FlashEA 1112 NC Elemental Analyzer within 12 hours of collection at 208C, with (CE Elantech, Lakewood, NJ, USA) (Barrett et al. nucleic acids moved to 808C storage within 48 2009). A 1:5 soil/deionized water slurry was hours of collection. All samples were returned to centrifuged and the supernatant analyzed for the Blacksburg, VA campus of Virginia Tech for major soluble ions using standard ion chroma- further analysis. tography methods (Thermo Scientific Dionex, 1 Sunnyvale, CA, USA). Soil nitrate (NO3 -N) Soil productivity and biogeochemistry concentrations were estimated using a 2 M KCl Chlorophyll a concentrations were measured soil extraction and subsequent Lachat QuikChem on composited surface mats and mineral soils (1 8500 Flow Injection Analyzer (Lachat Instru- mm layer) as a proxy for soil productivity. ments, Loveland, CO, USA) assay of centrifuged Although not a direct measure of production, extracts (QuikChem Methods 10-107-04-1-B). this remains a quick and efficient technique for Chloroform-labile carbon was used as an an index of potential productivity in soil and indication of soil microbial biomass and involved other habitats, as chlorophyll a is subject to rapid a 5-day fumigation of soil samples with gaseous photochemical degradation and is therefore chloroform (Cheng and Virginia 1993). Paired representative of living or only recently senesced fumigated and non-fumigated samples were then tissue (Metting 1994). Chlorophyll was extracted extracted with a 0.5 M K2SO4 solution and final from 2 mm sieved material using a dimethylsulf- extracts analyzed for total organic carbon using a oxide (DMSO) extraction procedure under low- OI Model 1010 Total Organic Carbon Analyzer light conditions adapted from Metting (1994) and (OI Analytical, College Station, TX, USA), where Castle et al. (2011). Sample extractions (5 mL final chloroform-labile carbon was calculated as DMSO:1.5 g sieved material standard ratio) took the difference between fumigated and non- place at 658C for 1 h with 20 s vortexing every 30 fumigated total soil organic carbon. Soil inverte-

v www.esajournals.org 5 October 2013 v Volume 4(10) v Article 127 GEYER ET AL. brates were extracted using a modified sugar involves use of a fluorescent primer during DNA centrifugation method (Freckman and Virginia amplification, amplicon digestion with one or 1993), enumerated by microscopy into the three more restriction enzymes to produce DNA major taxonomic groups (rotifers, tardigrades, fragments of varied length, and fragment sepa- and nematodes), and later pooled into total ration/quantification via capillary electrophore- invertebrate abundances. sis. Fragment relative abundance provides an Soil extracellular enzyme activity was assayed estimate of community diversity and structure targeting a- and b-glucosidase to characterize (Thies 2007). major organic matter degrading enzymes. These DNA was extracted from soils using a modi- hydrolytic enzymes are produced by microor- fied cetyltrimethylammonium bromide (CTAB) ganisms to initiate the decomposition of complex procedure that involves a mixture of 1% CTAB, extracellular organic compounds into simple 10% sodium dodecyl sulfate, phenol/chloroform/ units (e.g., glucose) that are easily transported isoamyl alcohol (pH ¼ 7.5), lysozyme (0.2 lg/lL), across the cell membrane. a-1,4 glycosidic bonds and proteinase K (20 lg/lL) with ;0.75 g soil. are common in starch and simple polysaccha- Extracted DNA was resuspended in Tris buffer rides, while b-1,4 glycosidic bonds typify more (pH¼ 8.0) and quantified via spectrophotometry structurally-complex compounds such as cellu- (NanoDrop 2000; Thermo Scientific, Wilmington, lose and chitin. The relative ratio of b/a-glucosi- DE, USA). PCR amplification took place in dase enzyme activity, therefore, may serve as an triplicate (25 lL reaction volume) using a indication of the relative complexity of soil standard 2 lL of diluted template, 5 units/lLof organic matter pools (Sinsabaugh et al. 2010). Taq Hot Start Polymerase (Promega Corporation, Potential enzyme activity was measured using Madison, WI, USA), and the universal bacterial 0.5 g soil incubations with the labeled substrates primers 8F (5’-AGAGTTTGATCMTGGCTCAG- 4-methylumbelliferyl(MUB)-a-D-glucopyrano- 3’) and 519R (5’-ACCGCGGCTGCTGGCAC-3’), side and 4-MUB-b-D-glucopyranoside in the the forward primer labeled with a 5’ 6-FAM fluorophore (Integrated DNA Technologies, Cor- presence of 50 mM NaHCO3 buffer (pH ¼ 8.2) following the methods of Zeglin et al. (2009). alville, IA, USA). Amplification was optimized Triplicate samples were incubated at room for concentrations of MgCl2 (2.5 mM per reac- l temperature on a platform shaker (250 rpm) for tion), BSA (1 L/reaction), annealing temperature 8 a minimum of 2 h and enzyme-induced fluores- (53 C), and final extension time (5 min). Ampli- cence measured by excitation (360 nm) and fication replicates were pooled and cleaned using emission (465 nm) using a Tecan SpectraFluor a QuickClean II PCR Extraction Kit (GenScript, Piscataway, NJ, USA). Successful amplifications Plus plate reader (Tecan, Mannedorf, Zurich, (32 of 48 total) were digested with HaeIII (New Switzerland). In addition to sample incubations, England BioLabs, Ipswich, MA, USA) in tripli- control (buffer only), substrate (substrate þ cate (20 lL reaction volume) for 3 h at 378C buffer), and standard (standard þ buffer) refer- following manufacturer’s suggested protocols. ences were analyzed to account for other sources Digestion replicates were then pooled and of fluorescence. Final activity was normalized to cleaned using GenScript extraction kits. Frag- sample soil organic carbon content and ex- ment separation/quantification took place in pressed as activity (nmol)h1g SOC1. quadruplicate with an ABI 3130xl Genetic Ana- lyzer (Applied Biosystems, Carlsbad, CA, USA) Bacterial communities and fragments binned using the GeneMarker A terminal restriction fragment length poly- software AFLP protocol. Resulting sample pro- morphism (TRFLP) procedure was chosen in files were standardized using the procedures order to provide an index of bacterial taxonomic outlined in Dunbar et al. (2001) to produce both a richness and community structure in the soils of consensus profile among replicates and final our local plots. TRFLP is a largely automated normalization of all sample profiles by total process suited for high sample through-put and sample fluorescence. remains particularly useful for tracking changes in microbial community structure over time and space (Schutte et al. 2008). The TRFLP method

v www.esajournals.org 6 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Fig. 2. Productivity (chlorophyll a) gradient observed for 48 local plots, log scale. Error bars represent standard deviation. Gradient divided by horizontal lines into three productivity classes created using k-means non- hierarchical clustering. Regional sites ranked by increasing maximum chlorophyll a content. Closed circles indicate locations where DNA could be extracted (32 of 48 plots).

Data analysis using a correlation distance metric and scores of All data analysis was restricted to a subset of the two primary eigenvectors were plotted. Non- 32 plots where DNA extraction was successful, metric multidimensional scaling (nMDS) analysis and thus bacterial community information avail- was performed using bacterial community data able. Plots were categorized into productivity and the Bray-Curtis distance metric with axes classes based on levels of surface chlorophyll a rotated to principle components. The final nMDS using k-means non-hierarchical clustering with ordination represents a plot of site scores for the JMP statistical software, specifying three conser- two primary axes. Canonical correspondence vative a priori groups (Fig. 2). Univariate analysis (CCA) was used to ordinate weighted statistics (simple regressions, partial regressions, average, scaled (by eigenvalue) site scores under Spearman rank correlations) were performed on the constraints of multiple linear regression with square-root transformedbiogeochemicaldata environmental variables. Multi-response permu- using JMP statistical software to explore the tation procedure (MRPP) was used to assess relatedness among measured variables (JMP, differences among groups of response variables Version 9, SAS Institute Inc., Cary, NC, USA). in ordination results. TRFLP results of fragment MANOVA and multivariate tests/ordinations abundance was used for calculations of bacterial were performed using R statistical freeware (R richness and Shannon-Weiner diversity, and were Development Core Team). Mantel tests involved square-root transformed to reduce the distortion- comparisons of three distance matrices (50000 al influence of high abundance taxa. permutations each): community (TRFLP) data transformed using a Bray-Curtis distance metric; RESULTS decimal degree geographic data which was transformed using the earth.dist (fossil) R pack- Soil productivity and biogeochemistry age (Vavrek 2011); and soil biogeochemistry data Soil biogeochemical properties exhibited or- (scaled and centered) transformed using a ders of magnitude variation across all 48 plots euclidean distance metric. Principle components (Tables 1 and 2; Appendix: Table A1). For analysis (PCA) was performed of soil properties example, chlorophyll a concentrations exhibited

v www.esajournals.org 7 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Table 3. Spearman rank correlations among soil variables.

Variable SOC TN Moisture pH EC MBC blug/agluc Total inverts Bacterial richness CHLA 0.80*** 0.80*** 0.51** 0.27 0.08 0.84*** 0.58*** 0.68*** 0.48** SOC 0.99*** 0.39* 0.29 0.18 0.90*** 0.65*** 0.63*** 0.48** TN 0.39* 0.27 0.16 0.90*** 0.68*** 0.61*** 0.50** Moisture 0.29 0.07 0.43* 0.33 0.25 0.35 pH 0.548** 0.29 0.08 0.07 0.34 EC 0.03 0.17 0.13 0.22 MBC 0.58*** 0.57** 0.45* blug/agluc 0.66*** 0.42* Total inverts 0.38* Notes: CHLA ¼ Chlorophyll a concentration (lg/g dry soil). All other abbreviations are as in Tables 1 and 2. *P 0.05; **P 0.01; ***P 0.001. a gradient of soil productivity spanning more predictor of TRFLP bacterial richness in a model than three orders of magnitude (Fig. 2) and are also including pH, conductivity, chlorophyll, comparable to those reported for hot desert soil microbial biomass carbon, and moisture (stan- biological crusts of North America (;5–10 lg dard coefficient ¼ 0.84, all variance inflation chla/g soil) (Castle et al. 2011), although the range factors ,5.8). here is considerably greater. Organic carbon Because the influence of soil productivity on (mean ¼ 733.1 mg/kg dry soil) and total nitrogen thesubsurfaceenvironmentwasofspecific (mean ¼ 92.0 mg/kg dry soil) concentrations also interest for this study, we used k-means cluster- exhibited wide variation, again ranging over an ing to bin local plots into productivity classes to order of magnitude, generating average molar serve as predictors of underlying soil properties C:N ratios of 9.2 6 1.7 SD. and microbial diversity in multivariate analyses. Mean a- and b-glucosidase activities were 3590 Fig. 2 depicts the results of this clustering, where and 4010 nmolh1g SOC1, respectively. These 8 plots grouped below 2 lg chla/g dry material values are higher than previously reported for (‘low productivity’), 15 plots between 2-35 lg dry mineral soils in the McMurdo Dry Valleys chla/g dry material (‘medium productivity’), and (Zeglin et al. 2009) and are more comparable to 9 plots above 35 lg chla/g dry material (‘high those found in semi-arid deserts (Zeglin et al. productivity’). With productivity class and re- 2007). Total invertebrate densities averaged 1200 gional site location as predictors, a multivariate individuals/kg dry soil, but ranged from none ANOVA (MANOVA) was used to examine the detected to over 6300 individuals/kg dry soil, amount of variability in soil properties which similar to modal densities reported by others in could be explained by these factors and their similarly productive Antarctic soil environments interaction. Results indicate that ad hoc soil (Barrett et al. 2006, Simmons et al. 2009). productivity classes explain very significant Microbial biomass averaged 25.5 mg/kg dry soil, levels of variation in organic carbon, total within the range previously reported for dry nitrogen, microbial biomass, and total inverte- valley soils (Barrett et al. 2006). brates (all p 0.001), while enzyme activity Correlations among these variables indicate a ratios and bacterial (TRFLP) richness were also strong relationship between biological parame- significantly constrained (p 0.01) (Table 4). In ters such as chlorophyll concentration, soil contrast, differences in geochemical parameters organic carbon, total nitrogen, microbial biomass such as pH and conductivity were associated carbon, enzyme activity, invertebrate abundance, with regional variation among the sites in Taylor and TRFLP bacterial richness (Table 3). However, and Wright Valley. only chlorophyll was significant (using partial regression analysis) in predicting soil carbon Bacterial community diversity and structure concentrations (standard coefficient ¼ 0.68, all Soil DNA amplification was successful for a variance inflation factors ,1.8) in a model also subset (32) of all 48 local plots, dictated primarily considering pH, conductivity, and moisture. In by levels of microbial biomass (multiple logistic turn, soil organic carbon was the only significant regression, p ¼ 0.012). Final TRFLP results

v www.esajournals.org 8 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Table 4. MANOVA results using productivity class (n ¼ 3) and regional site location (n ¼ 11) as predictors of common soil properties.

bgluc/ Source df SOC TN pH EC MBC agluc Total inverts Bacterial richness Productivity class 2 73.40*** 87.85*** 1.95 5.04* 20.76*** 7.60** 16.14*** 11.33** Site 10 8.23*** 12.93*** 2.76* 5.36** 0.48 1.03 4.02* 4.98** Interaction 6 7.16** 10.06*** 0.43 0.58 1.77 0.78 0.38 2.27 Notes: Abbreviations are as in Tables 1 and 2. *P 0.05; **P 0.01; ***P 0.001. indicate an average bacterial ribotype richness of 0.0001), while the correlation between communi- 23.4 and an average Shannon Index (H’) of 2.5 ties and geographic distance declined (Mantel’sR which is comparable, although somewhat lower, ¼ 0.173, p ¼ 0.018). Finally, a distance decay plot than those of Fierer and Jackson’s (2006) global using simple linear regression (data not shown) assessment of microbial biodiversity. A Mantel suggests that geographic separation explains test was performed to correlate distance matrices only a small proportion of the variance in of bacterial community similarity with both bacterial community similarity (r2 ¼ 0.044, p , geography (spatial distance) and soil properties 0.0001). (environmental distance). The results of these Principle components analysis (PCA) was used tests indicate a much stronger correlation of to examine how soil habitats differ with respect communities with soil properties than with to the primary soil properties of water content, geography (Mantel’sR¼ 0.542, p , 0.0001; pH, electrical conductivity, organic carbon, and Mantel’sR¼ 0.210, p ¼ 0.006, respectively). When total nitrogen (Fig. 3). PC1 (eigenvalue ¼ 1.53) the influences of both spatial proximity and soil and PC2 (eigenvalue ¼ 1.15) together were able to properties are controlled in these analyses (e.g., constrain 73.4% of the variation in soil properties, partial Mantel tests), the strength of correlation with little additional explanation from addition remained very high between bacterial communi- of further axes (eigenvalues , 1.0) (McCune and ties and soil properties (Mantel’sR¼ 0.533, p , Grace 2002). Correlations of soil properties with

Fig. 3. PCA ordination of the soil properties moisture, soil organic carbon (SOC), total nitrogen (TN), electrical conductivity, and pH for 32 local plots. Biplot of soil properties, as correlated with major axes, is overlain. Labels indicate productivity classes determined via clustering of local plots by soil surface chlorophyll a concentrations.

v www.esajournals.org 9 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Fig. 4. nMDS ordination of square-root transformed relative abundance bacterial (TRFLP) data for 32 local plots. Labels indicate productivity classes determined via clustering of local plots by chlorophyll a concentrations. Biplot of soil properties moisture, soil organic carbon (SOC), total nitrogen (TN), electrical conductivity, and pH (as correlated with major axes) is overlain. the primary axes indicate a strong correlation of (CCA) was used to examine the direct relation- PC1 with soil organic carbon (r ¼0.92), total ship of bacterial communities to the soil proper- nitrogen (r ¼0.89), moisture (r ¼0.66), and pH ties of organic carbon, pH, and electrical (r ¼ 0.50). PC2 was most strongly correlated with conductivity, chosen based on their lack of pH (r ¼0.73) and conductivity (r ¼0.57). multicollinearity (variance inflation factors , Multi-response permutation procedure (MRPP) 1.1) and strong correlation with nMDS axes of the ordination indicates a significant difference (Fig. 5). Five hundred test permutations yielded between points when grouped by the 3 ad hoc strong significance (p ¼ 0.002). The proportion of productivity classes (A ¼ 0.251, p ¼ 0.001). inertia (total variance) captured by constrained A non-metric multidimensional scaling axes was low (18.8%), and eigenvalues for the (nMDS) analysis was performed using TRFLP first two axes were also small (0.543, 0.288, bacterial relative abundance data (Fig. 4). A respectively). Correlation analysis indicates stable solution was achieved within 20 iterations CCA1 strongly related to conductivity (r ¼ of the test (stress ¼ 0.178). Resulting axes were 0.94), while CCA2 is strongly related to pH (r ¼ then correlated with soil environmental proper- 0.73) and soil organic carbon (r ¼ 0.62). ties to assess which soil characteristics were associated with the spread of ordinated commu- DISCUSSION nities. Axis 1 appears most strongly related to electrical conductivity (r ¼ 0.87), while Axis 2 was Heterogeneous biogeochemical conditions are negatively related to a number of properties a common characteristic of arid ecosystems associated with highly-productive plots, namely (Aguiar and Sala 1999, Wall and Virginia 1999). moisture (r ¼0.55) and soil organic carbon (r ¼ Such variation, particularly with respect to 0.48), along with a positive relationship to pH (r productivity and resource availability, has been ¼ 0.51). As with PCA, MRPP found significance widely used in macroecological research to among plots when grouped by productivity class examine the distribution and structure of both (A ¼ 0.042, p ¼ 0.001). plant and animal communities and to test Finally, canonical correspondence analysis ecological theory explaining patterns in biodiver-

v www.esajournals.org 10 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Fig. 5. CCA ordination of 32 local plots using weighted average site scores. Labels signify productivity classes determined via clustering of local plots by chlorophyll a concentrations. Biplot of soil properties soil organic carbon (SOC), electrical conductivity, and pH is overlain to emphasize the relationship between site bacterial communities and environment. sity (Abramsky and Rosenzweig 1984, Waide et scape history and soil development (Table 4), as al. 1999, Mittelbach et al. 2001). In the McMurdo has been previously described by others (Bock- Dry Valleys, productivity gradients associated heim 1997, Ugolini and Bockheim 2008). Similar with photosynthetic cryptogams are an impor- conclusions were drawn by Barrett et al. (2004) tant source of resources supporting soil organ- for biological communities in the Dry Valleys, isms (Simmons et al. 2009) and contribute to the where scale-dependent variation in soil proper- observed spatial patterns in diversity and com- ties (e.g., salinity, pH, and organic matter) munity structure of bacterial communites de- significantly influenced spatial variation in in- scribed here. Surface chlorophyll associated with vertebrate communities. a productivity gradient spanning both local and Bacterial community diversity was significant- regional scales was significantly correlated with ly correlated to soil organic carbon content subsurface soil properties including microbial suggesting that organic matter is a primary biomass carbon, total nitrogen, and invertebrate resource limitation to soil microbes in this system abundance. Chlorophyll was also the best pre- (Fig. 6). This relationship and other recent dictor of soil organic carbon, a vital resource in findings (Horner-Devine et al. 2003, Langen- this energy-limited ecosystem. This demonstrates heder and Prosser 2008, Logue et al. 2012) the linkage between above- and below-ground illustrate signficant productivity-diversity rela- processes and soil properties in driving patterns tionships for microbial communities. Although of diversity, as has been noted for numerous the polynomial relationship we found does not temperate systems (Wardle et al. 2004). conform to the unimodal curve often reported for Variation in biological properties such as total diversity-productivity studies of macroorgan- invertebrates, microbial biomass and diversity of isms (Waide et al. 1999, Mittelbach et al. 2001), bacterial communities was well explained by the the gradient we describe may represent a more stratification of study plots into productivity restricted range of resource availability, insuffi- classes determined by clustering. Variation in cient to support highly competitive copiotrophic geochemistry (e.g., pH, conductivity) was pri- bacterial taxa (Fierer et al. 2007) and the marily related to regional differences in land- competitive exclusion of other taxa as observed

v www.esajournals.org 11 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Fig. 6. Polynomial regression relating bacterial (TRFLP) richness to soil organic carbon (SOC) concentrations (mg/kg dry soil), square-root transformed. r2 ¼ 0.269, p ¼ 0.028. in more productive ecosystems. Dry valleys relationship with resources such as organic zones of high productivity are also likely carbon or water (Figs. 4 and 5). Thus, although composed of increasingly diverse primary pro- resource availability has a measurable influence ducer assemblages (e.g., mosses, chlorophytes, on structure, the most divergent bacterial com- diatoms and cyanobacteria) (Cary et al. 2010), the munities are primarily associated with extreme activity of which may increase the structural geochemical conditions (particularly when elec- complexity of the soil organic matter pool. trical conductivity exceeds 1000 lS/cm and pH is Indeed, we observed a significant increase in greater than 9.40). Such values may represent the activity of complex carbon acquiring en- threshold levels beyond which specialized taxa zymes (increasing ratio of b/a glucosidase activ- predominate. Analogous thresholds in geochem- ity) in the most productive habitats. Resource istry have been shown to influence the presence (carbon substrate) richness may play an impor- and absence of invertebrate species in this region tant role in facilitating the greater taxonomic (Treonis et al. 1999, Poage et al. 2008). diversity of microbial communities reported for Our results demonstrate that bacterial com- such locations (e.g., Grayston et al. 1998, Orwin munity dynamics are driven primarily by varia- et al. 2006), but further research will be needed to tion in soil properties (i.e., organic matter, pH determine the relative effects of resource quantity and salinity), however a weaker (yet still and quality (i.e., resource diversity) on microbial significant) effect of geographic distance is diversity and activity in such systems. detectable using Mantel and partial Mantel tests. In contrast to diversity, variation in bacterial A linear regression of geographic distance versus community structure appears to be more strong- community similarity indicates a weak, yet ly influenced by geochemical properties, such as significant, pattern of distance decay (likely pH and conductivity, than resource availability. inflated by the high number of pairwise obser- Ordination results (both nMDS and CCA) sug- vations, n ¼ 496). Taken together these results gest that although significant differences exist in suggest that although geographic processes may communities when grouped by ad hoc produc- play a significant role in determining bacterial tivity classes, the association of multivariate axes community structure in dry valley soils, for the with geochemistry (particularly electrical con- locations observed here the influence of local soil ductivity and pH) is generally stronger than the conditions appears to be the primary driver.

v www.esajournals.org 12 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Because sampling was performed specifically to was supported by the National Science Foundation’s encompass extremes in both spatial distance and Office of Polar Programs (grants #1027284 and soil conditions, this may provide evidence for 0838922). predominantly local, environmental controls over the distribution of dry valley microorgan- LITERATURE CITED isms. Abramsky, Z., and M. L. Rosenzweig. 1984. Tilman’s A bacterial diversity response to resource predicted productivity–diversity relationship availability, and community structure response shown by desert rodents. Nature 309:150–151. to environmental severity, mirrors effects dem- Adams, B. J., et al. 2006. Diversity and distribution of onstrated for both grassland plant ecosystems Victoria Land biota. Soil Biology & Biochemistry (Piper 1995) and experimental aquatic meso- 38:3003–3018. cosms (Chase 2010). Considering these examples, Aguiar, M. R., and O. E. Sala. 1999. Patch structure, the minimal response in community structure to dynamics and implications for the functioning of resource availability may be interpreted as arid ecosystems. Trends in Ecology & Evolution 14:273–277. taxonomic nestedness along a productivity gra- Aislabie, J., S. Jordan, J. Ayton, J. L. Klassen, G. M. dient, where oligotrophic communities represent Barker, and S. Turner. 2009. Bacterial diversity a diminished, tolerant subset of those taxa associated with ornithogenic soil of the Ross Sea normally present under more eutrophic condi- region, Antarctica. Canadian Journal of Microbiol- tions. Similarly, the lack of response in commu- ogy 55:21–36. nity diversity along an environmental severity Baas Becking, L. G. M. 1934. Geobiologie of inleiding gradient may be interpreted as turnover of tot de milieukunde. W.P. van Stockum and Zoon, community members better adapted to geochem- The Hague, The Netherlands. Barrett, J. E., R. A. Virginia, D. H. Wall, A. N. Parsons, ical extremes (saline/alkaline habitats) without L. E. Powers, and M. B. Burkins. 2004. Variation in modification of alpha richness. Still unknown are biogeochemistry and soil biodiversity across spatial the temporal responses of dry valley bacterial scales in a polar desert ecosystem. Ecology communities to changing abiotic conditions 85:3105–3118. caused by seasonal and other long-term dynam- Barrett, J. E., R. A. Virginia, D. H. Wall, C. S. Cary, B. J. ics, although it seems probable that alterations to Adams, A. L. Hacker, and J. M. Aislabie. 2006. Co- the physicochemical environment will induce a variation in soil biodiversity and biogeochemistry response in bacterial community structure. Alto- in northern and southern Victoria Land, Antarctica. Antarctic Science 18:535–548. gether this evidence suggests that bacterial Barrett, J. E., M. N. Gooseff, and C. Takacs-Vesbach. communities can respond to abiotic controls in 2009. Spatial variation in soil active-layer geochem- ways similar to macroorganisms, yet the re- istry across hydrologic margins in polar desert sponse is context dependent, contingent on the ecosystems. Hydrology and Earth System Sciences nature of environmental change (e.g., altered 13:2349–2358. resource availability or geochemical severity). Bockheim, J. G. 1997. Properties and classification of The functionality of the soil microbiome may be cold desert soils from Antarctica. Soil Science subsequently dependent on whether bacterial Society of America Journal 61:224–231. Broady, P. A. 1996. Diversity, distribution and dispersal diversity, structure, or both metrics respond to of Antarctic terrestrial algae. Biodiversity and shifting environmental conditions. Conservation 5:1307–1335. Bryant, J. A., C. Lamanna, H. Morlon, A. J. Kerkhoff, ACKNOWLEDGMENTS B. J. Enquist, and J. L. Green. 2008. Microbes on mountainsides: Contrasting elevational patterns of Many individuals played instrumental roles in bacterial and plant diversity. Proceedings of the various stages of this research. In particular, we would National Academy of Sciences USA 105:11505– like to thank the Crary Laboratory staff at McMurdo 11511. Station, as well as Raytheon Company, Inc. and Burkins, M. B., R. A. Virginia, C. P. Chamberlain, and Petroleum Helicopters, Inc. for their logistical support. D. H. Wall. 2000. Origin and distribution of soil We also thank Dr. Charles Lee for advice in project organic matter in Taylor Valley, Antarctica. Ecology development, as well as Bobbie Niederlehner and 81:2377–2391. several Virginia Tech collaborators for their contribu- Cary, C. S., I. R. McDonald, J. E. Barrett, and D. A. tions towards data acquisition and analysis. This work Cowan. 2010. On the rocks: The microbiology of

v www.esajournals.org 13 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Antarctic Dry Valley soils. Nature Reviews Micro- B. J. M. Bohannan. 2003. Bacterial diversity patterns biology 8:129–138. along a gradient of primary productivity. Ecology Castle, S. C., C. D. Morrison, and N. N. Barger. 2011. Letters 6:613–622. Extraction of chlorophyll a from biological soil Kennedy, A. D. 1993. Water as a limiting factor in the crusts: A comparison of solvents for spectrophoto- Antarctic terrestrial environment: A biogeograph- metric determination. Soil Biology & Biochemistry ical synthesis. Arctic and Alpine Research 25:308– 43:853–856. 315. Chase, J. M. 2010. Stochastic community assembly Langenheder, S., and J. I. Prosser. 2008. Resource causes higher biodiversity in more productive availability influences the diversity of a functional environments. Science 328:1388–1391. group of heterotrophic soil bacteria. Environmental Cheng, W., and R. A. Virginia. 1993. Measurement of Microbiology 10:2245–2256. microbial biomass in Arctic tundra soils using Lee, C. K., B. A. Barbier, E. M. Bottos, I. R. McDonald, fumigation-extraction and substrate-induced respi- and C. S. Cary. 2012. The Inter-Valley Soil ration procedures. Soil Biology and Biochemistry Comparative Survey: The ecology of Dry Valley 25:135–141. edaphic microbial communities. International Soci- Dunbar, J., L. O. Ticknor, and C. R. Kuske. 2001. ety for Microbial Ecology Journal 6:1046–1057. Phylogenetic specificity and reproducibility and Logue, J. B., S. Langenheder, A. F. Andersson, S. new method for analysis of terminal restriction Bertilsson, S. Drakare, A. Lanzen, and E. S. fragment profiles of 16S rRNA genes from bacterial Lindstrom. 2012. Freshwater bacterioplankton rich- communities. Applied and Environmental Micro- ness in oligotrophic lakes depends on nutrient biology 67:190–197. availability rather than on species-area relation- Eilers, K. G., C. L. Lauber, R. Knight, and N. Fierer. ships. International Society for Microbial Ecology 2010. Shifts in bacterial community structure Journal 6:1127–1136. associated with inputs of low molecular weight Martiny, J. B. H., et al. 2006. Microbial biogeography: carbon compounds to soil. Soil Biology & Bio- Putting microorganisms on the map. Nature chemistry 42:896–903. Reviews Microbiology 4:102–112. Fierer, N., and R. B. Jackson. 2006. The diversity and McCune, B., and J. B. Grace. 2002. Analysis of biogeography of soil bacterial communities. Pro- ecological communities. MjM Software Design, ceedings of the National Academy of Sciences USA Gleneden Beach, Oregon, USA. 103:626–631. McKnight, D. M., C. M. Tate, E. D. Andrews, D. K. Fierer, N., M. A. Bradford, and R. B. Jackson. 2007. Niyogi, K. Cozzetto, K. Welch, W. B. Lyons, and Toward an ecological classification of soil bacteria. D. G. Capone. 2007. Reactivation of a cryptobiotic Ecology 88:1354–1364. stream ecosystem in the McMurdo Dry Valleys, Finlay, B. J. 2002. Global dispersal of free-living Antarctica: A long-term geomorphological experi- microbial eukaryote species. Science 296:1061– ment. Geomorphology 89:186–204. 1063. Metting, F. B., Jr. 1994. DMSO extraction method for Fountain, A. G., et al. 1999. Physical controls on the quantification of chlorophyll in soil. Pages 443–444 Taylor Valley ecosystem, Antarctica. BioScience in J. M. Bigham, editor. Methods of soil analysis, 49:961–971. Part 2: Microbiological and biochemical properties. Freckman, D. W., and R. A. Virginia. 1993. Extraction Soil Science Society of America, Madison, Wiscon- of nematodes from Dry Valley Antarctic soils. Polar sin, USA. Biology 13:483–487. Michalet, R., R. W. Brooker, L. A. Cavieres, Z. Freckman, D. W., and R. A. Virginia. 1997. Low- Kikvidze, C. J. Lortie, F. I. Pugnaire, A. Valiente- diversity antarctic soil nematode communities: Banuet, and R. M. Callaway. 2006. Do biotic Distribution and response to disturbance. Ecology interactions shape both sides of the humped-back 78:363–369. model of species richness in plant communities? Grayston, S. J., S. Wang, C. D. Campbell, and A. C. Ecology Letters 9:767–773. Edwards. 1998. Selective influence of plant species Mitchell, K. R., and C. D. Takacs-Vesbach. 2008. A on microbial diversity in the rhizosphere. Soil comparison of methods for total community DNA Biology and Biochemistry 30:369–378. preservation and extraction from various thermal Hogg, I. D., C. S. Cary, P. Convey, K. K. Newsham, environments. Journal of Industrial Microbiology A. G. O’Donnell, B. J. Adams, J. Aislabie, F. Frati, & Biotechnology 35:1139–1147. M. I. Stevens, and D. H. Wall. 2006. Biotic Mittelbach, G. G., C. F. Steiner, S. M. Scheiner, K. L. interactions in Antarctic terrestrial ecosystems: Gross, H. L. Reynolds, R. B. Waide, M. R. Willig, Are they a factor? Soil Biology & Biochemistry S. I. Dodson, and L. Gough. 2001. What is the 38:3035–3040. observed relationship between species richness and Horner-Devine, M. C., M. A. Leibold, V. H. Smith, and productivity? Ecology 82:2381–2396.

v www.esajournals.org 14 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

Niederberger, T. D., I. R. McDonald, A. L. Hacker, mesofauna in above- and below-ground habitats: R. M. Soo, J. E. Barrett, D. H. Wall, and C. S. Cary. Taylor Valley, Antarctica. Polar Biology 32:1549– 2008. Microbial community composition in soils of 1558. Northern Victoria Land, Antarctica. Environmental Sinsabaugh, R. L., D. J. Van Horn, J. Shah, and S. Microbiology 10:1713–1724. Findlay. 2010. Ecoenzymatic stoichiometry in rela- Nkem, J. N., R. A. Virginia, J. E. Barrett, D. H. Wall, tion to productivity for freshwater biofilm and and G. Li. 2006. Salt tolerance and survival plankton communities. Microbial Ecology 60:885– thresholds for two species of Antarctic soil nema- 893. todes. Polar Biology 29:643–651. Soininen, J. 2012. Macroecology of unicellular organ- Noguez, A. M., H. T. Arita, A. E. Escalante, L. J. Forney, isms: patterns and processes. Environmental Mi- F. Garcia-Oliva, and V. Souza. 2005. Microbial crobiology Reports 4:10–22. macroecology: Highly structured prokaryotic soil Smith, V. H. 2007. Microbial diversity-productivity assemblages in a tropical deciduous forest. Global relationships in aquatic ecosystems. Federation of Ecology and Biogeography 14:241–248. European Microbiological Societies Microbiology Orwin, K. H., D. A. Wardle, and L. G. Greenfield. 2006. Ecology 62:181–186. Ecological consequences of carbon substrate iden- Smith, J. L., J. E. Barrett, G. Tusnady, L. Rejtoe, and tity and diversity in a laboratory study. Ecology C. S. Cary. 2010. Resolving environmental drivers 87:580–593. of microbial community structure in Antarctic Piper, J. K. 1995. Composition of prairie plant soils. Antarctic Science 22:673–680. communities on productive versus unproductive Stomeo, F., T. P. Makhalanyane, A. Valverde, S. B. sites in wet and dry years. Canadian Journal of Pointing, M. I. Stevens, C. S. Cary, M. I. Tuffin, and Botany 73:1635–1644. D. A. Cowan. 2012. Abiotic factors influence Poage, M. A., J. E. Barrett, R. A. Virginia, and D. H. microbial diversity in permanently cold soil hori- Wall. 2008. The influence of soil geochemistry on zons of a maritime-associated Antarctic Dry Valley. nematode distribution, McMurdo Dry Valleys, Federation of European Microbiological Societies Antarctica. Arctic Antarctic and Alpine Research Microbiology Ecology 82:326–340. 40:119–128. Takacs-Vesbach, C., L. Zeglin, J. E. Barrett, M. N. Pointing, S. B., Y. Chan, D. C. Lacap, M. C. Y. Lau, J. A. Gooseff, and J. C. Priscu. 2010. Factors promoting Jurgens, and R. L. Farrell. 2009. Highly specialized microbial diversity in the McMurdo Dry Valleys, microbial diversity in hyper-arid polar desert. Antarctica. Pages 221–257 in P. Doran, W. B. Lyons, Proceedings of the National Academy of Sciences and D. M. McKnight, editors. Life in Antarctic USA 106:19964–19969. deserts and other cold dry environments: Astro- Prosser, J. I. 2007. The role of ecological theory in biological analogs. Cambridge University Press, microbial ecology. Nature Reviews Microbiology Cambridge, UK. 5:384–392. Thies, J. E. 2007. Soil microbial community analysis R Development Core Team. 2011. R: A language and using terminal restriction fragment length poly- environment for statistical computing. R Founda- morphisms. Soil Science Society of America Journal tion for Statistical Computing, Vienna, Austria. 71:579–591. Ricklefs, R. E., and D. Schluter, editors. 1994. Species Tilman, D. 1982. Resource competition and community diversity in ecological communities. University of structure. Princeton University Press, Princeton, Chicago Press, Chicago, Illinois, USA. New Jersey, USA. Schutte, U. M. E., Z. Abdo, S. J. Bent, C. Shyu, C. J. Treonis, A. M., D. H. Wall, and R. A. Virginia. 1999. Williams, J. D. Pierson, and L. J. Forney. 2008. Invertebrate biodiversity in Antarctic dry valley Advances in the use of terminal restriction frag- soils and sediments. Ecosystems 2:482–492. ment length polymorphism (T-RFLP) analysis of Ugolini, F. C., and J. G. Bockheim. 2008. Antarctic soils 16S rRNA genes to characterize microbial commu- and soil formation in a changing environment: A nities. Applied Microbiology and Biotechnology review. Geoderma 144:1–8. 80:365–380. Vavrek, M. J. 2011. Fossil: Palaeoecological and palaeo- Seppelt, R. D., and T. G. A. Green. 1998. A bryophyte geographical analysis tools. Palaeontologia Elec- flora for Southern Victoria Land, Antarctica. New tronica 14:1T. http://palaeo-electronica.org/2011_1/ Zealand Journal of Botany 36:617–635. 238/index.html Silvertown, J., P. Poulton, E. Johnston, G. Edwards, M. Waide, R. B., M. R. Willig, C. F. Steiner, G. Mittelbach, Heard, and P. M. Biss. 2006. The Park Grass L. Gough, S. I. Dodson, G. P. Juday, and R. Experiment 1856-2006: Its contribution to ecology. Parmenter. 1999. The relationship between produc- Ecology 94:801–814. tivity and species richness. Annual Review of Simmons, B. L., D. H. Wall, B. J. Adams, E. Ayres, J. E. Ecology and Systematics 30:257–300. Barrett, and R. A. Virginia. 2009. Terrestrial Wall, D. H., and R. A. Virginia. 1999. Controls on soil

v www.esajournals.org 15 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

biodiversity: Insights from extreme environments. Gooseff, and C. D. Takacs-Vesbach. 2009. Land- Applied Soil Ecology 13:137–150. scape distribution of microbial activity in the Wardle, D. A., R. D. Bardgett, J. N. Klironomos, H. McMurdo Dry Valleys: Linked biotic processes, Setala, W. H. van der Putten, and D. H. Wall. 2004. hydrology, and geochemistry in a cold desert Ecological linkages between aboveground and ecosystem. Ecosystems 12:562–573. belowground biota. Science 304:1629–1633. Zeglin, L. H., C. N. Dahm, J. E. Barrett, M. N. Gooseff, Yergeau, E., S. Bokhorst, A. H. L. Huiskes, H. T. S. S. K. Fitpatrick, and C. D. Takacs-Vesbach. 2011. Boschker, R. Aerts, and G. A. Kowalchuk. 2007. Size Bacterial community structure along moisture and structure of bacterial, fungal and nematode gradients in the parafluvial sediments of two communities along an Antarctic environmental gradient. Federation of European Microbiological ephemeral desert streams. Microbial Ecology Societies Microbiology Ecology 59:436–451. 61:543–556. Zeglin, L. H., M. Stursova, R. L. Sinsabaugh, and S. L. Zhou, J. Z., B. C. Xia, D. S. Treves, L. Y. Wu, T. L. Collins. 2007. Microbial responses to nitrogen Marsh, R. V. O’Neill, A. V. Palumbo, and J. M. addition in three contrasting grassland ecosystems. Tiedje. 2002. Spatial and resource factors influenc- Oecologia 154:349–359. ing high microbial diversity in soil. Applied and Zeglin, L. H., R. L. Sinsabaugh, J. E. Barrett, M. N. Environmental Microbiology 68:326–334.

v www.esajournals.org 16 October 2013 v Volume 4(10) v Article 127 GEYER ET AL.

SUPPLEMENTAL MATERIAL

APPENDIX

Table A1. Regional sampling site locations and average (n ¼ 3) soil characteristics including a-glucosidase activity (activity/g organic carbon/hr), b-glucosidase activity (activity/g organic carbon/hr), nitrate-N (mg/kg dry soil), and sulfate (mg/kg dry soil) along with average soil characteristics of ad hoc productivity classes; standard deviation in parentheses.

Latitude/ a-glucosidase b-glucosidase 1 2 Site ID or class Landscape location Valley longitude activity activity NO3 -N SO4 1 Onyx River margin Wright 77.442367/ 1857.50 1581.48 3.52 74.78 162.667550 (493.92) (766.04) (5.53) (65.31) 2 Hypolith site, uplands Taylor 77.626400/ 2126.28 2249.36 0.03 31.73 below Aiken Glacier 163.331100 (221.82) (1061.10) (0.03) (8.63) 3 Unnamed Pond Wright 77.565750/ 3327.02 2124.80 12.71 45.79 margin, Labyrinth 160.830800 (1664.88) (1109.35) (2.30) (15.15) 4 Lower Green Creek Taylor 77.622950/ 4842.90 4612.74 0.42 10.47 margin 163.065300 (5049.09) (4617.62) (0.23) (3.00) 5 Unnamed Pond Wright 77.557817/ 1700.04 1827.10 9.46 9.09 margin, Labyrinth 160.950267 (1212.43) (1513.63) (11.85) (2.57) 6 Gupwell Pond margin, Wright 77.557733/ 1873.00 1617.74 2.94 5.86 Labyrinth 160.909900 (504.11) (506.55) (1.61) (4.99) 7 Spaulding Pond Taylor 77.658400/ 2205.58 2856.44 1.03 8.45 margin 163.102517 (132.35) (248.30) (0.25) (8.79) 8 Snowpack margin, Taylor 77.637333/ 3192.96 3503.59 0.68 3.96 near south shore 162.881200 (1080.89) (1172.56) (0.50) (0.27) Lake Hoare 9 Upper Green Creek Taylor 77.624400/ 2744.05 2692.25 0.43 1.21 margin 163.05403 (578.63) (1252.97) (0.21) (0.53) 10 Murray Pond margin, Wright 77.558667/ 3563.30 4375.43 1.01 4.86 Labyrinth 160.940317 (1142.35) (1543.20) (0.56) (2.33) 11 Canada Stream margin Taylor 77.615417/ 4553.24 7420.94 1.38 4.02 163.041450 (1567.90) (3835.68) (1.02) (5.42) 12 Bonney Riegel, near Taylor 77.733333/ 5571.71 5908.31 0.80 8.46 Wormherder Creek 162.320183 (2285.91) (2635.58) (0.08) (3.03) 13 Unnamed pond Wright 77.442417/ 3638.05 4245.76 1.40 462.08 margin, near L. 162.733850 (1945.01) (1984.87) (1.26) (367.13) Brownworth 14 Unnamed pond Wright 77.442433/ 4971.52 7302.43 1.63 108.07 margin, near L. 162.758483 (2017.54) (3615.60) (0.46) (54.59) Brownworth 15 Bonney Riegel, near Taylor 77.730383/ 6294.70 6697.27 1.57 10.34 Wormherder Creek 162.334400 (1202.78) (1910.42) (1.19) (10.36) 16 Snowpack margin, Wright 77.442533/ 4966.40 5156.92 1.84 10.82 near L. Brownworth 162.744950 (1336.84) (2267.71) (1.51) (7.33) Ad hoc productivity classes Low (n ¼ 8) 2821.01 2677.99 1.76 95.25 (1533.99) (1828.61) (3.41) (146.90) Medium (n ¼ 15) 3223.98 3516.69 0.80 13.87 (1133.27) (983.74) (0.61) (24.45) High (n ¼ 9) 5247.96 7302.18 1.79 132.20 (1790.74) (2928.05) (0.85) (271.35)

v www.esajournals.org 17 October 2013 v Volume 4(10) v Article 127