Soil Water Availability and Microsite Mediate Fungal and Bacterial

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by DigitalCommons@University of Nebraska

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

U.S. Department of Agriculture: Agricultural Publications from USDA-ARS / UNL Faculty Research Service, Lincoln, Nebraska

2011

Soil water availability and microsite mediate fungal and bacterial phospholipid fatty acid biomarker abundances in Mojave Desert soils exposed to elevated atmospheric CO2

V. L. Jin USDA-ARS, [email protected]

S. M. Schaeffer University of California, Santa Barbara

S. E. Ziegler Memorial University of Newfoundland

R. D. Evans Washington State University

Follow this and additional works at: https://digitalcommons.unl.edu/usdaarsfacpub

Jin, V. L.; Schaeffer, S. M.; Ziegler, S. E.; and Evans, R. D., "Soil water availability and microsite mediate fungal and bacterial phospholipid fatty acid biomarker abundances in Mojave Desert soils exposed to elevated atmospheric CO2" (2011). Publications from USDA-ARS / UNL Faculty. 1622. https://digitalcommons.unl.edu/usdaarsfacpub/1622

This Article is brought to you for free and open access by the U.S. Department of Agriculture: Agricultural Research Service, Lincoln, Nebraska at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Publications from USDA-ARS / UNL Faculty by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116, G02001, doi:10.1029/2010JG001564, 2011

Soil water availability and microsite mediate fungal and bacterial phospholipid fatty acid biomarker abundances in Mojave Desert soils exposed to elevated atmospheric CO2 V. L. Jin,1 S. M. Schaeffer,2 S. E. Ziegler,3 and R. D. Evans4 Received 29 September 2010; revised 9 January 2011; accepted 26 January 2011; published 14 April 2011.

[1] Changes in the rates of nitrogen (N) cycling, microbial carbon (C) substrate use, and extracellular enzyme activities in a Mojave Desert ecosystem exposed to elevated atmospheric CO2 suggest shifts in the size and/or functional characteristics of microbial assemblages in two dominant soil microsites: plant interspaces and under the dominant shrub Larrea tridentata. We used ester‐linked phospholipid fatty acid (PLFA) biomarkers as a proxy for microbial biomass to quantify spatial and temporal differences in soil microbial communities from February 2003 to May 2005. Further, we used the 13C signature of the fossil CO2 source for elevated CO2 plots to trace recent plant C inputs into soil organic matter (SOM) and broad microbial groups using d13C(‰). Differences 13 13 between individual d CPLFA and d CSOM for fungal biomarkers indicated active metabolism of newer C in elevated CO2 soils. Total PLFA‐C was greater in shrub microsites compared to plant interspaces, and CO2 treatment differences within microsites increased under higher soil water availability. Total, fungal, and bacterial PLFA‐C increased with decreasing soil volumetric water content (VWC) in both microsites, suggesting general adaptations to xeric desert conditions. Increases in fungal‐to‐bacterial PLFA‐C ratio with decreasing VWC reflected functional group‐specific responses to changing soil water availability. While temporal and spatial extremes in resource availability in desert ecosystems contribute to the difficulty in identifying common trends or mechanisms driving microbial responses in less extreme environments, we found that soil water availability and soil microsite interacted with elevated CO2 to shift fungal and bacterial biomarker abundances in Mojave Desert soils. Citation: Jin, V. L., S. M. Schaeffer, S. E. Ziegler, and R. D. Evans (2011), Soil water availability and microsite mediate fungal and bacterial phospholipid fatty acid biomarker abundances in Mojave Desert soils exposed to elevated atmospheric CO2, J. Geophys. Res., 116, G02001, doi:10.1029/2010JG001564.

1. Introduction resources (e.g., quantity and quality), specific microbial substrate demands related to stoichiometric constraints (i.e., [2] The responses of soil microbial assemblages to bacterial versus fungal C:N ratio) [Ziegler and Billings, changing atmospheric CO concentrations will depend pri- 2 2011], and the physiological differences between active marily on the indirect effects of elevated CO on plant pro- 2 soil microbial functional groups [Billings and Ziegler, 2005; cesses which provide the key energy resources for microbes Strickland and Rousk, 2010]. Microbial dynamics are also (e.g., root exudates, rhizodeposition, litter) [Kandeler et al., involved in nutrient feedback mechanisms that affect plant 1998; Hungate et al., 2000; Pendall et al., 2004; Drigo productivity. Soil microbial activity can be C‐limited [Zak et al., 2008]. These microbial community‐level responses et al., 1996; Schaeffer et al., 2003], so shifts in microbial will be mediated by the interactions between available soil C community composition driven by plant C inputs could impact microbially mediated N cycling processes that can feed back 1Agroecosystem Management Research Unit, Agricultural Research to plant growth [Smith et al., 2000; Zak et al., 2000; Jin and Service, U.S. Department of Agriculture, University of Nebraska at Evans, 2007]. Some studies have found no compositional Lincoln, Lincoln, Nebraska, USA. 2Department of Ecology, Evolution, and Marine Biology, University of changes in microbial assemblages associated with altered California, Santa Barbara, California, USA. microbial activities under elevated CO2 [Zak et al., 2000; 3Department of Earth Sciences, Memorial University of Newfoundland, Rønn et al., 2002; Ebersberger et al., 2004], while others St. John’s, Newfoundland and Labrador, Canada. 4 show no measureable changes in soil microbial activity School of Biological Sciences and Laboratory for Biotechnology and although compositional changes under elevated CO are evi- Bioanalysis 2 Stable Isotope Core, Washington State University, Pullman, 2 Washington, USA. dent [He et al., 2010]. Several studies, however, have linked CO2‐induced changes in microbial processes with changes in Copyright 2011 by the American Geophysical Union. soil microbial community composition [Montealegre et al., 0148‐0227/11/2010JG001564

G02001 1of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

2002; Billings and Ziegler, 2005; Janus et al., 2005]. In shifts in the functional composition or structure of soil addition, changes in microbial function can still occur in the microbial assemblages. absence of compositional shifts as measured by PLFAs or [6] The objective of this study, therefore, is to determine polymerase chain reaction–denaturing gradient gel electro- whether previously reported changes in microbial process phoresis (PCR–DGGE) [Chung et al., 2006]. rates are concomitant with measurable changes in fungal [3] The climate in extreme desert environments, coupled and/or bacterial abundances and C utilization relationships with both temporal and spatial extremes in resource avail- in soil microbial communities in Mojave Desert soils at the ability, contribute to the difficulty in identifying common NDFF. We assessed changes in microbial functional struc- trends or mechanisms that drive microbial responses in less ture and functional group abundances by using ester‐linked extreme environments. Field studies on soil microbial phospholipid fatty acid biomarkers (PLFAs) to identify responses to elevated CO2 are dominated by experiments broad classes of soil microorganisms [Ringelberg et al., 1989; conducted in more mesic forest, agricultural, or grassland Tunlid and White, 1992; Zelles, 1999]. We examined soil ecosystems, with a few studies from more xeric scrub or PLFA biomarkers from the NDFF from February 2003 to Mediterranean climates. Soil microbial responses to elevated May 2005 in soils from the two dominant microsites, plant CO2 in extreme arid ecosystems, however, is limited to interspaces and under Larrea tridentata. Specific shifts in a single experimental site in the Mojave Desert, Nevada microbial community diversity or species composition can- Desert FACE (Free‐Air Carbon Dioxide Enrichment) Facility not be resolved using PLFAs, but biochemical profiling via (NDFF; Nevada, USA) [Jordan et al., 1999]. Deserts com- PLFA biomarkers can be used to evaluate spatial and tem- prise almost one‐third of the terrestrial biosphere [Nobel poral patterns in general microbial community characteristics, et al., 1996], and plant and soil processes in arid ecosystems such as the abundances of different microbial functional are likely to be among the most responsive to increasing groups. Further, changes in microbial substrate utilization can concentrations of atmospheric CO2 [Melillo et al., 1993; also be assessed using the stable C isotopic composition of Hungate et al.,1997;Owensby et al., 1999]. Shifts in micro- different PLFA biomarkers. bial community compositions and functions in arid ecosys- [7] The use of an isotopically distinct CO2 source for tem responses to rising atmospheric CO2, therefore, could implementing the elevated CO2 treatment provides a pow- have significant impacts on global biogeochemical cycling erful tool for assessing changes in natural assemblages of of C and N. soil microorganisms via the stable carbon isotope ratios [4] In semiarid and arid environments, temporal and of individual PLFAs [Abraham et al., 1998; Waldrop and spatial heterogeneity in resource availability as well as in the Firestone, 2004]. At the NDFF, a change in CO2 source distribution of soil microbial communities themselves in February 2003 resulted in a −11.4 ± 2.3‰ difference in 13 interact to constrain biological activity. Pulses in water and the d CofCO2 in the air above elevated treatment plots nutrient availability following rainfall events trigger pulses relative to ambient plots [Schaeffer, 2005]. This change in 13 in biological activity [Fierer and Schimel, 2002; Austin the d C of atmospheric CO2 was used to follow the transfer et al., 2004; Xiang et al., 2008]. Aridland soil resources of recently fixed C from plants into soil organic matter and also vary spatially within a resource‐poor matrix of plant microbial PLFAs during the 2 year study period. Previously interspace areas punctuated by “islands of fertility” under measured changes in C use and extracellular enzyme activi- perennial vegetation [Schlesinger et al., 1996; Titus et al., ties, particularly those involving organic N and P cycling, 2002]. Biological activity in these arid environments is suggested increases in soil fungal activity and/or fungal thus pulse‐adapted in both time and space, with water abundance in soils exposed to elevated CO2. In this study, availability acting as the primary control on activity. Thus, therefore, we hypothesized: (1) an increase in the relative flows of nutrient and energy in arid ecosystems under fungal component of the soil microbial community in soils increasing atmospheric CO2 will also depend on how soil exposed to elevated CO2 relative to ambient soils; (2) that microbial communities respond to variability in current and changes in microbial community structure over time (i.e., future rainfall patterns [Mikha et al., 2005; Schimel et al., fungal‐to‐bacterial PLFA‐C ratio) and concomitant shifts in 13 2007; Gordon et al., 2008]. the use of recently fixed C would be reflected in d CPLFA 13 13 [5] In the Mojave Desert, elevated CO2 has enhanced relative to the d C of soil organic matter (d CSOM) in plots aboveground plant C inputs into the soil [Smith et al., 2000; exposed to elevated CO2; and (3) that soil moisture avail- Nowak et al., 2004; Housman et al., 2006] and increased ability will be positively correlated with soil microbial rates of microbial C and N transformations [Billings et al., biomass (i.e., total PLFA‐C), particularly in interspace soils 2003, 2004; Schaeffer et al., 2003; Jin and Evans, 2007]. exposed to elevated CO2 given previously reported increases In addition, the amount and diversity of C substrates used by in microbial activities under high water availability [Jin and soil microbes as well as extracellular enzyme activities Evans, 2007]. involved in microbial C, N, and P cycling (cellobiohydrolase, leucine aminopeptidase, alkaline phosphatase) has increased 2. Materials and Methods under elevated CO2 in soils under the dominant shrub Larrea tridentata [Jin and Evans, 2007]. In that study, it 2.1. Study Site was hypothesized that general increases in microbial activi- [8] The study was conducted at the Nevada Desert Free‐ ties under elevated CO2 were attributable to greater micro- Air Carbon Dioxide Enrichment Facility (NDFF), 15 km bial biomass in interspace soils, and to increased microbial north of Mercury, NV, USA (36°49′N, 115°55′W; elevation turnover rates and/or activity levels rather than microbial 965–970 m) [Jordan et al., 1999]. The community is domi- pool size in shrub microsite soils. These observed changes nated by shrubs (Larrea tridentata (DC.) Cov., Ambrosia in microbial C and N cycling further suggest potential dumosa (A. Gray) Payne, Lycium andersonii (A. Gray),

2of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

Lycium pallidum (Miers var. oligospermum C. Hitchc.)), from either microsite was not removed. Prior to freezing, and perennial grasses (Pleuraphis rigida Thurber, 20 g subsamples were weighed, oven‐dried at 50°C for 72 h, Achnatherum hymenoides (Roemer and Schultes) Bark- and reweighed for soil water content. Soil water content worth). Crytobiotic soil crusts can dominate plant inter- was then converted to volumetric water content (%) using spaces [Jordan et al., 1999], and contain both autotrophic measured soil bulk densities. Frozen soil samples were and heterotrophic organisms including lichens, bryophytes, lyophilized and stored at −70°C in the dark until extraction and cyanobacteria. The long‐term mean annual precipita- for phospholipids. Values are presented as treatment means tion (MAP) is 140 mm [Hunter, 1994]. MAP during this per g dry soil ± one standard error (se). study was 149 mm, 123 mm, and 242 mm in 2003, 2004, d13 and 2005, respectively, with significant within‐year varia- 2.3. Soil Organic Matter C Determination Larrea tridentata d13 tion [Charlet, 2007]. Soils are Typic Haplocalcids derived and Foliar C from calcareous alluvium, with textures ranging from loamy [12] Soil subsamples were analyzed for the C isotopic 13 sands in surface horizons to coarse sands in the subsoil composition of soil organic matter (d CSOM). Prior to iso- horizons and no caliche layer in subsoils [Jordan et al., topic analysis of organic C, inorganic C (i.e., carbonates) 13 1999]. Soil pH is alkaline, and ranges from 8 to 9 at all was removed to determine carbonate‐free d CSOM. Acid depths [Jordan et al., 1999]. was added to soils (3 N H3PO4; 1 g: 3 mL soil:acid ratio) [9] The site was established in April 1997, consisting of and allowed to sit overnight, rinsed three times with deio- nine 23 m diameter circular plots exposed to three CO2 nized (DI) water to remove residual acid, then dried at 60°C treatments using free‐air carbon dioxide enrichment (FACE) and reground. Carbonate content based on total mass loss technology [Hendrey and Kimball, 1994]. Three plots during acid‐washing ranged from 76% to 84%. Testing of each were maintained at either ambient atmospheric CO2 this method on native carbonate‐free soils indicated that −1 13 (380 mmol CO2 mol ) or at 1.5 X ambient CO2 (550 mmol d CSOM was not significantly different between acid‐ −1 CO2 mol ). Treatment concentrations were applied to plots washed soils compared to soils washed with deionized water by pushing in either ambient air or ambient air mixed with or unwashed soils [Schaeffer, 2005]. Soil samples were pure CO2 (added by a mass flow controller) through vent analyzed on an elemental analyzer coupled to a Thermo‐ pipes [Jordan et al., 1999]. There were also three nonblower Finnigan Delta+XP via a Finnigan GC/CIII combustion control plots (i.e., no FACE ring structure). Treatments were interface at the Washington State Laboratory for Biotech- maintained continuously except when air temperatures fall nology and Bioanalysis 2 Stable Isotope Core. −1 13 below 3°C or at wind speeds > 7 m s , resulting in oper- [13] Larrea tridentata foliar d C was measured in re- ating 98.4% to 98.1% of the time from 2003 to 2005. Site plicates (n = 3 to 6) of foliage sampled from each of the six operational statistics are available online (http://www.unlv. CO2 treatment plots immediately prior to CO2 source edu/Climate_Change_Research/NDFF/co2_treatment.htm). change and then throughout March to June 2003 to 2005. 13 [10] The d CofCO2 in the air in ambient plots varies All plant samples were oven‐dried for 48 h at 60°C, ground between −8.2 ± 0.3 and −8.5 ± 0.4‰. The d13C of source to a fine powder, and analyzed with an elemental analyzer CO2 measured prior to atmospheric mixing was approxi- coupled to an isotope‐ratio mass spectrometer (Thermo‐ mately −6‰ until February 2003, after which time the CO2 Finnigan MAT, Bremen, Germany) at the University of supply was changed due to a shift in commercial supplier. Arkansas Stable Isotope Laboratory (Fayetteville, AK, USA), The new source d13C was approximately −30‰. This shift the University of Georgia’s Stable Isotope/Soil Biology in the source CO2 value resulted in a consistent −8to−11‰ Laboratory (Athens, GA, USA), or Washington State Uni- difference in the atmospherically mixed CO2 in the air versity’s Stable Isotope Core (Pullman, WA, USA). Plant above elevated plots compared to ambient plots [Schaeffer, standards were analyzed in each laboratory to ensure ana- 2005]. The CO2 source change was reflected in the foliar lytical consistency across isotope laboratories. isotopic composition of Larrea tridentata (Figure 1). 2.4. Phospholipid Fatty Acid Analysis 2.2. Soil Sampling [14] Approximately 50 g of lyophilized soil was extracted [11] Soil samples were collected from a suspended sled using a modified Bligh‐Dyer procedure to provide adequate attached to a rotating walkway to limit soil disturbance and microbial biomass for determining the C content, relative prevent destruction of the cryptobiotic soil crust [Jordan abundance, and isotopic composition of PLFA biomarkers et al., 1999]. Soils from the top 5 cm of the soil horizon [Bligh and Dyer, 1959; White and Ringelberg, 1998; Pinkart were sampled in the six treatment plots using a 5 cm et al., 2002]. A silicic acid phase column on a vacuum diameter PVC soil sampling device at seven sampling dates: manifold was used to separate the neutral lipids, free fatty February, April, July, and September 2003; October 2004; acids, and phospholipids from one another in each sample and February and May 2005. Soils were sampled from three after lipid extraction [Dobbs and Findlay, 1993; White and separate, randomly selected locations within each plot from Ringelberg, 1998]. Following mild alkali hydrolysis with each of two microsite types: plant interspaces and under KOH and acidification with HCl, fatty acids in the phos- the dominant shrub Larrea tridentata. Cryptobiotic crust, if pholipid fraction were converted to their corresponding fatty present, was gently brushed off before soil collection. acid methyl esters (FAME) using BF3 in methanol [Dobbs Samples were composited by microsite for a total sample size and Findlay, 1993]. of ∼100 g fresh weight. Soils were thoroughly homogenized, [15] Processed FAME samples were analyzed at the passed through a 2 mm sieve, and frozen within 3 h of University of Arkansas in 2003 and at Washington State sample collection. Roots that passed through the sieve were University in 2004–2005. All lipid extraction methods were removed by hand, but particulate organic matter in soils identical at both locations. Fatty acid methyl esters were

3of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

Figure 1. Isotopic compositions (d13C; ‰) of bulk soil organic matter, Larrea tridentata foliage (prior to and after CO2 source change), and individual PLFA biomarkers from different microbial functional groups. Data shown are untransformed mean values ± standard error (n = 3). Asterisks and daggers indi- 13 cate significant CO2 treatment effects on d CPLFA and D values, respectively (P ≤ 0.05). For each 13 microsite, mean d CSOM by CO2 treatment are shown (dotted line, ambient; solid line, elevated) because values did not differ between sampling dates (P ≥ 0.05).

4of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001 identified using a gas chromatograph (Hewlett Packard Post‐hoc multiple comparisons between fixed treatment HP 5890 series II plus, Palo Alto, CA, USA) with a 70% means were tested for differences using the Bonferroni‐ cyanopropyl polysilphenylene‐siloxane column (SGE BPX‐ adjusted Fisher’s least significant difference (LSD) proce- 70; 60 m × 0.25 mm × 0.25 mm), and interfaced with a mass dure. Data were tested for normality using the Shapiro‐Wilk selective detector (HP 5970B). Unknown FAMEs were statistic and transformed when necessary. Linear regressions identified by comparing retention times and mass spectra were used to evaluate changes in total and functional group against a combination of known standards (Bacterial Fatty PLFA‐C, fungal‐to‐bacterial PLFA‐C ratio, or functional Acid Methyl Esters and 37 component FAME standards, group relative abundance against soil volumetric water Supelco Co., Rockford, IL, USA) and mass spectral libraries content. Statistical tests were developed in consultation with (National Bureau of Standards). The concentrations of indi- the Statistics Department at Washington State University, vidual FAMEs in 2003 were determined by GC‐flame ioni- and all tests were performed using SAS 9.1 statistical soft- zation detection (GC‐FID) using the same GC parameters as ware (SAS, Inc., Cary, NC, USA). for GC‐MS analysis. The isotopic compositions of individual fatty acids were determined using an HP6890 GC 3. Results coupled to a Finnigan Delta+ via a Finnigan GC/CIII com- 13 bustion interface located at the University of Arkansas Stable 3.1. Soil Organic Matter d C and L. tridentata 13 Isotope Laboratory. Concentrations and isotopic composi- Foliar d C 13 13 tions of individual fatty acids from 2004 and 2005 were [18] Soil organic matter d C(d CSOM) did not change determined using a Trace‐GC coupled to a Thermo‐Finnigan over time (Date, P = 0.090), but was significantly affected + Delta XP via a Finnigan GC/CIII combustion interface at by microsite and CO2 treatments with no treatment inter- 13 the Washington State Laboratory for Biotechnology and action effects. Specifically, d CSOM was lower (e.g., more Bioanalysis 2 Stable Isotope Core. Weight percent recovery negative) under elevated CO2 compared to ambient condi- was determined from a known amount of phospholipid tions (CO2, P = 0.044) and was also more negative in soils standard (1,2‐diheptadecanoyl‐sn‐glycero‐3‐phosphocholine; under L. tridentata compared to interspace soils (Microsite, Avanti Polar Lipids) added to subsamples of soil which were P = 0.009) (Figure 1). Differences between treatments extracted, methylated, and analyzed along with unamended for the main effects of each microsite and CO2 factors, samples. however, were less than 0.5‰. The isotopic composition of [16] A correction for the methyl carbon addition from L. tridentata foliage decreased over time under elevated BF3/methanol derivatization was calculated for each fatty CO2 compared to ambient conditions (Microsite*CO2, P < acid by mass balance from the analysis of free and meth- 0.0001), indicating that the photosynthetic‐fixation of the ylated internal standards of tridecanoic and tricosanoic acids fossil CO2 used in elevated plots changed foliar isotopic m −1 [Abrajano et al., 1994]. PLFA biomarker C ( gCg ) composition within 4 months after the CO2 source change was used as a proxy for examining microbial biomass C (Figure 1). dynamics. Functional groups were assigned to PLFAs fol- d13 lowing Ringelberg et al. [1989], Zelles et al. [1992], and 3.2. PLFA Biomarker C and Changes Relative d13 Zelles [1999]: general (14:0, 15:0, 16:0, 17:0, 18:0, 18:1w9c, to CSOM 18:1w9t, 20:0, 21:0, 22:0, 24:0); Gram‐negative bacteria [19] Twenty‐nine individual PLFA biomarkers were (16:1w7c, 16:1w5c, 16:1w9, cy17:0, 17:1w7c, cy19:0); Gram‐ identified in Mojave Desert soils across all microsites and positive bacteria (i15:0, a15:0, i16:0, i17:0, a17:0); actino- CO2 treatments over the 2 year study period, with 11 gen- bacteria (3me16:0, 8me17:0, 10me17:0, 10me18:0); and fungi eral, bacterial, and fungal biomarkers common to all soils at (18:2w6c,t, 18:3w3, 20:1w9). all times plus four actinobacterial biomarkers occasionally present in the different treatments. Stable isotopic compo- 2.5. Statistical Analyses 13 sitions (d CPLFA) for all biomarkers varied by date, with [17] A repeated measures mixed effects analyses of vari- variable effects of microsite and CO2 treatment for dif- ance (ANOVA) was used to assess the fixed effects of ferent biomarkers (Date*Microsite*PLFA, Date*CO2*PLFA, 13 sampling date and CO2 on L. tridentata foliar d C(‰). P < 0.0001; Microsite*CO2*PLFA, P = 0.0148). For the 11 13 Split‐plot, repeated measures mixed effects ANOVAs were common biomarkers, d CPLFA in shrub microsites were used to assess the fixed effects of sampling date, CO2, and more negative than in plant interspace soils for general microsite on soil organic matter d13C(‰), soil total PLFA‐C biomarkers 14:0, 16:0, 18:0; Gram negative bacterial bio- abundance (mgCg−1), functional group PLFA‐C, PLFA marker cy17:0; and Gram positive bacterial biomarkers 13 13 relative abundance (%), and PLFA d C(‰) (PROC i15:0, i16:0 (Microsite, P ≤ 0.05). Biomarker d CPLFA MIXED; SAS, Inc, Cary, NC). Microsite was identified as the was also more negative in soils treated with elevated CO2 subplot factor, plot as the random factor, and the covariance compared to ambient soils for general biomarkers 14:0, structure of all analyses accounted for unevenly spaced 16:0, 18:0; Gram negative bacterial biomarker 16:1w9; and 13 sampling dates [Littell et al., 1996]. Changes in d CPLFA in Gram positive bacterial biomarkers a15:0, i16:0 (CO2, P ≤ response to CO2 treatments were assessed by calculating the 0.05). No treatment interactions between date, microsite, or 13 difference between individual PLFA d C and soil organic CO2 treatment were significant for these PLFAs. In contrast, 13 matter d C(D; ‰) for each ambient and elevated CO2 significant three‐way treatment interactions occurred in treatments [Billings and Ziegler, 2005]. Because SOM did general biomarker 18:1w9t and fungal biomarkers 18:2w6c,t not vary by sampling date (see results below), D values for and 18:3w3 (Date*Microsite*CO2, P ≤ 0.05). For general each PLFA were calculated using the mean SOM value biomarker 18:1w9t, significant CO2 treatment differences averaged over all dates by microsite for each CO2 treatment. were more prevalent in soils under L. tridentata than in

5of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

13 13 Table 1. Mean D (d CSOM − d CPLFA; ± se) After CO2 Source Change and After 2 Years of Treatment Under Ambient or Elevated a CO2 in Interspace Soils and Soils Under Larrea tridentata Interspace Soils Soils Under Larrea tridentata February 2003 May 2005 February 2003 May 2005 Fatty Acid Ambient Elevated Ambient Elevated Ambient Elevated Ambient Elevated *14:0 −2.0 ± 0.8 −0.6 ± 2.3 −5.1 ± 0.6 −6.5 ± 0.6 −2.5 ± 1.4 −4.6 ± 0.6 −6.8 ± 0.6 −8.4 ± 0.2 16:0 −2.3 ± 1.3 −3.3 ± 0.4 −4.3 ± 0.6 −8.0 ± 0.7 −2.1 ± 1.7 −2.2 ± 0.5 −5.9 ± 0.7 −7.0 ± 0.2 *18:0 −3.9 ± 1.0 −3.9 ± 1.6 −2.1 ± 0.7 −3.0 ± 0.5 −5.8 ± 0.8 −5.3 ± 0.8 −4.5 ± 0.6 −4.5 ± 0.2 *18:1w9t 1.4 ± 0.1 0.8 ± 0.3 −1.4 ± 1.6 −1.7 ± 1.3 1.4 ± 1.2 −1.7 ± 0.7 −3.9 ± 1.4 −2.9 ± 1.1 16:1w9 −0.7 ± 0.3 −4.9 ± 3.0 3.2 ± 1.1 1.5 ± 0.4 −2.5 ± 0.5 −1.3 ± 1.0 −3.5 ± 1.1 1.4 ± 1.1 *cy17:0 −5.6 ± 1.6 −3.4 ± 1.2 −4.3 ± 0.7 −5.3 ± 1.0 −5.1 ± 1.6 −3.6 ± 0.8 −2.8 ± 1.0 −3.9 ± 0.4 *a15:0 2.7 ± 0.6 0.7 ± 0.3 0.3 ± 0.2 −1.6 ± 0.5 0.3 ± 0.3 0.6 ± 0.6 0.5 ± 0.8 −0.3 ± 0.5 *i15:0 1.1 ± 0.3 2.0 ± 1.8 −0.3 ± 0.6 −0.9 ± 0.1 −0.4 ± 0.4 −0.1 ± 0.8 −0.2 ± 0.6 −1.1 ± 0.5 *i16:0 −0.6 ± 1.3 −3.8 ± 0.9 −5.0 ± 0.6 −5.9 ± 0.9 −2.2 ± 0.3 −2.5 ± 0.7 −3.3 ± 0.9 −4.4 ± 0.2 *18:2w6c,t −7.2 ± 0.9 −1.3 ± 0.3 −4.8 ± 0.4 −9.8 ± 0.8 −8.9 ± 1.0 −5.2 ± 0.8 −7.7 ± 1.5 −6.3 ± 0.1 *18:3w3 −10.7 ± 0.6 −0.4 ± 1.0 −8.8 ± 0.2 −12.3 ± 1.6 −8.2 ± 0.8 −7.5 ± 0.9 −9.0 ± 1.0 −12.6 ± 1.6

a Bolded values indicate significant CO2 treatment differences at each sampling date, and asterisks indicate significant microsite differences (P ≤ 0.05).

interspace soils, whereas significant CO2 treatment effects increased with decreasing soil VWC, and responses were for both fungal biomarkers occurred more in interspace soils microsite‐specific (Figures 3a–3d). In interspace soils, the 13 (Figure 1). For all three biomarkers, d CPLFA tended to magnitude of responses to soil VWC did not differ between become more negative under elevated CO2 later during the CO2 treatments for either fungal PLFA‐C(y = −0.07x + 13 2 study period. The d CPLFA of actinobacterial biomarkers 1.18; r = 0.34, P < 0.001) or bacterial PLFA‐C(y = −0.07x + were not affected by any treatment factors. 1.41; r2 = 0.19, P = 0.004). Changes in functional group PLFA‐C in shrub soils under ambient CO2 (fungal: y = d13 d13 D 2 3.3. Changes in PLFA C Relative to CSOM ( ) −0.23x + 3.10; r = 0.69, P < 0.001; bacterial: y = −0.22x + 13 13 2 [20] The differences between d CSOM and d CPLFA (D) 4.07; r = 0.45, P = 0.001), however, were more sensitive to in ambient and elevated CO2 soils, respectively, were cal- changes in soil VWC compared to under elevated CO2 13 − 2 culated for each individual PLFA (Table 1). Bulk d CSOM (fungal: y = 0.13x + 2.92; r = 0.23, P = 0.032; bacterial: did not vary by sampling date, so D values were calculated y = −0.10x + 3.50; r2 = 0.23, P = 0.033). Fungal‐to‐bacterial 13 ‐ using the mean d CSOM averaged over all dates by micro- PLFA C ratios varied by date (P < 0.0001) and microsite ≤ ‐ ‐ ‐ site for each CO2 treatment. Values for D varied by bio- (interspace > shrub, P 0.0573). Fungal to bacterial PLFA C marker and date, with variable effects of microsite and CO2 ratios also increased with decreasing soil VWC in inter- treatments (Date*Microsite*PLFA, Date*CO2*PLFA, P < space soils, with no differences between CO2 treatment (y = − 2 0.0001; Microsite*CO2*PLFA, P = 0.0147). Significant 0.03x + 0.91, r = 0.13, P = 0.021; Figure 3e). Increases in main effects of microsite on D remained the same as fungal‐to‐bacterial PLFA‐C ratios with decreasing soil VWC 13 d CPLFA for all biomarkers as reported previously. In con- was significant in shrub soils under ambient CO2 only (y = 13 − 2 trast, all significant main effects of CO2 on d CPLFA were not 0.05x + 0.85, r = 0.51, P < 0.001; Figure 3f). found for D values, except for the general biomarker 16:0. [23] The relative abundances of all functional groups varied The three‐way interaction effects and response patterns over time (P < 0.0001) particularly between microsites, with 13 observed for d CPLFA for general biomarker 18:1w9t and few differences related to CO2 treatment (Figures 4a and 4b). fungal biomarkers 18:2w6c,t and 18:3w3 were also signifi- The relative abundance of general biomarkers was greater in cant for respective D values (Date*Microsite*CO2, P ≤ 0.05) shrub microsites compared to interspaces in April, Sep- (Figure 1). tember 2003, and February 2005 (Date*Microsite, P = 0.0476). In contrast, the relative abundances of actinobacteria 3.4. Soil PLFA Biomarker Abundance and Relative and fungi were greater in interspaces (Microsite, P < 0.01). Distribution Functional group relative abundances did not show consistent [21] Soil volumetric water content (VWC, %) did not main or interaction effects of CO2 treatment, although general differ between microsites or CO2 treatments, but varied and actinobacterial biomarkers showed a marginal response significantly with date (P ≤ 0.0001). Soil VWC ranged from in interspace soils (Microsite*CO2; P = 0.0911 and P = 0.6 ± 0.2% to 13.8 ± 2.4% over the study period (Figure 2a). 0.0778, respectively). Total PLFA‐C(mgCg−1 dry soil) contents were greater in soils under L. tridentata compared to interspace soils and varied significantly over time (Date*Microsite, P = 0.0038) 4. Discussion (Figure 2b). Total PLFA‐C concentrations were higher under [24] Changes in the isotopic compositions and relative elevated CO2 compared to ambient conditions in July 2003 abundances of PLFA biomarkers indicate that microbial interspace soils and in 2004–2005 soils under L. tridentata community structure in two dominant soil microsites has (P < 0.05). In September 2003, however, total PLFA‐C was shifted in the Mojave Desert over 8 years of exposure to higher under ambient conditions in shrub microsites only. elevated atmospheric CO2. We found that fungal biomass [22] Total PLFA‐C increased with decreasing soil VWC estimated by PLFA‐C tended to be greater under elevated in both interspace soils and soils under L. tridentata. Spe- CO2, supporting our first hypothesis and consistent with a cifically, both fungal and bacterial biomarker C abundance controlled growth chamber study using soils collected from

6of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

Figure 2. (a) Soil water content (SWC; g water g−1 dry soil) and (b) total PLFA‐C(mgg−1 dry soil) in microsite soils from ambient or elevated CO2 over time. Data shown are untransformed mean values ± standard error (n = 3). Different letters indicate temporal differences in SWC (P ≤ 0.05), and asterisks indicate CO2 differences for each microsite type (P ≤ 0.05). this site [Jin and Evans, 2010]. Increases in fungal PLFA‐C and between microsites under elevated CO2 only was eval- 13 under elevated CO2 were also associated with the use of uated by calculating the difference between d CSOM and 13 recently fixed C as indicated by changes in the differences d CPLFA (D) in ambient and elevated CO2 soils for each 13 13 between d CSOM and d CPLFA for two fungal biomarkers, individual PLFA. This calculation accounted for potential 13 13 supporting our second hypothesis of variable use of recent C shifts in d CSOM derived from the C label of supplemental by different microbial functional groups. In contrast, nega- CO2 used in elevated plots, assuming the even incorporation tive correlations between total PLFA‐C and soil water of the 13C label into metabolizable SOM [Billings and Ziegler, availability (i.e., volumetric water content, VWC), particu- 2005]. Thus, D values for a PLFA would remain the same larly in interspace soils, did not support our third hypothesis between the ambient and treated soils if elevated CO2 had no that increased activities in this microsite may be due to effect on the active microbial community composition. 13 increased microbial biomass. Fungal‐to‐bacterial PLFA‐C [26] While significant decreases in d CPLFA were observed ratios generally increased with decreasing soil VWC in both in various biomarkers in elevated CO2 plots, D values indi- microsites, and responses appeared less sensitive to changes cated that consistent incorporation of newly fixed C occurred in soil water availability in shrub microsite soils exposed to only into two general biomarkers (16:0, 18:1w9t) and two elevated CO2. The general negative relationship between fungal biomarkers (18:2w6c,t; 18:3w3). Other biomarkers fungal‐to‐bacterial PLFA‐C and soil VWC also suggested a that showed significant CO2 effects on D included the gen- greater level of drought tolerance in fungi compared to eral biomarker 14:0, the Gram‐negative biomarker 16:1w9, bacteria, although high levels of variability indicated that and two Gram‐positive biomarkers i16:0 and a15:0. The other soil physical, chemical, or biological factors (i.e., incorporation of new photosynthate into these biomarkers texture, pH, trophic interactions) are also influencing varied temporally after the CO2 source change. Occasional microbial community structure. increases in D indicated the use of other SOC substrate (e.g., nonlabeled SOC) as measured for Gram‐negative bio- 4.1. Use of Recent C by Soil Microbial Communities marker 16:1w9 and fungal biomarker 18:2w6c,t, but gen- 13 [25] The lack of an equivalent C tracer in ambient CO2 eral decreases in D later in the study indicated significant plots prevents any straightforward comparisons of changes use of newly fixed C. in specific microbial activity level between ambient and [27] We originally expected larger differences in D values elevated CO2 treatments [Pataki et al., 2003; Hobbie et al., between CO2 treatments in shrub microsites compared to 2004; Billings and Ziegler, 2005]. Therefore, the use of vegetation‐free interspace soils as a result of greater avail- recently fixed C by microbial functional groups over time ability of recently fixed plant C. While general biomarker

7of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

Figure 3. Soil volumetric water content effects on fungal and bacterial PLFA‐C(mgCg−1) and fungal‐ to‐bacterial PLFA‐C ratios in interspace soils and shrub microsites. Relationships did not differ between CO2 in interspace soils, so data were pooled for regressions. Regression statistics are shown within each panel.

18:1w9t showed greater use of new C in soils under L. arid ecosystems [Evans and Ehleringer, 1993; Evans and tridentata as expected, both fungal biomarkers showed Belnap, 1999]. Indeed, the fungal biomarker 18:3w3 found greater use of new C in interspace soils. The latter suggests in the present study has also been identified in the N‐fixing that recently fixed C is also contributing C substrate for photobiont of the dominant lichen in biological soil crusts microbial use in resource‐poor microsites, possibly from [Bowker et al., 2008]. Thus, significant changes in these roots extending into plant interspaces. Plant inputs via root biomarkers in interspace soils but not in soils under L. exudation, fungal‐root litter, or fine root turnover from tridentata indicate microsite‐specific changes in soil C and larger woody vegetation are possible for shrubs such as L. N processes in the Mojave Desert in response to increasing tridentata which can extend its roots beyond 5 m horizon- atmospheric CO2 concentrations. tally from the stem base [Jordan et al., 1999; Hartle et al., [28] In contrast, the absence of any change in biomarker 2006]. Alternatively, recently fixed C in interspace micro- D values after long‐term exposure to elevated CO2 suggests sites also could be derived directly through exudates from that (1) certain microorganisms or microbial groups may be the cryptobiotic soil crust which contains both autotrophic more heavily dependent on older soil organic matter (SOM) and heterotrophic organisms, some of which also are capa- compared to newer plant‐derived C, (2) significant transfer ble of N fixation and can be the primary source of N in some of newly fixed C into the older SOM compartments used by

8of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

Figure 4. Relative abundances of functional groups (% total PLFA‐C) in (a) interspace soils and (b) soils under L. tridentata. Data shown are untransformed mean values ± standard error (n = 3). Aster- isks indicate CO2 differences (P ≤ 0.05). these microorganisms has not occurred within the 2 years bacterial species are involved in N fixation [Chalk, 1991; following the CO2 source change, and/or (3) recent photo- Huss‐Danell, 1997]. synthate C allocation into some PLFA biomarkers may have been too low to detect. While quantifying microbial use of 4.2. Soil Microbial Pool Size specific soil C pools or microbial C turnover is beyond the [29] Previously observed increases in microbial substrate scope of this study, the absence of changes in D values of use diversity and gross N cycling rates under elevated CO actinobacterial biomarkers suggested little to no change in 2 have been attributed to changes in biomass or activity of soil C substrates use by actinobacteria. Actinobacteria spe- different functional groups, depending on microsite. Jin and cialize in metabolizing less labile substrates such as chitin, Evans [2007] hypothesized that increases in microbial cellulose, and hemicellulose, particularly in alkaline soils activity under elevated CO in interspace soils were driven or soils that experience significant water or heat stress 2 by greater microbial biomass as evidenced by higher [Killham, 1994]. Actinobacteria have been identified as a microbial biomass N, while higher microbial activity in potential indicator of declines in the quality of the miner- shrub microsites was due to greater microbial turnover more alizable SOM fraction under elevated CO , and studies from 2 so than changes in biomass. In contrast to these hypothe- other ecosystems show an active actinobacterial component sized trends, this study showed that total PLFA‐C in inter- in soils exposed to elevated CO [Ringelberg et al., 1997; 2 spaces was not affected by CO treatment and increased in Billings and Ziegler, 2005; Denef et al., 2007]. In this study 2 soils under L. tridentata under elevated CO . These differ- however, actinobacterial relative abundance in interspace 2 ences indicate that measured changes in microbial biomass soils exposed to elevated CO was lower than under ambient 2 N may not translate directly to changes in microbial biomass conditions. This decrease, coupled with the absence of C (as estimated using PLFA‐C). Rather, changes in micro- changes in D values, may signal microsite‐specific changes bial biomass N, such as those observed in interspace soils, in SOM cycling by actinobacteria in these Mojave Desert could reflect increases in cellular solute concentrations of soils under elevated CO . Such changes could also con- 2 N‐bearing compounds (i.e., amino compounds) which are tribute to changes in soil N cycling because some actino- used in osmoregulation to cope with drought stress in some

9of12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001 microorganisms [Miller and Wood,1996;Schimel et al., 2007; lations could also affect fungal‐to‐bacterial ratios by Williams and Xia, 2009]. differentially affecting rates of predation on soil bacteria or [30] Microsite effects on total PLFA‐C were also apparent fungi which can alter soil C and N cycling [Bouman and in negative relationships with soil volumetric water content Zwart, 1994]. (VWC). Increasing total PLFA‐C with decreasing VWC [34] While the impact of elevated CO2 on fungal‐to‐ suggested that soil microbial biomass or membrane mate- bacterial ratios is not straightforward and will also depend rials as biomass were more abundant under drier conditions, on other factors (i.e., moisture, temperature, nutrient avail- reflecting the long‐term adaptations that enable microbial ability and their interactions) [Castro et al., 2010], we found communities to maintain and even grow under the environ- that changes in fungal‐to‐bacterial PLFA‐C ratio became mental stresses imposed [Walker et al., 2006]. Similar in- less sensitive to decreasing VWC in soils exposed to ele- creases in microbial phospholipid abundance under dry vated CO2 compared to ambient conditions in shrub mi- conditions compared to postprecipitation have also been crosites, but not in interspace soils. The increased variability reported in more arid zones of the Judean Desert [Steinberger in fungal‐to‐bacterial responses under elevated CO2 illus- et al., 1999]. trates the difficulty in partitioning causality among many correlated factors [Strickland and Rousk, 2010], which is 4.3. Fungal and Bacterial Responses further complicated by both temporal and spatial extremes in [31] The magnitude, direction, and timing of microbial resource availability in this extreme desert environment. biomarker responses in soils exposed to elevated CO2 were strongly dependent on soil water content and on microsite, 5. Conclusions consistent with previous studies at this site [Smith et al., 2000; Billings et al., 2002a, 2002b, 2004; Schaeffer et al., [35] Overall, the transfer of plant C recently fixed from 2003, 2007; Jin and Evans, 2007] and in other ecosystems fossil CO2 sources was evident in two general and two [Schortemeyer et al., 1996; Montealegre et al., 2002]. Fungal fungal biomarkers in soils treated with elevated CO2 com- PLFA‐C tended to be greater under elevated CO2, and pared to ambient soils. Fungal and bacteria PLFA quantities though not significant, this trend was consistent with other with elevated CO2 depended strongly on soil microsite studies finding greater fungal activity and/or growth under and water availability, illustrating how temporal and spatial elevated CO2 [Hungate et al., 2000; Weimken et al., 2001; heterogeneity in resource availability as well as in the dis- Phillips et al., 2002; Janus et al., 2005; Lipson et al., 2006; tribution of soil microbial communities themselves inter- Treseder and Turner, 2006]. Bacterial PLFA‐C, however, acted to affect microbial community responses to elevated ‐ was unaffected by CO2 treatment. These results are similar atmospheric CO2 in this pulse adapted ecosystem. to a study in a chaparral ecosystem [Allen et al., 2005; [36] Because the capacities for growth, nutrient acquisi- Lipson et al., 2006] and a temperate grassland [Sonnemann tion, or activity levels are likely to differ between microbial and Wolters, 2005], but contrast with trends found in species within any functional group [Billings and Ziegler, another grassland ecosystem [He et al., 2010]. 2005], broad changes in the composition of microbial bio- [32] The ratio of fungal to bacterial PLFA‐C was not markers could indicate potential changes to processes that affected by elevated CO2, but both fungal and bacterial are mediated by these microorganisms [Hungate et al., 1996; PLFA‐C increased with decreasing VWC. This negative Montealegre et al., 2002; Janus et al., 2005]. The negative relationship persisted for fungal‐to‐bacterial PLFA‐C ratios relationship between soil VWC and PLFA‐C suggested a under ambient CO , but the relationship decoupled under greater abundance or biomass of soil microorganisms under 2 ‐ elevated CO2. The general negative relationship between drier conditions, reflecting the long term adaptations that fungal and bacterial PLFA‐C and soil VWC under ambient enable microbial communities to maintain and even grow CO2 suggests that fungi may be more drought‐tolerant than under the extreme environmental stresses imposed. Increases bacteria [Schimel et al., 2007]. Under water‐limited condi- in fungal PLFA‐C and water‐related shifts in fungal‐ tions, the availability of other soil resources (i.e., N) is also to‐bacterial ratios may have also contributed to measured likely to be limited in these desert soils. Compared to bac- increases in fungal‐associated extracellular enzyme activities teria, fungi can maintain their activities under more stressful [Jin and Evans, 2007]. While other covarying factors are environmental conditions (e.g., lower nutrient and water likely to influence microbial functional structure, our data availability, osmotic shock during soil wetting events) suggest that fungal responses to changes in soil water avail- [Harris, 1981; Holland and Coleman, 1987] and are more ability may be key to how soil organic C and nutrients are efficient in transforming soil C substrates into biomass cycled in this desert ecosystem. Further, the strong control because of stoichiometric differences [Killham, 1994; Zak exerted by soil water availability on soil fungal and bacterial et al., 1996]. This relationship also may have reflected abundances emphasize the importance of stress on microbial faster growth by bacteria relative to soil fungi when the responses to environmental changes, such as elevated availability of soil water (and subsequently, soil nutrients) atmospheric CO2, particularly in extreme environments. increased. [33] Fungal‐to‐bacterial ratios are highly dynamic, how- [37] Acknowledgments. The authors acknowledge the support from ever, and can reflect multiple covarying factors [Hungate the U.S. National Science Foundation (NSF‐DEB‐0424979, NSF‐MRI‐ et al., 2000; Strickland and Rousk, 2010]. Functional group‐ 0421478 to RDE) for this research, and the U.S. Department of Energy specific responses that affect soil trophic interactions could Terrestrial Carbon Processes program (award DE‐FG02‐03ER63651) for also contribute to changes in fungal‐to‐bacterial dominance. research and operational support to the NDFF. S.M.S. was supported by aNSF‐EPSCoR cellular and molecular biology fellowship and NSF‐Doctoral For example, differences in the drought tolerance of bac- Dissertation Improvement Grant (DEB‐0309219). The authors also thank terivorous or fungivorous protozoan and nematode popu- C. Chambers, B. Harlow, and A. Koyama for laboratory assistance and

10 of 12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

R. Alldredge for statistical consulting at Washington State University. This Harris, R. (1981), Effect of water potential on microbial growth and activity, manuscript was greatly improved by excellent comments from P. Fay. in Water Potential Relations in Soil Microbiology, edited by J. Parr, W. Gardner, and L. Elliot, pp. 23–95, Soil Sci. Soc. of Am., Madison, Wis. Hartle, R. T., G. C. J. Fernandez, and R. S. Nowak (2006), Horizontal and vertical zones of influence for root systems of four Mojave Desert shrubs, References J. Arid Environ., 64, 586–603. ‐ He, X., M. Xu, Y. Deng, S. Kang, L. Kellogg, L. Wu, J. D. Van Nostrand, Abraham, W. R., C. Hesse, and O. Pelz (1998), Ratios of carbon isotopes S. E. Hobbie, P. B. Reich, and J. Zhou (2010), Metagenomic analysis in microbial lipids as an indicator of substrate usage, Appl. Environ. – reveals a marked divergence in the structure of belowground microbial Microbiol., 64, 4202 4209. communities at elevated CO , Ecol. Lett., 13, 564–575. Abrajano, T. A., Jr., D. E. Murphy, J. Fang, P. Comet, and J. M. Brooks 2 13 12 Hendrey, G. R., and B. A. Kimball (1994), The FACE Program, Agric. For. (1994), C/ C ratios in individual fatty acids of marine mytilids with – – Meteorol., 70,3 14. and without bacterial symbionts, Org. Geochem., 21, 611 617. Hobbie, E. A., et al. (2004), Isotopic estimates of new carbon inputs into Allen, M. F., J. N. Klironomos, K. K. Treseder, and W. C. Oechel (2005), litter and soils in a four‐year climate change experiment with Douglas‐fir, Responses of soil biota to elevated CO2 in a chaparral ecosystem, Ecol. – – Plant Soil, 259,331 343. Appl., 15, 1701 1711. Holland, E. A., and D. C. Coleman (1987), Litter placement effects on Austin, A. T., L. Yahdjian, J. M. Stark, J. Belnap, A. Porporato, U. Norton, microbial and organic‐mater dynamics in an agroecosystem, Ecology, D. A. Ravetta, and S. M. Schaeffer (2004), Water pulses and biogeochem- – – 68, 425 433. ical cycles in arid and semiarid ecosystems, Oecologia, 141, 221 235. Housman, D. C., E. Naumburg, T. E. Huxman, T. N. Charlet, R. S. Nowak, Billings, S. A., and S. E. Ziegler (2005), Linking microbial activity and soil and S. D. Smith (2006), Increases in desert shrub productivity under ele- organic matter transformations in forest soils under elevated CO2, Global – – vated carbon dioxide vary with water availability, Ecosystems, 9, 374 385. Change Biol., 11,1 10. Hungate, B. A., R. B. Jackson, C. B. Field, and F. S. Chapin III (1996), Billings, S. A., S. M. Schaeffer, and R. D. Evans (2002a), Trace N gas Detecting changes in soil carbon in CO2 enrichment experiments, Plant losses and N mineralization in Mojave Desert soils exposed to elevated – – Soil, 187, 135 145. CO2, Soil Biol. Biochem., 34, 1777 1784. Hungate, B. A., F. S. Chapin III, H. Zhong, E. A. Holland, and C. B. Field Billings, S. A., S. M. Schaeffer, S. Zitzer, T. Charlet, S. C. Smith, and R. D. (1997), Stimulation of grassland nitrogen cycling under carbon dioxide Evans (2002b), Alterations of nitrogen dynamics under elevated CO2 in enrichment, Oecologia, 109, 149–153. an intact Mojave Desert ecosystem, evidence from nitrogen‐15 natural – Hungate, B. A., C. H. Jaeger III, G. Gamara, F. S. Chapin III, and C. B. abundance, Oecologia, 131, 463 467. Field (2000), Soil microbiota in two annual grasslands response to ele- Billings, S. A., S. F. Zitzer, H. Weatherly, S. M. Schaeffer, T. Charlet, J. A. vated atmospheric CO2, Oecologia, 124, 589–598. Arnone, and R. D. Evans (2003), Effects of elevated carbon dioxide on Hunter, R. B. (1994), Status of the flora and fauna on the Nevada Test Site, green leaf tissue and leaf litter quality in an intact Mojave Desert ecosys- 1994, DOE/NV/11432–195 UC‐721, Natl. Tech. Inf. Serv., Springfield, Va. tem, Global Change Biol., 9, 729–735. Huss‐Danell, K. (1997), Actinorhizal symbioses and their N2 fixation, Billings, S. A., S. M. Schaeffer, and R. D. Evans (2004), Soil microbial New Phytol., 136, 375–405. activity and N availability with elevated CO2 in Mojave Desert soils, Janus, L. R., N. L. Angeloni, J. McCormack, S. T. Rier, N. C. Tuchman, Global Biogeochem. Cycles, 18, GB1011, doi:10.1029/2003GB002137. and J. J. Kelly (2005), Elevated atmospheric CO alters soil microbial Bligh, E. G., and W. J. Dyer (1959), A rapid method of total lipid extraction 2 – communities associated with trembling aspen (Populus tremuloides) and purification, Can. J. Biochem. Physiol., 37, 911 917. roots, Microb. Ecol., 50, 102–109. Bouman, L. A., and K. B. Zwart (1994), The cology of bacterivorous proto- Jin, V. L., and R. D. Evans (2007), Elevated CO increases microbial car- zoans and nematodes in arable soil, Agric. Ecosyst. Environ., 51, 145–160. 2 ‐ bon substrate use and nitrogen cycling in Mojave Desert soils, Global Bowker, M. A., N. C. Johnson, J. Belnap, and G. W. Koch (2008), Short Change Biol., 13, 452–465. term monitoring of aridland lichen cover and biomass using photography 13 – Jin, V. L., and R. D. Evans (2010), Microbial C utilization patterns via and fatty acids, J. Arid Environ., 72, 869 878. stable isotope probing of phospholipid biomarkers in Mojave Desert soils Castro, H. F., A. T. Classen, E. E. Austin, R. J. Norby, and C. W. Schadt exposed to ambient and elevated atmospheric CO2, Global Change Biol., (2010), Soil microbial community responses to multiple experimental cli- – – 16, 2334 2344. mate change drivers, Appl. Environ. Microbiol., 76, 999 1007. Jordan, D. N., S. F. Zitzer, G. R. Hendry, K. F. Lewin, J. Nagy, R. S. Nowak, Chalk, P. M. (1991), The contribution of associative and symbiotic nitrogen S. D. Smith, J. S. Coleman, and J. R. Seeman (1999), Biotic, abiotic and fixation to the nitrogen nutrition of non‐legumes, Plant Soil, 132,29–39. performance aspects of the Nevada Desert Free‐Air CO2 Enrichment Charlet, D. (2007), NDFF precipitation data, Carbon Dioxide Inf. Anal. (FACE) Facility, Global Change Biol., 5, 659–668. Cent., Oak Ridge Natl. Lab., U.S. Dep. of Energy, Oak Ridge, Tenn. Kandeler, E., D. Tscherko, R. D. Bardgett, P. J. Hobbs, C. Kampichler, and (Available at http://cdiac.ornl.gov.) T. H. Jones (1998), The response of soil microorganisms and roots to ele- Chung, H., D. R. Zak, and E. A. Lilleskov (2006), Fungal community com- vated CO2 and temperature in a terrestrial model ecosystem, Plant Soil, position and metabolism under elevated CO2 and O3, Oecologia, 147, – – 202, 251 262. 143 154. Killham, K. (1994), Soil Ecology, Cambridge Univ. Press, Cambridge, U. K. Denef, K., H. Bubenheim, K. Lenhart, J. Vermeulen, O. Van Cleemput, Lipson,D.A.,M.Blair,G.Barron‐Gafford, K. Grieve, and R. Murthy P. Boeckx, and C. Muller (2007), Community shifts and carbon transloca- (2006), Relationships between microbial community structure and soil tion within metabolically active rhizosphere microorganisms in grasslands – processes under elevated atmospheric carbon dioxide, Microb. Ecol., under elevated CO2, Biogeosciences, 4, 769 779. 51, 302–314. Dobbs, F. C., and R. H. Findlay (1993), Analysis of microbial lipids to Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Winfinger (1996), determine biomass and detect the response of sedimentary microorganisms SAS System for Mixed Models, SAS Inst., Cary, N. C. to disturbance, in Handbook of Methods in Aquatic Microbial Ecology, – Melillo,J.M.,D.A.McGuire,D.W.Kicklighter,B.Moore,C.J. edited by P. F. Kemp et al., pp. 347 358, Lewis, Boca Raton, Fla. Vorosmarty, and A. L. Schloss (1993), Global climate change and terres- Drigo, B., G. A. Kowalchuk, and J. A. vanVen (2008), Climate change trial net primary production, Nature, 363, 234–240. goes underground: Effects of elevated atmospheric CO2 on microbial Mikha, M. M., C. W. Rice, and G. A. Milliken (2005), Carbon and nitro- community structure and activities in the rhizosphere, Biol. Fertil. Soils, – gen mineralization as affected by drying and wetting cycles, Soil Biol. 44, 667 679. Biochem., 37, 339–347. Ebersberger, D., N. Wermbter, P. A. Niklaus, and E. Kandeler (2004), Miller, K. J., and J. M. Wood (1996), Osmoadaptation by rhizosphere bac- Effects of long term CO2 enrichment on microbial community structure in – – teria, Annu. Rev. Microbiol., 50, 101 136. calcareous grassland, Plant Soil, 264, 313 323. Montealegre, C. M., C. van Kessel, M. P. Russelle, and M. J. Sadowsky Evans, R. D., and J. Belnap (1999), Long‐term consequences of distur- – (2002), Changes in microbial activity and composition in a pasture eco- bance on nitrogen dynamics in an arid ecosystem, Ecology, 80, 150 160. system exposed to elevated atmospheric carbon dioxide, Plant Soil, 243, Evans, R. D., and J. R. Ehleringer (1993), A break in the nitrogen cycle of 197–207. aridland: Evidence from d15N of soils, Oecologia, 94, 314–317. ‐ Nobel, I. R., H. Gitay, A. N. Alwelaie, M. T. Hoffman, and A. R. Saunders Fierer, N., and J. P. Schimel (2002), Effects of drying rewetting frequency (1996), Deserts in a changing climate, Impacts, in Contribution of Working on soil carbon and nitrogen transformations, Soil Biol. Biochem., 34, – Group II to the Second Assessment Report of the Intergovernmental Panel 777 787. on Climate Change, edited by R. T. Watson et al., pp. 159–169, Cambridge Gordon, H., P. M. Haygarth, and R. D. Bardgett (2008), Drying and rewet- Univ. Press, New York. ting effects on soil microbial community composition and nutrient leach- Nowak, R. S., D. S. Ellsworth, and S. D. Smith (2004), Functional responses ing, Soil Biol. Biochem., 40, 302–311. of plants to elevated atmospheric CO2—Do photosynthetic and productiv-

11 of 12 G02001 JIN ET AL.: SOIL MICROBIAL RESPONSES TO CO2 G02001

ity from FACE experiments support early predictions?, New Phytol., 162, Strickland, M. S., and J. Rousk (2010), Considering fungal:bacterial dom- 253–280. inance in soils—Methods, controls, and ecosystem implications, Soil Owensby, C. E., J. M. Ham, A. K. Knapp, and L. M. Auen (1999), Biomass Biol. Biochem., 42, 1385–1395. production and species composition change in a tallgrass prairie ecosys- Titus, J. H., R. S. Nowak, and S. D. Smith (2002), Soil resource heteroge- tem after long‐term exposure to elevated atmospheric CO2, Global neity in the Mojave Desert, J. Arid Environ., 52, 269–292. Change Biol., 5, 497–506. Treseder, K. K., and K. M. Turner (2006), Glomalin in ecosystems, Soil Pataki, D. E., et al. (2003), Tracing changes in ecosystem function under Sci. Soc. Am. J., 71, 1257–1266. elevated carbon dioxide conditions, BioScience, 53, 805–818. Tunlid, A., and D. C. White (1992), Biochemical analysis of biomass, com- Pendall, E., et al. (2004), Below‐ground process responses to elevated CO2 munity structure, nutritional status, and metabolic activity of microbial and temperature, a discussion of observations, measurement methods, communities in soil, in Soil Biochemistry, vol. 7, edited by G. Stotzky and models, New Phytol., 162, 311–322. and J. M. Bollag, pp. 229–262, Marcel Dekker, New York. Phillips, R. L., D. R. Zak, W. E. Holmes, and D. C. White (2002), Micro- Waldrop, M. P., and M. K. Firestone (2004), Altered utilization patterns of bial community composition and function beneath temperate trees young and old soil C by microorganisms caused by temperature shifts exposed to elevated atmospheric carbon dioxide and ozone, Oecologia, and N additions, Biogeochemistry, 67, 235–248. 131, 236–244. Walker, V. K., G. R. Palmer, and G. Voordouw (2006), Freeze‐thaw toler- Pinkart, H. C., D. B. Ringelberg, Y. M. Piceno, S. J. MacNaughton, and ance and clues to the winter survival of a soil community, Appl. Environ. D. C. White (2002), Biochemical approaches to biomass measure- Microbiol., 72, 1784–1792. ments and community structure analysis, in Manual of Environmental Weimken, V., E. Laczko, K. Ineichen, and T. Boller (2001), Effects of ele- Microbiology, edited by C. J. Hurst et al., pp. 101–113, Am. Soc. of vated carbon dioxide and nitrogen fertilization on mycorrhizal fine roots Microbiol. Press, Washington, D. C. and the soil microbial community in beech‐spruce ecosystems on sili- Ringelberg, D. B., et al. (1989), Validation of signature polarlipid fatty acid ceous and calcareous soil, Microb. Ecol., 42, 126–135. biomarkers for alkane‐utilizing bacteria in soils and subsurface aquifer White, D. C., and D. B. Ringelberg (1998), Signature lipid biomarker anal- materials, FEMS Microbiol. Ecol., 62,39–50. ysis, in Techniques in Microbial Ecology, edited by R. D. Burlage et al., Ringelberg, D. B., J. O. Stair, J. Almeida, R. J. Norby, E. G. O’Neill, and pp. 255–272, Oxford Univ. Press, New York. D. C. White (1997), Consequences of rising atmospheric carbon dioxide Williams, M. A., and K. Xia (2009), Characterization of the water soluble levels for the belowground, J. Environ. Qual., 26, 495–503. soil organic pool following the rewetting of dry soil in a drought‐prone Rønn, R., M. Gavito, J. Larsen, I. Jakobsen, H. Frederiksen, and S. Chris- tallgrass prairie, Soil Biol. Biochem., 41,21–28. tensen (2002), Response of free‐living soil protozoa and microorganisms Xiang, S.‐R., A. Doyle, P. A. Holden, and J. P. Schimel (2008), Drying and to elevated atmospheric CO2 and presence of mycorrhizal, Soil Biol. Bio- rewetting effects on C and N mineralization and microbial activity in sur- chem., 34, 923–932. face and subsurface California grassland soils, Soil Biol. Biochem., 40, Schaeffer, S. M. (2005), Determining the effects of global change on soil 2281–2289. nitrogen cycling in arid ecosystems, Ph.D. dissertation, Dep. of Biol. Zak, D. R., D. B. Ringelberg, K. S. Pretiztzer, D. L. Randlett, D. C. White, Sci., Univ. of Ark., Fayetteville. and P. S. Curtis (1996), Soil microbial communities beneath Populus Schaeffer, S. M., S. A. Billings, and R. D. Evans (2003), Responses of soil gradidentata grown under elevated atmospheric CO2, Ecol. Appl., 6, nitrogen dynamics in a Mojave Desert ecosystem to manipulations in soil 257–262. carbon and nitrogen availability, Oecologia, 134, 547–553. Zak, D. R., K. S. Pretiztzer, P. S. Curtis, and W. E. Holmes (2000), Atmo- Schaeffer, S. M., S. A. Billings, and R. D. Evans (2007), Laboratory incu- spheric CO2 and the composition and function of soil microbial commu- bations reveal potential responses of soil nitrogen cycling to changes in nities, Ecol. Appl., 10,47–59. soil C and N availability in Mojave Desert soils exposed to elevated Zelles, L. (1999), Fatty acid patterns of phospholipids and lipopolysacchar- atmospheric CO2, Global Change Biol., 13, 854–865. ides in the characterisation of microbial communities in soil, a review, Schimel, J., T. C. Balser, and M. Wallenstein (2007), Microbial stress‐ Biol. Fertil. Soils, 29, 111–129. response physiology and its implications for ecosystem function, Ecol- Zelles, L., Q. Y. Bai, T. Beck, and F. Besse (1992), Signature fatty acids in ogy, 88, 1386–1394. phospholipids and lipopolysaccharides as indicators of microbial biomass Schlesinger, W. H., J. F. Raikes, A. E. Hartley, and A. F. Cross (1996), On and community structure in agricultural soils, Soil Biol. Biochem., 24, the spatial pattern of soil nutrients in desert ecosystems, Ecology, 77, 317–323. 364–374. Ziegler, S., and S. Billings (2011), Soil nitrogen status as a regulator of car- Schortemeyer,M.,U.A.Hartwig,G.R.Hendrey,andM.J.Sadowsky bon substrate flows through microbial communities with elevated CO2, (1996), Microbial community changes in the rhizospheres of white clover J. Geophys. Res., 116, G01011, doi:10.1029/2010JG001434. and perennial ryegrass exposed to free air carbon dioxide enrichment – (FACE), Soil Biol. Biochem., 28, 1717 1724. R. D. Evans, School of Biological Sciences, Washington State Smith, S. D., T. E. Huxman, S. F. Zitzer, T. N. Charlet, D. C. Housman, J. S. University, Pullman, WA 99164‐4236, USA. Coleman, L. K. Fenstermaker, J. R. Seeman, and R. S. Nowak (2000), Ele- V. L. Jin, Agroecosystem Management Research Unit, Agricultural vated CO2 increases productivity and invasive species success in an arid – Research Service, U.S. Department of Agriculture, University of ecosystem, Nature, 408,79 82. Nebraska at Lincoln, 137 Keim Hall, E. Campus, Lincoln, NE 68583‐ Sonnemann, I., and V. Wolters (2005), The microfood web of grassland 0938, USA. ([email protected]) soils responds to a moderate increase in atmospheric CO2, Global – S. M. Schaeffer, Department of Ecology, Evolution, and Marine Biology, Change Biol., 11, 1148 1155. University of California, Santa Barbara, CA 93106‐9610, USA. Steinberger, Y., L. Zelles, Q. Y. Bai, M. von Lützow, and J. C. Munch S. E. Ziegler, Department of Earth Sciences, Memorial University of (1999), Phospholipid fatty acid profiles as indicators for the microbial Newfoundland, St. John’s, NL A1B 3X5, Canada. community structure in soils along a climatic transect in the Judean Desert, Biol. Fertil. Soils, 28, 292–300.

12 of 12