Abstract
Imbalances in C:N:P supply ratios may cause bacterial resource limitations and constrain biogeochemical processes, but the importance of shifts in soil stoichiometry are complicated by the nearly limitless interactions between an immensely rich species pool and a multiple chemical resource forms. To more clearly identify the impact of soil C:N:P on bacteria, we evaluated the cumulative effects of single and coupled long-term nutrient additions (i.e., C as mannitol, N as equal concentrations NH4+ and NO3−, and P as Na3PO4) and water on communities in an Antarctic polar desert, Taylor Valley. Untreated soils possessed relatively low bacterial diversity, simplified organic C sources due to the absence of plants, limited inorganic N, and excess soil P potentially attenuating links between C:N:P. After 6 years of adding resources, an alleviation of C and N colimitation allowed one rare Micrococcaceae, an Arthrobacter species, to dominate, comprising 47% of the total community abundance and elevating soil respiration by 136% relative to untreated soils. The addition of N alone reduced C:N ratios, elevated bacterial richness and diversity, and allowed rare taxa relying on ammonium and nitrite for metabolism to become more abundant [e.g., nitrite oxidizing Nitrospira species (Nitrosomonadaceae), denitrifiers utilizing nitrite (Gemmatimonadaceae) and members of Rhodobacteraceae with a high affinity for ammonium]. Based on community co-occurrence networks, lower C:P ratios in soils following P and CP additions created more diffuse and less connected communities by disrupting 73% of species interactions and selecting for taxa potentially exploiting abundant P. Unlike amended nutrients, water additions alone elicited no lasting impact on communities. Our results suggest that as soils become nutrient rich a wide array of outcomes are possible from species dominance and the deconstruction of species interconnectedness to the maintenance of biodiversity.
Introduction
Environmental conditions dramatically structure soil bacterial communities; however, only a few environmental variables such as pH, salinity, and C substrate quality and quantity are known to drive community assemblages (; ; ). Borrowing from foundations in plant community ecology, following light and water, nutrient additions dramatically alter species abundance within communities often allowing certain species to dominate and biodiversity to decline (; ). For example, in a grassland, even low continual additions of N reduced species richness in just 2 years and depressed the number of species for as long as 20 years (). Classically, soil bacterial metabolism and growth is limited by water availability, then the quality and quantity of C substrates, and finally nutrient concentrations. But the role of major nutrients, such as N and P, remains incomplete even though changes in nutrient availability shape the responses of specific bacterial species or species interactions within soil communities (; ). Unlike plants, bacterial responses to resource constraints are complicated by interactions within consortia or communities where functionally disparate taxa (e.g., decomposers, nitrifiers, and methanogens) potentially dictate the form and availability of specific C substrates and nutrients necessary for other bacteria to become metabolically active and grow. In most soil communities such interactions are further complicated because soil organic C substrates are extremely numerous and diverse, containing both labile and more recalcitrant sources structuring the availability of N and P (; ). Furthermore, the C, N, and P requirements of bacterial biomass differ among species and ecosystems, and are not homeostatic through time (; ).
Ecological stoichiometry is a unifying body of theory in ecology predicting relationships between the organismal biochemistry of plants, invertebrates, and microorganisms and the availability and recycling of nutrient elements in the environment (). Ecological stoichiometry may also help identify the resource requirements of bacterial taxa and the conditions allowing certain bacteria to become metabolically active. Ecological stoichiometric theory was developed in aquatic ecosystems, but is universally valid, and over the last decades was also successfully applied to terrestrial systems (; ; ; ). Soil (186:13:1) and soil microbial (60:7:1) C:N:P stoichiometry, like Redfield ratios for planktonic biomass (C:N:P = 106:16:1) (), are well-constrained across multiple biomes () offering incredible utility in understanding bacterial resource limitations and constraints on biogeochemical processes (; ). The C:N:P stoichiometry of plant residues, soil organic matter, and bacterial biomass influence litter decomposition rates (; ), N and P mineralization rates (), and C-use-efficiency of bacteria, determining metabolic activity and trace gas flux (). Community structure is intimately connected to C:N:P ratios (). Specifically, soil C:N:P stoichiometry sheds light on the potential for N and P availability to influence bacterial community structure. For example, higher N:P ratios in afforested soils of the Loess Plateau in China reflected P deficiencies among bacteria, leading to lower diversity but a higher abundance of Proteobacteria, Acidobacteria, and Nitrospirae (). Further, a decrease in soil C:P ratios caused Gram-positive bacterial biomass to increase by 22% and the abundance of arbuscular mycorrhizal fungi to increase by 46% in a pasture following slash-and-burn agriculture in the South Ecuadorian Andes (). Taken together, investigating soil C:N, C:P, and N:P ratios are instrumental in identifying patterns of ecological coherence among responding bacteria under varying resource conditions.
Soil ecosystems of the McMurdo Dry Valleys, Eastern Antarctica are a model system for investigating stoichiometric controls over soil communities and ecosystem processes (). The extreme environment limits biota to cryptogrammic vegetation, a few taxa of metazoan invertebrates, and microbial dominated food webs. Phylum-level bacterial diversity in Antarctic soils is surprisingly high considering the environmental extremes and dearth of resources, i.e., organic matter and available nutrients (; ). However, these soils host comparatively low diversity at the family or genus level relative to other biomes (), attenuating the nearly limitless possibilities of links between C:N:P stoichiometry and communities potentially present in high diversity ecosystems. Further, soil C:N:P ratios in most systems are necessarily complicated by plant residues with multiple different stoichiometric ratios that may mask links between C sources and release rate of N and P (). Alternatively, due to the absence of vascular plants, Antarctic Dry Valley soils have some of the lowest soil organic matter concentrations on Earth (; ) with much of the soil organic matter being “legacy” C accumulated over thousands of years by cryptoendolithic bacteria, paleolake deposition, and minor inputs of contemporary algae and cyanobacteria from lakes, intermittent streams and saturated zones (). The concentrations of inorganic N and P are relatively low with N entering the system via atmospheric deposition (), endolithic and hypolithic cyanobacterial N2 fixation over millions of years (), and from dust, while P enters soils through mineral weathering (; ). Both N and P concentrations vary among soils occurring on glacial tills with distinct exposure age and mineralogy (). Accompanying nutrient limitations, water is a universal resource essential for polar bacteria. In the McMurdo Dry Valleys, low precipitation inputs and sublimation and ablation processes () cause dehydration stress and limit substrate diffusion to bacterial cells (). Water additions to Antarctic soils do create higher bacterial growth rates, elevate soil respiration, and decrease soil community diversity (; ). Therefore, water, in addition to nutrients, may have direct and indirect effects on community composition.
In this study, we explored the effects of long-term, coupled resource additions and water on bacterial species responses and ecosystem processes in a cold desert of Antarctica. After treating soils with six different resource additions including combinations of water, C as mannitol, N as equal NH4+ NO3−, and P as Na3PO4 annually over 6 years in the field, we evaluated shifts in bacterial community composition metrics such as richness, alpha diversity and evenness, taxa co-occurrence patterns, and soil respiration. Based on polar desert resource conditions, our initial soil C:N:P ratio (mean = 167:8:1, n = 8; more initial soil chemistry data is provided in the first section of the “Materials and Methods”), and the modal C:N:P ratio of soils (186:13:1, ), we hypothesized that C, N, CN, and CP additions will alleviate resource limitations and provide organic C and inorganic N for a subset of the community to exploit, while P is not limiting and adding more inorganic P will not alter community composition.
Materials and Methods
Study Site and Initial Soil Chemistry
Our study was conducted in a polar desert of the McMurdo Dry Valleys (76°30′ – 78°00′S, 160°00 – 165°00′E) at the McMurdo Long Term Ecological Research (LTER) site in Antarctica. The site was located in the Lake Fryxell basin (77°36.5′S, 163°14.9′E) of Taylor Valley, on Ross Sea drift soils (late-Quaternary) ().
Dominant soil microflora include multiple species of algae from the division Chlorophyta and Heterokontophyta; microfauna comprised of nematodes, tardigrades and rotifers; and cyanobacteria, such species as, Leptolyngbya frigida and Nostoc commune in aquatic and terrestrial habitats (). The basin experiences fewer than 50 days where average temperatures exceed 0°C within the summer months of December, January, February when mean annual temperature is -4.21°C ± 0.80 SD (n = 24). Soils receive less than 10 cm per yr−1 of effective mean annual precipitation falling as snow (). Soils are Typic Haploturbels with shallow surface layer (≈0–10 cm depth), which experience continual cryoturbation, and a perennial permafrost layer (≈30–300 cm depth) (; ). All soils are poorly developed silty-loams with an average pH of 9.69 ± 0.12 SEM (n = 8) and an electrical conductivity of 258 ± 115 μS cm−1 (n = 8). Initial soil chemistry demonstrated that soils were generally extremely low in organic C (organic C = 0.03% ± 0.003, total soil C = 0.13% ± 0.01, n = 8) and possessed relatively high soil P (2.35 ± 0.18 μg g−1 soil, n = 8) but low levels of soil N of (0.003% ± 0.0004, n = 8).
Stoichiometry and Water Long-Term Manipulations
To gain insights into the resource controls on microbial community assembly, we conducted a 6-year field stoichiometry experiment (austral summer field season 2006 – 2007 to 2011 – 2012) altering the stoichiometry of major nutrients (i.e., C, N, and P) and water availability. The experiment was a randomized block design with plots (1 m × 1 m) consisting of six treatments and an un-amended control: water only (W); C as mannitol and water (C); N as equal concentrations NH4+ and NO3− with water (N); P as Na3PO4 and water (P); C, N, and water (CN); C, P, with water (CP); and the untreated control (U). The C additions, as mannitol, mirrored nutrient inputs from more contemporary algae and cyanobacteria. The N and P additions closely followed the Redfield ratio (106:16:1) to mimic new biomass from photosynthetic organisms entering organic matter-impoverished soils (). Annually, all nutrients were delivered as aqueous solutions to bring the soils to field capacity with concentrations of 15.3 g C m−2, 2.69 g N m−2 as NH4NO3, and 0.37 g P m−2 as Na3PO412H2O and water of 12.7 L H2O m−2. For more information on the treatments and treatment application in the field see . The present study comprises an analysis of the Fryxell basin site only.
Soil C:N:P and Chemistry
To measure post-amendment changes in C:N:P stoichiometry we calculated C:N:P ratios from total C and N, and extractable P, and measured soil organic C and inorganic N. Soils were collected from all treatments (5 nutrient additions with water, 1 water addition, and a control × 8 replicates = 56) with a plastic scoop to a soil depth of 10 cm (approximately 500 g), sieved (2 mm sieve), and frozen until processing. Total C and N were measured on a Elantech Flash EA 1200 (CE Elantech, NJ, United States). Extractable soil P, as phosphate, was measured in 10 g soil with 0.5 M NaHCO3 (1:5 w/v) at pH 8.5, acidified with 3 mL of 6 N HCl, and analyzed on a Lachat Autoanalyzer (). We measured dissolved organic C on a Shimadzu TOC-5000A (Shimadzu Corporation, Columbia, MD, United States). Inorganic N (μg N-NH4+ g soil−1, μg N-NO3− g soil−1) was evaluated from 20 g of soil extracted with 2 M KCl extraction (1:2.5 w/v), passed through a Whatman #1 filter, and measured on an a Lachat Quikchem 8500 (Lachat Instruments, Loveland, CO, United States). We tested for the effect of the additions on our response variables and soil C:N:P ratios using one-way ANOVA and Tukey’s HSD test to identify significant differences among the treatments in R (). For stoichiometric analyses data were converted into molar ratios.
Bacterial Community Responses to Changes in Soil C:N:P and Water
After maintaining the treatments for more than half a decade, we characterized soil communities in treatment soils using barcoded sequencing of the 16S rRNA gene. Soils were collected from three randomly selected replicates in all treatments (5 nutrient additions with water, 1 water addition, and a control × 3 replicates = 21) to a depth of 10 cm using sterile plastic scoops. All soils were transported from the field in an insulated chest, sieved to 2 mm, and stepped-down to −20°C over 24 h. Nucleic acids were extracted from 1.5 g of soil using a PowerSoil DNA Isolation Kit (Mo Bio Corporation, Carlsbad, CA, United States). We PCR amplified the V4 region of the16S rRNA gene using bacterial specific primer set 515F and 806R with unique 12-nt error correcting Golay barcodes (; ). The thermal cycle conditions consisted of an initial denaturing step at 94°C for 3 min followed by 35 cycles of denaturing at 94°C for 45 s, annealing at 50°C for 30 s, and amplifying at 72°C for 90 s. After purifying (Agencourt AMPure XP PCR Purification Beckman Coulter Inc., Brea, CA, United States) and pooling PCR amplicons at approximately equimolar concentrations, samples were sequenced at the Brigham Young University DNA Sequencing Center1 using a 454 Life Sciences genome sequencer FLX (Roche, Branford, CT, United States). All sequences were trimmed and cleaned using mothur [v. 1.31.2; All sequences were trimmed and cleaned using mothur (v. 1.31.22; )]. After removing barcodes and primers, we eliminated sequences that were <250 bp in length or possessed homopolymers longer than 8 bp. We then denoised the sequences with AmpliconNoise (), removed chimeras using UCHIME (), and eliminated chloroplast, mitochondrial, archaeal, and eukaryotic 16S rRNA gene sequences based on reference sequences from the Ribosomal Database Project (). We aligned the sequences against the SILVA database () with the SEED aligner, created operational taxonomic units (OTUs) based on uncorrected pairwise distances at 97% sequence similarity, and determined the phylogenetic identity of OTUs using the SILVA database.
To assess the effects of resources on soil bacterial communities, we first visualized differences in community composition using Principal Coordinates Analysis (PCoA) based on a Bray–Curtis distance matrix with the ‘vegan’ package in R (). We then quantitated the effects of the different resource treatments [i.e., W, C, single nutrients N and P, and combined C and nutrients (CN and CP)] on community compositions using permutational multivariate analyses of variance (PERMANOVA) () performed with the adonis function in the vegan package in R (). Second, we quantified and constructed 95% confidence intervals for estimated richness as the total number of OTUs, alpha diversity as the inverse Shannon index, and taxa evenness using Pielous’s evenness based on 1000 iterations of 900 random resampled seqeunces from each replicate (). Last, we calculated the relative recovery of nine phyla and three subclasses to identify differences in the distribution of major taxonomical groups (recovery ≥ 1.0%) due to the resource additions. Taxonomic trends of 24 families (recovery ≥ 1.0% in at least one replicate) were visualized in a heat map with hierarchal clustering using the heatmap function in the ‘gplot’ package in R ().
Bacterial Community Network Models
To assess changes within communities at the OTU taxonomical level, we created network co-occurrence models for combinations of resources based on maximal information coefficient (MIC) analysis. We calculated all possible linear and non-linear associations between OTUs using the mic function and the ‘minerva’ package in R (), which belongs to a class of maximal information-based non-parametric exploration statistics for identifying and classifying relationships (). The nodes in the networks represented individual OTUs at 97% identity, while edges corresponded to valid or significant co-occurrence connections that occurred in at least 75% of all samples and had a MIC that was both >0.7 and statistically significant (P-value = 0.01) (). This filtering facilitated the determination of the OTUs interacting within the treatments and removed poorly represented OTUs reducing network complexity (). We described the network through a series of topological parameters: mean path length, mean degree, mean clustering coefficient, density, and modularity (). Network graphs in the graphml format were generated using ‘igraph’ package in R () and visualized with the interactive platform Gephi (v. 0.8.2-beta) (). To identify the taxonomy of bacteria within the networks, we elevated nodes at the order taxonomical rank. We calculated the node number as the total number of nodes within each of the nineteen orders comprising the networks, and the relative recovery of nodes as the summation of the mean relative recoveries of the nodes within an a given order from P and CP or U and W communities.
Soil Respiration
To investigate the links between microbial communities and ecosystem processes, we measured soil respiration (μmoles C-CO2 m−2 soil sec−1) in all resource treatments (5 nutrient additions with water, 1 water addition, and a control × 8 replicates = 56). Within 1 week of resource additions, soil CO2 flux in the field was evaluated using a Li-COR 8100 (LI-COR Biosciences, Lincoln, NE, United States) with a 10-cm diameter PVC ring inserted 2 cm into the soil at least 1 h prior to measurement ().
Results
N and P Additions Altered Soil C:N:P
Resource additions clearly altered soil C:N:P leading to increases in N and P availability. Following multiple years of N additions, C:N were lower and N:P were higher in N and CN than P and CP soils (Table 1). The shifts in ratios were highlighted by inorganic NH4+ (one-way ANOVA, df = 6, F = 39.1, P < 0.001) and NO3− (one-way ANOVA, df = 6, F = 14.9, P < 0.001) concentrations being at least forty-four- and nine-times higher, respectively, in N and CN than all other soil treatment (Table 2). Accompanying soil P additions, C:P were lower in P and CP soils than all other treatments besides CN (Table 1). Similarly, extractable P increased more than 53.0% in P and CP relative to W, C, and N (one-way ANOVA, df = 6, F = 6.26, P < 0.001, Table 2). Conversely, the additions of C had no apparent effect on total soil C only slightly increasing SOC in the C compared to P and W treatments (one-way ANOVA, df = 6, F = 3.80, P = 0.004, Table 2). The C:N:P ratios for each treatment are listed in Table 1.
Table 1
| C:N:P | C:N | C:P | N:P | |
|---|---|---|---|---|
| U | 182:7:1 | 28.2 ± 2.69 ab | 182 ± 29.7 ab | 6.85 ± 1.01 bcd |
| W | 247:8:1 | 31.6 ± 5.76 ab | 247 ± 45.4 a | 8.28 ± 0.768 bcd |
| C | 218:8:1 | 30.0 ± 4.82 ab | 218 ± 21.0 ab | 8.43 ± 1.31 bc |
| N | 223:16:1 | 14.4 ± 1.17 b | 223 ± 30.8 a | 15.5 ± 1.89 a |
| P | 90:3:1 | 33.0 ± 49 a | 90.3 ± 4.65 c | 3.02 ± 0.400 d |
| CN | 165:12:1 | 14.4 ± 1.68 b | 165 ± 26.5 abc | 11.9 ± 1.85 ab |
| CP | 108:3:1 | 37.8 ± 6.00 a | 108 ± 5.93 bc | 3.47 ± 0.675 cd |
Molar C:N:P ratios of soils following 6 years of six resource additions in Fryxell lake basin of the Taylor Valleys in Antarctica.
Treatments abbreviations include: un-amended control (U); water only (W); C as mannitol and water (C); N as equal concentrations NH4+ and NO3− with water (N); P as Na3PO4 and water (P); C, N, and water (CN); C, P, with water (CP). Data are mean (±SEM, n = 8) with different letters indicating differences among treatments (P < 0.05) from ANOVA and Tukey’s HSD.
Table 2
| U | W | C | N | P | CN | CP | |
|---|---|---|---|---|---|---|---|
| pH | 9.7 ± 0.15 | 9.9 ± 0.05 | 9.9 ± 0.12 | 9.5 ± 0.20 | 9.8 ± 0.10 | 9.6 ± 0.17 | 9.8 ± 0.12 |
| EC (ds m−1) | 220 ± 42.1 | 183 ± 45.7 | 224 ± 40.9 | 331 ± 146 | 220 ± 42.0 | 301 ± 97.4 | 266 ± 73.8 |
| Total C (%) | 0.14 ± 0.01 | 0.12 ± 0.01 | 0.14 ± 0.01 | 0.14 ± 0.02 | 0.13 ± 0.01 | 0.15 ± 0.01 | 0.16 ± 0.02 |
| Total N (%) | 0.004 ± 0.001 bc | 0.004 ± 0.001 c | 0.005 ± 0.001 bc | 0.008 ± 0.001 ab | 0.004 ± 0.001 c | 0.01 ± 0.001 a | 0.004 ± 0.001 bc |
| P (μg N g soil−1) | 3.6 ± 0.86 ab | 2.2 ± 0.37 b | 2.7 ± 0.39 b | 2.5 ± 0.23 b | 5.6 ± 0.67 a | 4.2 ± 0.74 ab | 5.8 ± 0.54 a |
| SOC (μg N g soil−1) | 0.03 ± 0.004 ab | 0.03 ± 0004 b | 0.06 ± 0.008 a | 0.03 ± 0.004 ab | 0.03 ± 0.004 b | 0.04 ± 0.007 ab | 0.05 ± 0.005 ab |
| N-NH4+ (μg N g soil−1) | 0.08 ± 0.09 b | 0.27 ± 0.05 b | 0.04 ± 0.01 b | 17 ± 1.6 a | 0.29 ± 0.10 b | 13 ± 2.6 a | 0.05 ± 0.01 b |
| N-NO3− (μg N g soil−1) | 0.64 ± 0.30 c | 1.1 ± 0.57 c | 0.09 ± 0.03 c | 22 ± 4.2 a | 0.77 ± 0.26 c | 10 ± 2.5 b | 1.1 ± 1.1 c |
Soil chemistry following 6 years of resource additions.
For treatment abbreviations see Table 1. Data are mean (±SEM, n = 8) with different letters indicating significant differences among treatments (P < 0.05) based on ANOVA and Tukey’s HSD.
CN Reduced Evenness and Diversity but N Alone Enhanced Richness and Diversity
Bacterial evenness and diversity was reduced following CN additions, while diversity and richness was enhanced by N additions (Figure 1). Specifically, the addition of CN dramatically depressed taxa evenness by at least 20.1% relative to all other treatments, and alpha diversity by at least 24.5% relative to the C, N, and CP treatments. In contrast, N additions stimulated OTU richness by 48.4 and 38.9% relative to bacterial communities in C and P soils, respectively. Diversity also increased by more than 13% in N compared to C, P, and CN treatments. In general, all resource additions reduced the variability surrounding richness and diversity metrics. All community inferences were based on the recovery of 138,458 quality sequences and 1,450 unique OTUs with samples possessing an average sequencing coverage of 98.4% ± 0.21 (mean and SEM). All sequence data were submitted to NCBI and are available as BioProject PRJNA476992.
FIGURE 1
All Nutrient Additions Created Distinct Communities, Especially CN
The CN treatment dramatically influenced bacterial communities, most notably by the separation of communities along axis one, which explained 31.6% of the variation among communities (Figure 2). Further, the addition of any nutrient (i.e., C, N, P, and CP) reduced the variability among communities compared to the untreated control and water only addition along axis 2, which explained 19.6% of the variation. PERMANOVA results supported these interpretations, as communities were distinct among treatments (F = 3.5, R2 = 0.60, P < 0.001, df = 6).
FIGURE 2
Family-Specific Responses to Nutrients
All resource additions, except water, promoted taxonomic shifts in bacterial OTU abundance in 11 families across five phyla (Figure 3). The most pronounced increase in relative recovery occurred in the Micrococcaceae (Actinobacteria) in CN soils where a single bacterium, an Arthrobacter species, was relatively rare (0.06% ± 0.05) in untreated soils but constituted 47% ± 5.6 of the community in CN-amended soils. The only other bloom occurred in the Trueperaceae (Deinococcus), where one OTU was classified as intermediate (0.42% ± 0.09) in untreated but became abundant (9.6% ± 3.8) in CP soils. In general, Actinobacteria were abundant in all soils and accounted for at least 36% of the community composition in all treatments (Supplementary Figure S1), but not all Actinobacteria responded positively to CN. For example, three Actinobacteria families, Solirubrobacteracea, Solirubrobacterales unclassified, and Rubrobacteriaceae, decreased from 2.1- to 4.9-fold in CN compared to all other treatments (Figure 3). CN additions also stimulated Xanthomonadaceae (2.7% ± 1.2, Gammaproteobacteria) and Sphingobacteriaceae (1.8% ± 0.48, Bacteroidetes). With the addition of N, the Nitrosomonadaceae (Betaproteobacteria) increased in recovery 5.2-times allowing the N treatment to have the highest recovery of Betaproteobacteria (3.1% ± 0.17). Annual N additions also enhanced the recovery of Rhodobacteraceae (0.51% ± 0.16, Alphaproteobacteria) and Gemmatimonadaceae (2.7% ± 0.27, Gemmatimonadetes) by at least 1.8-times relative to all other treatments. Both CP and P additions enhanced the recovery of Chitinophagaceae (Bacteroidetes) and Spartobacteria unclassified (Verrucomicrobia) in particular. The recovery of these families was at least 1.5-times higher in P and CP than all other soils and caused the recovery of Bacteroidetes to increase upward of 100% (P = 6.3% ± 2.2, CP = 6.8% 1.8) in both treatments.
FIGURE 3
Deinococcus-Thermus were present in all soils (Supplementary Figure S1), but the addition of CP, C, and P stimulated the recovery of these taxa causing the Deinococcaceae to range from 5.3 to 11% (Figure 3). The addition of C alone promoted different families with Sphingomonadaceae (Alphaproteobacteria) and the Geodermatophilaceae (Actinobacteria) increasing upward of 3.0- and 39-times, respectively, but combined these two taxa only accounted for less than 4.1% of the community in C-amended soils.
Network Community Modeling
Nutrient additions disrupted interactions among community assemblages. Due to the requirement of more than three samples to create reliable community network models with MIC, we only created community network models for two combined treatments (U and W, and P and CP) that were relatively similar (PERMANOVA: U and W, F = 5.8, R2 = 0.59, P = 0.1, df = 1; PW and CPW, F = 1.2, R2 = 0.22, P = 0.5, df = 1). After years of P and CP additions, multiple aspects of the community broke down relative to the untreated and water only soils (Figure 4 and Table 3). For example, the number of significant nodes or taxa, and edges or connections between taxa was 51 and 73% lower, respectively, in P and CP networks. The mean degree (number of connections per node to other nodes) declined twofold from U and W to P and CP networks, and mean path length (number of nodes needed to link any one node to any other in the network) decreased from 3.5 in the U and W to 2.8 in P and CP models. Within the two networks, specific orders were favored and the nodes were often major contributors to the recovery of the community. For example, the Trueperaceae represented 3 nodes in the combined P and CP network and 7.7% relative recovery in the P and the CP communities, but only 1 taxon in U and W network and 0.61% of the recovery in U and W communities. Alternatively, in the U and W network, Phycisphaerae (unclassified; 10 nodes, 2.1% relative recovery), Intrasporangiaceae (2 nodes, 0.13% relative recovery), and Xanthomonadaceae (2 nodes, 0.55% relative recovery) were present but completely absent from the P and CP network. The Spartobacteria (unclassified), Micrococcaceae, and Chitinophagaceae had similar numbers of nodes in all models, but contributed substantially more in abundance in the P and CP communities, 14, 3.2, and 3.1%, respectively. The Solirubrobacteriaceae and Solirubrobacterales (unclassified) consistently contributed to both models with 14 and 10 nodes in the P and CP, and the U and W networks, respectively, and comprised no less than 20% of the relative recovery from either community.
Table 3
| Network topographical parameters | U and W | P and CP | ||
|---|---|---|---|---|
| Nodes, edges (#) | 113, 1321 | 75, 433 | ||
| Mean path length | 2.8 | 3.5 | ||
| Mean degree | 23 | 11.5 | ||
| Mean clustering coefficient | 0.84 | 0.74 | ||
| Density | 0.21 | 0.16 | ||
| Modularity | 0.54 | 0.51 | ||
| U and W | P and CP | |||
| Network taxonomy (order) | Node # | Relative recovery of nodes (%) | Node # | Relative recovery of nodes (%) |
| Acidobacteriaceae | 10 | 14 | 7 | 13 |
| Intrasporangiaceae | 2 | 0.13 | 0 | 0 |
| Micrococcaceae | 1 | 0.06 | 1 | 3.2 |
| Nocardioidaceae | 1 | 0.10 | 0 | 0 |
| Rubrobacteriaceae | 4 | 1.0 | 3 | 0.83 |
| Conexibacteraceae | 1 | 0.40 | 1 | 0.16 |
| Solirubrobacteriaceae | 2 | 1.3 | 4 | 2.0 |
| Solirubrobacterales Unclassified | 8 | 22 | 10 | 18 |
| Chitinophagaceae | 5 | 0.32 | 4 | 3.1 |
| Sphingobacteriaceae | 0 | 0 | 1 | 0.90 |
| Caldilineaceae | 1 | 0.28 | 1 | 1.5 |
| Trueperaceae | 1 | 0.51 | 3 | 7.7 |
| Gemmatimonadaceae | 1 | 0.65 | 1 | 0.16 |
| Phycisphaerae unclassified | 10 | 2.1 | 0 | 0 |
| Sphingomonadaceae | 1 | 0.98 | 2 | 1.8 |
| Alcaligenaceae | 1 | 0.12 | 1 | 0.14 |
| Oxalobacteraceae | 1 | 0.18 | 1 | 0.50 |
| Xanthomonadaceae | 2 | 0.55 | 0 | 0 |
| Spartobacteria unclassified | 4 | 7.3 | 4 | 14 |
Community network model characteristics for soil bacteria in P-amended and control soils.
For treatment abbreviations see Table 1 and for an explanation of network topographical parameters see network community modeling in the results section. For network taxonomy, node # is the total number of nodes or OTUs in a given order for a network. The relative recovery of nodes is the summation of mean relative recoveries of the nodes within the order from P and CP or U and W communities.
FIGURE 4
CN Elevated Soil Respiration
Only the CN resource addition elevated soil respiration 1 week following the nutrient additions. As a result, we observed a 114–234% increase in soil respiration in CN soils compared to all other treatments (one-way ANOVA, df = 6, F = 11.6, P < 0.001, Figure 5). The soil treatments exhibited no differences in temperature at the time of sampling (one-way ANOVA, df = 6, F = 1.46, P = 0.24). The mean temperature of all treatments was 7.2°C ± 0.32 SD (n = 56).
FIGURE 5
Discussion
Stoichiometric shifts of C:N:P in the coldest and driest soils on Earth () alleviated resource limitations, created species-specific bacterial responses, and altered ecosystem processes. Our long-term coupled resource additions dramatically altered soil C:N:P leading to increases in inorganic N and P availability, but only a slight increase in soil organic C content, which was presumably consumed by bacteria. As hypothesized, C, N, CN, and CP additions created unique communities, relative to untreated soils, with CN and N having the most pronounced effect on bacterial species responses. We found that the alleviation of a C and N co-limitation facilitated the dominance of an Arthrobacter species (family, Micrococcaceae) that ultimately elevated soil respiration, and that shifts in C:N ratios may remove nutrient constraints on bacteria enhancing species richness and diversity. Contrary to our hypothesis, the addition of P, even to our relatively P-rich soils (; ), helped create unique communities for all single and coupled resource additions.
Colimitation of CN Facilitates Species Dominance and Enhanced Respiration
The colimitation of organic C, N, and/or P are common in marine and freshwater systems where the abundance of photoautotrophs to organoheterotrophs is often influenced by two or more nutrients (; ; ; ). In soils, co-limitation exists but is harder to identify due to the high levels of bacterial diversity and the wide variety of resource substrates for species to exploit, from extremely labile C substrates to recalcitrant soil organic matter. Over the 6 years of the present study, as CN limitations were eased, (e.g., C:N decreased and N:P increased in CN amended soils), an Arthrobacter species (family = Micrococcaceae, Actinobacteria) went from being rare (0.06% ± 0.05) to dominant (47% ± 5.6). Arthrobacter species are common psychrotrophs found in Adelie penguin guano () and Antarctic epilithic lichens () seeming to capitalize on the localized nutrient-rich penguin feces in the otherwise nutrient poor landscape. Arthrobacter strains can respond quickly to changes in nutrient conditions by breaking dormancy and growing within an hour of the removal of starvation stress (). Compared to temperate Arthrobacter species, Antarctic Arthrobacter possess lower metabolic versatility () but similar genes to many psychrophilic/psychrotolerant species (e.g., cold active hydrolytic enzymes; sigma factors; signal transduction pathways; carotenoid biosynthesis pathway; and genes induced by cold-shock, oxidative, and osmotic stresses (; ). Further, our unclassified Arthrobacter OTU falls within a genus whose members readily decompose almost any algal and cyanobacterial bioproducts, from cyanotoxins () to cellobiose, the final derivative of cellulose utilization (). Thus, the Arthrobacter we recovered is likely a well-adapted psychrophile poised to exploit common bacterial and algal derived C sources when N is available.
The functional consequences of Arthrobacter dominance were easily distinguishable. Often the functional consequences of soil bacterial community change is exceptionally difficult due to discern due to levels of functional redundancy (; ) and large fractions of bacterial diversity being dormant or metabolically inactive at any given time (). In contrast, only in soils where Arthrobacter bacteria achieved dominance did soil respiration dramatically increase (114–234%). Our results are consistent with the findings of who showed that the addition of glucose, glycine, and ammonium stimulated the mineralization of lacustrine detritus and soil organic matter across different geomorphically defined landscapes in Garwood Valley, Antarctica. Even though the link between our dominant species and respiration is implied rather than explicit, dominant or abundant bacteria often contribute proportionally to universal soil processes such as respiration (). Thus, Arthrobacter most likely exploited mannitol and ammonium or nitrate to a greater extent than other species to become more metabolically active. Bacterial competition for essential resources may loosely be classified into two competition categories, scramblers and contesters (). Scramble competition or exploitation competition involves rapid utilization of resources without directly interacting with other bacteria, while contest competition or interference competition involves direct antagonistic interactions between competitors. While both scramblers and contesters occur in most soils, the effects of competition for resources on bacterial taxa are often only implied (; ) and potential interactions among limiting nutrients are often neglected (). In our CN-enriched soils, Arthrobacter is most likely a scrambler, better suited to capitalize on emerging resources. Its rise to dominance resulted in a decline in bacterial evenness and diversity while allowing for the persistence of rare taxa, as evidenced by similar species richness levels exhibited among the different resource treatments.
Inorganic N Opened New Bacterial Niches for Rare Taxa
The removal of ammonium limitations opened new niches for once rare taxa to exploit. With the immense elevation of ammonium levels following N additions, bacterial richness increased upward of 48% in comparison to soils that receive only C or P additions. Higher levels of ammonium increased bacterial richness directly and indirectly by potentially stimulating nitrifying bacteria relying on ammonium and nitrite. We found that ammonium additions enhanced the abundance of two nitrite oxidizing Nitrospira species from the family Nitrosomonadaceae. Antarctic Nitrospira species may also contain amoA sequences () and participate in complete nitrification (). Our findings are consistent with the suggestion that the extreme abiotic severity of the McMurdo Dry Valley soil habitat drives the presence of ammonia oxidizing bacteria (AOB) like Nitrospira (; ). Only after the soils became considerably less harsh with N additions were AOB found in soils. Presumably the increase in nitrite triggered certain species to increase in abundance. For example, six Gemmatimonadaceae species with the ability to reduce nitrite to nitric oxide via the NirK gene (clade II) () increased in abundance at least 1.8-fold under N additions relative to all other treatments. Also, four Rhodobacteraceae taxa with a high affinity for ammonium, characterized by high transcript abundance for ammonium transporters (), increased in abundance as N limitations were lifted. Many of these taxa responding to ammonium and possibly nitrite were rare () with an abundance in untreated soils <0.1% of the total recovery (). Thus, the enhanced availability of reduced forms of N following N-amendments stimulates microbial biodiversity in Antarctic soils.
CP and P Deconstructed Species Assemblages
Lower soil C:P following P additions disrupted community co-occurrence patterns and facilitated new nutrient-related interactions among taxa. Even an increase in soil P, a nutrient that was not predicted to be limiting based on the initial soil C:N:P ratios, influenced species interactions. Excess soil P disrupted potential interactions among community assemblages, as evidenced by more than 50% of the network (i.e., significant species and interactions) disappearing and 14 taxa from three families (i.e., Phycisphaerae unclassified, Intrasporangiaceae, and Xanthomonadaceae) vanishing from the co-occurrence network. However, as P limitations were alleviated a more diffuse and less connected network potentially centered on P availability emerged. For example, two new Trueperaceae species (Deinococcus) were incorporated into the model and the collective abundance of the three Trueperaceae species was 13-times higher in P amended soils. Members of the Trueperaceae family are remarkably resistant to ionizing radiation and able to grow under multiple extreme conditions, including alkaline, moderately saline, and high temperature environments (). Deinococcus-Thermus taxa in general possess a remarkable number of genes encoding for catabolic enzymes including phosphatases (), suggesting that the access to P is potentially linked to radiation resistance and/or helps boost the survival of Trueperaceae in extreme soils. Additionally, multiple shared taxa between our two networks [i.e., Spartobacteria unclassified (Verrucomicrobia), Micrococcaceae (Actinobacteria), and Chitinophagaceae (Bacteroidetes)] increased in abundance under P additions. The abundance of taxa within these families may track the availability of P in soils and water (; ). Even with P additions communities seemed to remain partially reliant on Solirubrobacteriaceae taxa. Members of the Solirubrobacteriaceae family may enhance the weathering of volcanic rocks, which are common in soils occurring on Ross Sea Till (). Despite P additions, Solirubrobacteriaceae taxa were integral to all models constituting upward of 10% of species and 20% of the recovery suggesting that mineral weathering is essential to enhance micronutrient availability under both high and low nutrient conditions. As soils transition from a nutrient-poor to a nutrient-rich state, the excess soil P potentially disrupted interactions among bacterial taxa. For example, the addition of water and organic matter in McMurdo Dry Valley soils caused certain bacteria (members of the and Actinobacteria, Proteobacteria, and Firmicutes) to become active demonstrating a potential taxonomical shift from species adapted to dry oligotrophic to moist copiotrophic conditions (). Thus, future climate-driven changes that ameliorate the current stoichiometric imbalances of the dry valley soils (; ), may deconstruct current bacterial communities and reorganizing them into communities dominated by more copiotrophic taxa. Co-occurrence networks do provide insights into potential interactions among taxa within a community (; ), but to fully understand interactions among bacteria a more direct approach is needed.
Annual Water Additions Failed to Elicit Bacterial Response
Water is necessary for imbalances in stoichiometric nutrient ratios to influence communities; however, our one-time water addition alone was not enough to create lasting effects on bacteria community structure. Frequent water additions do influence bacterial activity across Dry Valleys (; ) where soil moisture is ephemeral and extremely patchy. Surface hydrogeological features such as water tracks in soils (), lateral margins of stream and lake margins (), and discontinuous patches of soils often form in the same location due to wind sheltering and microtopography () enhance bacterial metabolic activity and alter species distributions.
Conclusion
Stoichiometric additions of C, N, and P reduced resource limitations, created species-specific bacterial responses, and in one case altered a fundamental ecosystem process. The most dramatic effects of changes in ecosystem stoichiometry occurring around C and N additions in our initially N limited soils. C as mannitol and N as equal molar concentrations ammonium and nitrate induced an almost twofold reduction in C:N ratios; caused bacterial evenness and diversity to decline; allowed one rare Micrococcaceae, an Arthrobacter species, to dominate community abundance; and elevated soil respiration by 136% compared to untreated soils. N additions alone also reduced C:N ratios, and in contrast to CN-additions, increased species richness and diversity by at least 48 and 13%, respectively, compared to soils receiving a single resource as C or P, and enhanced the abundance of rare taxa dependent on N for metabolism and growth. The addition of P to levels well below the C:P ratio necessary for balanced microbial growth also influenced soil microbial communities. Based on community co-occurrence networks, lower C:P ratios in soils following P and CP additions reduced the number of taxa interacting with one another by 51% and the number of interactions among taxa by 73% relative to untreated and watered soils. Our results suggest that the alleviation of C and N co-limitation facilitated the dominance of single species ultimately altering ecosystem processes; the reduced forms of inorganic N open multiple niches for bacteria to exploit, and excess soil P disrupted interactions within communities.
Statements
Author contributions
ZA, SS, BB, DW, JB, RV, and BA conducted the experiments. ZA, SS, BB, DW, JB, MM, NG, RV, AB, and BA analyzed and interpreted the data. ZA, SS, DW, JB, NG, MM, RV, AB, and BA helped to write and review the manuscript. ZA agrees to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Funding
Our research was supported by the National Science Foundation Office of Polar Programs grants to the McMurdo Long-Term Ecological Research Program ANT-0423595 and OPP-1115245.
Acknowledgments
We thank Ashley Shaw providing valuable comments that improved the paper.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2018.01401/full#supplementary-material
References
1
AanderudZ. T.JonesS. E.FiererN.LennonJ. T. (2015). Resuscitation of the rare biosphere contributes to pulses of ecosystem activity. Front. Microbiol.6:24. 10.3389/fmicb.2015.00024
2
AanderudZ. T.JonesS. E.SchoolmasterD. R.FiererN.LennonJ. T. (2013). Sensitivity of soil respiration and microbial communities to altered snowfall.Soil Biol. Biochem.57217–227. 10.1016/j.soilbio.2012.07.022
3
AdamsB. J.BardgettR. D.AyresE.WallD. H.AislabieJ.BamforthS.et al (2006). Diversity and distribution of Victoria Land biota.Soil Biol. Biochem.383003–3018. 10.1016/j.soilbio.2006.04.030
4
AislabieJ. M.ChhourK. L.SaulD. J.MiyauchiS.AytonJ.PaetzoldR. F.et al (2006). Dominant bacteria in soils of Marble Point and Wright Valley, Victoria Land, Antarctica.Soil Biol. Biochem.383041–3056. 10.1016/j.soilbio.2006.02.018
5
AndersonM. J. (2001). A new method for non-parametric multivariate analysis of variance.Austral Ecol.2632–46.
6
AnejaM. K.SharmaS.FleischmannF.StichS.HellerW.BahnwegG.et al (2006). Microbial colonization of beech and spruce litter - Influence of decomposition site and plant litter species on the diversity of microbial community.Microb. Ecol.52127–135. 10.1007/s00248-006-9006-3
7
AustinA. T.VitousekP. M. (2012). Introduction to a virtual special issue on ecological stoichiometry and global change.New Phytol.196649–651. 10.1111/j.1469-8137.2012.04376
8
BaldrianP.KolarikM.StursovaM.KopeckyJ.ValaskovaV.VetrovskyT.et al (2012). Active and total microbial communities in forest soil are largely different and highly stratified during decomposition.ISME J.6248–258. 10.1038/ismej.2011.95
9
BallB. A.AdamsB. J.BarrettJ. E.WallD. H.VirginiaR. A. (2018). Soil biological responses to C, N and P fertilization in a polar desert of Antarctica.Soil Biol. Biochem.1227–18. 10.1016/j.soilbio.2018.03.025
10
BarberanA.BatesS. T.CasamayorE. O.FiererN. (2012). Using network analysis to explore co-occurrence patterns in soil microbial communities.ISME J.6343–351. 10.1038/ismej.2011.119
11
BarrettJ. E.VirginiaR. A.HopkinsD. W.AislabieJ.BargagliR.BockheimJ. G.et al (2006). Terrestrial ecosystem processes of Victoria Land, Antarctica.Soil Biol. Biochem.383019–3034. 10.1016/j.soilbio.2006.04.041
12
BarrettJ. E.VirginiaR. A.LyonsW. B.McKnightD. M.PriscuJ. C.DoranP. T.et al (2007). Biogeochemical stoichiometry of Antarctic Dry Valley ecosystems.J. Geophys. Res.112:G01010. 10.1029/2005JG000141
13
BastianM.HeymannS.JacomyM. (2009). “Gephi: an open source software for exploring and manipulating networks,” inProceedings of the 3rd International AAAI Conference on Weblogs and Social Media. Available at: http://www.aaai.org/ocs/index.php/ICWSM/09/paper/view/154
14
BedfordB. L.WalbridgeM. R.AldousA. (1999). Patterns in nutrient availability and plant diversity of temperate North American wetlands.Ecology802151–2169. 10.1890/0012-9658(1999)080[2151:PINAAP]2.0.CO;2
15
BobbinkR.HicksK.GallowayJ.SprangerT.AlkemadeR.AshmoreM.et al (2010). Global assessment of nitrogen deposition effects on terrestrial plant diversity: a synthesis.Ecol. Appl.2030–59. 10.1890/08-1140.1
16
BockheimJ. G.McLeodM. (2008). Soil distribution in the McMurdo Dry Valleys, Antarctica.Geoderma14443–49. 10.1016/j.geoderma.2007.10.015
17
BockheimJ. G.PrenticeM. L.McLeodM. (2008). Distribution of glacial deposits, soils, and permafrost in Taylor Valley, Antarctica.Arctic Antarct. Alp. Res.40279–286. 10.1657/1523-0430(06-057)[BOCKHEIM]2.0.CO;2
18
BouvyM.TroussellierM.GotP.ArfiR. (2004). Bacterioplankton responses to bottom-up and top-down controls in a West African reservoir (Selingue, Mali).Aquat. Microb. Ecol.34301–307. 10.3354/ame035301
19
BuelowH. N.WinterA. S.Van HornD. J.BarrettJ. E.GooseffM. N.SchwartzE.et al (2016). Microbial community responses to increased water and organic matter in the arid soils of the McMurdo Dry Valleys, Antarctica.Front. Microbiol.7:1040. 10.3389/fmicb.2016.01040
20
BurkinsM. B.VirginiaR. A.ChamberlainC. P.WallD. H. (2000). Origin and distribution of soil organic matter in Taylor Valley, Antarctica.Ecology812377–2391. 10.1890/0012-9658
21
CaporasoJ. G.LauberC. L.WaltersW. A.Berg-LyonsD.LozuponeC. A.TurnbaughP. J.et al (2010). Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample.Proc. Natl. Acad. Sci. U.S.A.1084516–4522. 10.1073/pnas.1000080107
22
CaryS. C.McDonaldI. R.BarrettJ. E.CowanD. A. (2010). On the rocks: the microbiology of Antarctic Dry Valley soils.Nat. Rev. Microbiol.8129–138. 10.1038/nrmicro2281
23
CherifM.LoreauM. (2007). Stoichiometric constraints on resource use, competitive interactions, and elemental cycling in microbial decomposers.Am. Nat.169709–724. 10.1086/516844
24
ClarkC. M.TilmanD. (2008). Loss of plant species after chronic low-level nitrogen deposition to prairie grasslands.Nature451712–715. 10.1038/nature06503
25
ClevelandC. C.LiptzinD. (2007). C : N : P stoichiometry in soil: is there a “Redfield ratio” for the microbial biomass?Biogeochemistry85235–252. 10.1007/s10533-007-9132-0
26
CockellC. S.KellyL. C.MarteinssonV. (2013). Actinobacteria-an ancient phylum active in volcanic rock weathering.Geomicrobiol. J.30706–720. 10.1080/01490451.2012.758196
27
ColeJ. R.WangQ.CardenasE.FishJ.ChaiB.FarrisR. J.et al (2009). The ribosomal database project: improved alignments and new tools for rRNA analysis.Nucleic Acids Res.37D141–D145. 10.1093/nar/gkn879
28
CowanD. A.MakhalanyaneT. P.DennisP. G.HopkinsD. W. (2014). Microbial ecology and biogeochemistry of continental Antarctic soils.Front. Microbiol.5:154. 10.3389/fmicb.2014.00154
29
CsardiG.NepuszT. (2006). The igraph software package for complex network research.InterJournal Complex Syst.16951–9.
30
DaimsH.LebedevaE. V.PjevacP.HanP.HerboldC.AlbertsenM.et al (2015). Complete nitrification by Nitrospira bacteria.Nature528504–509. 10.1038/nature16461
31
DalyM. J.GaidamakovaE. K.MatrosovaV. Y.KiangJ. G.FukumotoR.LeeD. Y.et al (2010). Small-molecule antioxidant proteome-shields in Deinococcus radiodurans.PLoS One5:e12570. 10.1371/journal.pone.0012570
32
DecleyreH.HeylenK.TytgatB.WillemsA. (2016). Highly diverse nirK genes comprise two major clades that harbour ammonium-producing denitrifiers.BMC Genomics17:155. 10.1186/s12864-016-2465-0
33
DoranP. T.McKayC. P.ClowG. D.DanaG. L.FountainA. G. A.NylenT.et al (2002). Valley floor climate observations from the McMurdo Dry Valleys, Antarctica, 1986-2000.J. Geophys. Res.1074772–4784. 10.1029/2001JD002045
34
DsouzaM.TaylorM. W.TurnerS. J.AislabieJ. (2015). Genomic and phenotypic insights into the ecology of Arthrobacter from Antarctic soils.BMC Genomics16:18. 10.1186/s12864-015-1220-2
35
EdgarR. C.HaasB. J.ClementeJ. C.QuinceC.KnightR. (2011). UCHIME improves sensitivity and speed of chimera detection.Bioinformatics272194–2200. 10.1093/bioinformatics/btr381
36
ElserJ. J.HamiltonA. (2007). Stoichiometry and the new biology - The future is now.PLoS Biol. 5:1403–1405. 10.1371/journal.pbio.0050181
37
ElserJ. J.SternerR. W.GorokhovaE.FaganW. F.MarkowT. A.CotnerJ. B. (2000). Biological stoichiometry from genes to ecosystems.Ecol. Lett.3540–550. 10.1046/j.1461-0248.2000.00185.x
38
FaninN.HattenschwilerS.SchimannH.FrominN. (2015). Interactive effects of C, N and P fertilization on soil microbial community structure and function in an Amazonian rain forest.Funct. Ecol.29140–150. 10.1111/1365-2435.12329
39
FiererN.LeffJ. W.AdamsB. J.NielsenU. N.BatesS. T.LauberC. L.et al (2012). Cross-biome metagenomic analyses of soil microbial communities and their functional attributes.Proc. Natl. Acad. Sci. U.S.A.10921390–21395. 10.1073/pnas.1215210110
40
FiererN.StricklandM. S.LiptzinD.BradfordM. A.ClevelandC. C. (2009). Global patterns in belowground communities.Ecol. Lett.121238–1249. 10.1111/j.1461-0248.2009.01360
41
FilosiM.VisintainerR.AlbaneseD. (2017). Minerva: Maximal Information-Based Nonparametric Exploration for Variable Analysis. R Package Version 1.4.7. Available at: https://CRAN.R-project.org/package=minerva
42
FountainA. G.LyonsW. B.BurkinsM. B.DanaG. L.DoranP. T.LewisK. J.et al (1999). Physical controls on the Taylor Valley Ecosystem, Antarctica.BioScience49961–971. 10.1525/bisi.1999.49.12.961
43
FreedmanZ. B.UpchurchR. A.ZakD. R. (2016). Microbial potential for ecosystem N loss is increased by experimental N deposition.PLoS One11:e0164531. 10.1371/journal.pone.0164531
44
FreedmanZ. B.ZakD. R. (2015). Atmospheric N deposition alters connectance, but not functional potential among saprotrophic bacterial communities.Mol. Ecol.243170–3180. 10.1111/mec.13224
45
FreilichS.ZareckiR.EilamO.SegalE. S.HenryC. S.KupiecM.et al (2011). Competitive and cooperative metabolic interactions in bacterial communities.Nat. Commun.2:589. 10.1038/ncomms1597
46
GeyerK. M.AltrichterA. E.Takacs-VesbachC. D.Van HornD. J.GooseffM. N.BarrettJ. E. (2014). Bacterial community composition of divergent soil habitats in a polar desert.FEMS Microbiol. Ecol.89490–494. 10.1111/1574-6941.12306
47
GooseffM. N.BarrettJ. E.AdamsB. J.DoranP. T.FountainA. G.LyonsW. B. (2017). Decadal ecosystem response to an anomalous melt season in a polar desert in Antarctica.Nat. Ecol. Evol.11334–1338. 10.1038/s41559-017-0253-0
48
GooseffM. N.BarrettJ. E.DoranP. T.FountainA. G.LyonsW. B.ParsonsA. N.et al (2003). Snow-patch influence on soil biogeochemical processes and invertebrate distribution in the McMurdo Dry Valleys, Antarctica.Arctic Antarct. Alp. Res.3591–99. 10.1657/1523-0430(2003)035[0091:SPIOSB]2.0.CO;2
49
HartmanW. H.RichardsonC. J. (2013). Differential nutrient limitation of soil microbial biomass and metabolic quotients (qCO2): Is there a biological stoichiometry of soil microbes?PLoS One8:e57127. 10.1371/journal.pone.0057127
50
HeckyR. E.CampbellP.HendzelL. L. (1993). The stoichiometry of carbon, nitrogen, and phosphorus in particulate matter of lakes and oceans.Limnol. Oceanogr.38709–724. 10.4319/lo.1993.38.4.0709
51
HeindelR. C.SpickardA. M.VirginiaR. A. (2017). Landscape-scale soil phosphorus variability in the McMurdo Dry Valleys, Antarctica.Antarct. Sci.29252–263. 10.1017/S0954102016000742
52
HernandezD. L.HobbieS. E. (2010). The effects of substrate composition, quantity, and diversity on microbial activity.Plant Soil335397–411. 10.1007/s11104-010-0428-9
53
HibbingM. E.FuquaC.ParsekM. R.PetersonS. B. (2010). Bacterial competition: surviving and thriving in the microbial jungle.Nat. Rev. Microbiol.815–25. 10.1038/nrmicro2259
54
HopkinsD. W.SparrowA. D.ElberlingB.GregorichE. G.NovisP. M.GreenfieldL. G.et al (2006). Carbon, nitrogen and temperature controls on microbial activity in soils from an Antarctic dry valley.Soil Biol. Biochem.383130–3140. 10.1016/j.soilbio.2006.01.012
55
HowarthR. W.MarinoR. (2006). Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: Evolving views over three decades.Limnol. Oceanogr.51364–376. 10.4319/lo.2006.51.1_part_2.0364
56
IvanovaN.RohdeC.MunkC.NolanM.LucasS.Del RioT. G.et al (2011). Complete genome sequence of Truepera radiovictrix type strain (RQ-24(T).Stand. Genomic Sci.491–99. 10.4056/sigs.1563919
57
JobbagyE. G.JacksonR. B. (2000). The vertical distribution of soil organic carbon and its relation to climate and vegetation.Ecol. Appl.10423–436. 10.1890/1051-0761(2000)010[0423:TVDOSO]2.0.CO;2
58
JohnsonW. H.DouglasM. R.LewisJ. A.StueckerT. N.CarboneroF. G.AustinB. J.et al (2017). Do biofilm communities respond to the chemical signatures of fracking? A test involving streams in North-central Arkansas.BMC Microbiol.17:12. 10.1186/s12866-017-0926-5
59
KeiblingerK. M.HallE. K.WanekW.SzukicsU.HammerleI.EllersdorferG.et al (2010). The effect of resource quantity and resource stoichiometry on microbial carbon-use-efficiency.FEMS Microbiol. Ecol.73430–440. 10.1111/j.1574-6941.2010.00912
60
LawtonL. A.WelgamageA.ManageP. M.EdwardsC. (2011). Novel bacterial strains for the removal of microcystins from drinking water.Water Sci. Technol.631137–1142. 10.2166/wst.2011.352
61
LeeC. K.LozierJ. D.BarbierB. A.MillsN. J.BottosE. M.CaryS. C. (2011). The Inter-Valley Soil Comparative Survey: the ecology of Dry Valley edaphic microbial communities.ISME J.61046–1057. 10.1038/ismej.2011.170
62
LennonJ. T.JonesS. E. (2011). Microbial seed banks: the ecological and evolutionary implications of dormancy.Nat. Rev. Microbiol.9119–130. 10.1038/nrmicro2504
63
LevyJ. S.FountainA. G.GooseffM. N.BarrettJ. E.VantreeseR.WelchK. A.et al (2014). Water track modification of soil ecosystems in the Lake Hoare basin, Taylor Valley, Antarctica.Antarct. Sci.26153–162. 10.1017/S095410201300045
64
LozuponeC. A.KnightR. (2007). Global patterns in bacterial diversity.Proc. Natl. Acad. Sci. U.S.A.10411436–11440. 10.1073/pnas.0611525104
65
MagalhaesC. M.MachadoA.Frank-FahleB.LeeC. K.CarS. C. (2014). The ecological dichotomy of ammonia-oxidizing archaea and bacteria in the hyper-arid soils of the Antarctic Dry Valleys.Front. Microbiol.5:515. 10.3389/fmicb.2014.00515
66
MarschnerP.KandelerE.MarschnerB. (2003). Structure and function of the soil microbial community in a long-term fertilizer experiment.Soil Biol. Biochem.35453–461. 10.1016/S0038-0717(02)00297-3
67
MichalskiG.BockheimJ. G.KendallC.ThiemensM. (2005). Isotopic composition of Antarctic Dry Valley nitrate: implications for NOy sources and cycling in Antarctica.Geophys. Res. Lett.32:L13817. 10.1029/2004GL022121
68
MikiT.YokokawaT.MatsuiK. (2014). Biodiversity and multifunctionality in a microbial community: a novel theoretical approach to quantify functional redundancy.Proc. Biol. Sci.281:20132498. 10.1098/rspb.2013.2498
69
MooshammerM.WanekW.SchneckerJ.WildB.LeitnerS.HofhanslF.et al (2012). Stoichiometric controls of nitrogen and phosphorus cycling in decomposing beech leaf litter.Ecology93770–782. 10.1890/11-0721.1
70
MuscarellaM. E.BirdK. C.LarsenM. L.PlacellaS. A.LennonJ. T. (2014). Phosphors resource heterogeneity in microbial food webs.Aquat. Microb. Ecol.73259–272. 10.3354/ame01722
71
NielsenU. N.BallB. A. (2014). Impacts of altered precipitation regimes on soil communities and biogeochemistry in arid and semi-arid ecosystems.Glob. Change Biol.211407–1421. 10.1111/gcb.12789
72
OksanenJ.GuillaumeF.KindtB.KindtR.LegendreP.MinchinP. R.et al (2013). Vegan: Community Ecology Package. Available at: https://cran.r-project.org/web/packages/vegan/index.html
73
OrwinK. H.WardleD. A.GreenfieldL. G. (2006). Ecological consequences of carbon substrate identity and diversity in a laboratory study.Ecology87580–593. 10.1890/05-0383
74
Pedros-AlioC. (2012). The rare bacterial biosphere.Ann. Rev. Mar. Sci.4449–466. 10.1146/annurev-marine-120710-100948
75
PfreundtU.SpunginD.BonnetS.Berman-FrankI.HessW. R. (2016). Global analysis of gene expression dynamics within the marine microbial community during the VAHINE mesocosm experiment in the southwest Pacific.Biogeosciences134135–4149. 10.5194/bg-13-4135-2016
76
PruesseE.QuastC.KnittelK.FuchsB. M.LudwigW. G.PepliesJ.et al (2007). SILVA: A comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB.Nucleic Acids Res.357188–7196. 10.1093/nar/gkm864
77
QuinceC.LanzenA.DavenportR. J.TurnbaughP. J. (2011). Removing noise from pyrosequenced amplicons.BMC Bioinformatics1210.1186/1471-2105-12-38
78
R Development Core Team (2013). R: A Language and Environment for Statistical Computing.Vienna: R Foundation for Statistical Computing.
79
RamirezK. S.LauberC. L.KnightR.BradfordM. A.FiererN. (2010). Consistent effects of nitrogen fertilization on soil bacterial communities in contrasting systems.Ecology913463–3470. 10.1890/10-0426.1
80
RedfieldA. C. (1958). The biological control of chemical factors in the environment.Am. Sci.46205–221.
81
ReinersW. A. (1986). Complementary models for ecosystems.Am. Nat.46205–221. 10.1086/284467
82
RenC. J.ZhaoF. Z.KangD.YangG. H.HanX.TongH. G.et al (2016). Linkages of C:N:P stoichiometry and bacterial community in soil following afforestation of former farmland.Forest Ecol. Manag.37659–66. 10.1016/j.foreco.2016.06.004
83
ReshefD. N.ReshefH. K.FinucaneS. R.GrossmanG.McVeanP. J.TurnbaughP. J.et al (2011). Detecting novel associations in large data sets.Science3341518–1524. 10.1126/science.1205438
84
SaitoM. A.GoepfertT. J.RittJ. T. (2008). Some thoughts on the concept of colimitation: three definitions and the importance of bioavailability.Limnol. Oceanogr.53276–290. 10.4319/lo.2008.53.1.0276
85
SchellenbergerS.DrakeH. L.KolbS. (2011). Functionally redundant cellobiose-degrading soil bacteria respond differentially to oxygen.Appl. Environ. Microbiol.776043–6048. 10.1128/AEM.00564-11
86
SchlossP. D.WestcottS. L.RyabinT.HallJ. R.HartmannM.HollisterE. B.et al (2009). Introducing mothur: Open-source, platform-independent, community-supported software for describing and comparing microbial communities.Appl. Environ. Microbiol.757537–7541. 10.1128/AEM.01541-09
87
SchwartzE.Van HornD. J.BuelowH. N.OkieJ. G.GooseffM. N.BarrettJ. E.et al (2014). Characterization of growing bacterial populations in McMurdo Dry Valley soils through stable isotope probing with 18O-water.FEMS Microb. Ecol.89415–425. 10.1111/1574-6941.12349
88
SelbmannL.ZucconiL.RuisiS.GrubeM.CardinaleM.OnofriS. (2010). Culturable bacteria associated with Antarctic lichens: affiliation and psychrotolerance.Polar Biol.3371–83. 10.1007/s00300-009-0686-2
89
SinghR. N.GabaS.YadavA. N.GaurP.GulatiS.KaushikR.et al (2016). First high quality draft genome sequence of a plant growth promoting and cold active enzyme producing psychrotrophic Arthrobacter agilis strain L77.Stand. Genomic Sci.11:54. 10.1186/s40793-016-0176-4
90
SoginM. L.MorrisonH. G.HuberJ. A.WelchM. D.HuseS. M.NealP. R.et al (2006). Microbial diversity in the deep sea and the unexplored ”rare biosphere.”Proc. Natl. Acad. Sci. U.S.A.3212115–12120. 10.1073/pnas.060512710
91
SolyanikovaI. P.SuzinaN. E.EgozarianN. S.PolivtsevaV. N.PrisyazhnayaN. V.El-RegistanG. I.et al (2017). The response of soil Arthrobacter agilis lush13 to changing conditions: transition between vegetative and dormant state.J. Environ. Sci. Health B52745–751. 10.1080/03601234.2017.1356665
92
StarkJ. M.FirestoneM. K. (1995). Mechanisms for soil moisture effects on the activity of nitrifying bacteria.Appl. Environ. Microbiol.61218–221.
93
TianH. Q.ChenG. S.ZhangC.MelilloJ. M.HallC. A. S. (2010). Pattern and variation of C:N:P ratios in China’s soils: a synthesis of observational data.Biogeochemistry98139–151. 10.1007/s10533-009-9382-0
94
TischerA.WerischM.DobbelinF.CamenzindT.RilligM. C.PotthastK.et al (2015). Above- and belowground linkages of a nitrogen and phosphorus co-limited tropicalmountain pasture system-responses to nutrient enrichment.Plant Soil391333–352. 10.1007/s11104-015-2431-7
95
VirginiaR. A.WallD. H. (1999). How soils structure communities in the Antarctic dry valleys.Bioscience49973–983. 10.1525/bisi.1999.49.12.973
96
WangR.SunQ. Q.WangY.LiuQ. F.DuL. L.ZhaoM.et al (2017). Temperature sensitivity of soil respiration: synthetic effects of nitrogen and phosphorus fertilization on Chinese Loess Plateau.Sci. Total Environ.5741665–1673. 10.1016/j.scitotenv.2016.09.001
97
WarnesG. R.BolkerB.LodewijkB.GentlemanR.LiawW. H. A.LumleyT.et al (2014). gplots: Various R Programming Tools for Plotting Data Package. Available at: https://cran.r-project.org/package=gplots
98
XuX. F.ThorntonP. E.PostW. M. (2013). A global analysis of soil microbial biomass carbon, nitrogen and phosphorus in terrestrial ecosystems.Glob. Ecol. Biogeogr.22737–749. 10.1111/geb.12029
99
YinB.CrowleyD.SparovekG.De MeloW. J.BornemanJ. (2000). Bacterial functional redundancy along a soil reclamation gradient.Appl. Environ. Microbiol.664361–4365. 10.1128/AEM.66.10.4361-4365.2000
100
ZdanowskiM. K.WeglenskiP.GolikP.SasinJ. M.BorsukP.ZmudaM. J.et al (2004). Bacterial diversity in Adelie penguin, Pygoscelis adeliae, guano: molecular and morpho-physiological approaches.FEMS Microbiol. Ecol.50163–173. 10.1016/j.femsec.2004.06.012
101
Zechmeister-BoltensternS.KeiblingerK. M.MooshammerM.PenuelasJ.RichterA.SardansJ.et al (2015). The application of ecological stoichiometry to plant-microbial-soil organic matter transformations.Ecol. Monograph.85133–155. 10.1890/14-0777.1
102
ZeglinL. H.DahmC. N.BarrettJ. E.GooseffM. N.FitpatrickS. K.Takacs-VesbachC. D. (2011). Bacterial community structure along moisture gradients in the Parafluvial sediments of two ephemeral desert streams.Microb. Ecol.61543–556. 10.1007/s00248-010-9782-7
103
ZhouJ. Z.XiaB. C.TrevesD. S.WuL. Y.MarshT. L.O’NeillR. V.et al (2002). Spatial and resource factors influencing high microbial diversity in soil.Appl. Environ. Microbiol.68326–334. 10.1128/AEM.68.1.326-334.2002
Summary
Keywords
ecological stoichiometry, Lake Fryxell Basin, McMurdo Dry Valleys, network community modeling, nutrient colimitation, Solirubrobacteriaceae
Citation
Aanderud ZT, Saurey S, Ball BA, Wall DH, Barrett JE, Muscarella ME, Griffin NA, Virginia RA, Barberán A and Adams BJ (2018) Stoichiometric Shifts in Soil C:N:P Promote Bacterial Taxa Dominance, Maintain Biodiversity, and Deconstruct Community Assemblages. Front. Microbiol. 9:1401. doi: 10.3389/fmicb.2018.01401
Received
21 February 2018
Accepted
07 June 2018
Published
03 July 2018
Volume
9 - 2018
Edited by
Thulani Peter Makhalanyane, University of Pretoria, South Africa
Reviewed by
David Anthony Pearce, Northumbria University, United Kingdom; Richard Allen White III, RAW Molecular Systems (RMS) LLC, United States
Updates
Copyright
© 2018 Aanderud, Saurey, Ball, Wall, Barrett, Muscarella, Griffin, Virginia, Barberán and Adams.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Zachary T. Aanderud, zachary_aanderud@byu.edu
This article was submitted to Extreme Microbiology, a section of the journal Frontiers in Microbiology
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.